close

Вход

Забыли?

вход по аккаунту

?

Microwave-induced combustion synthesis of aluminum oxide-titanium carbide powder

код для вставкиСкачать
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of th e
copy subm itted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted.
Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy.
Higher quality 6” x 9" black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
Bell & Howell Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MICROWAVE-INDUCED COMBUSTION SYNTHESIS OF AI20 3-TiC POWDER
By
DUANGDUEN ATONG
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
2000
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 9997771
___
UMI
UMI Microform 9997771
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To my dearest twin sister, my very best friend, Dr. Duangdao Atong, whose love and
support has always been there since the day we were bora.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
I would like to thank a number of people who have contributed to my education and
my well-being throughout my graduate career. First and foremost, I would like to express
my sincere gratitude to my advisor, Dr. David E. Clark, for his guidance, advice,
encouragement, and continuous support throughout my program here in the University of
Florida. I am very grateful for what he helped to show me how to succeed in the real
professional world.
I also would like to thank my committee members, Dr. Hassan El-Shall, Dr John
Mecholsky, Dr. Eric Enholm and Dr. E. Dow Whitney for taking the time to guide me along
this educational challenge. A special thanks goes out to Dr. El-Shall for his assistance and
suggestion regarding statistical analysis of the results.
I also would like to express my appreciation to my present and former colleagues in
Dr. Clark’s research group, including Attapon Boonyapiwat, Greg Darby, Mark Moore,
Kristie Leiser, Robert DiFiore, and Diane Folz for their friendship and research assistance.
In particular I would like to thank Attapon for his help since the first day I came to
Gainesville and also for answering my endless questions about microwave technology.
Many thanks are due to my high school and college friends, including Ning, Ong,
Wut, Raj, Joe, Tak, Tar. I would like to thank Joe for technical discussions and suggestions.
I would like to thank Tar for his moral support over the past few years. In addition, I would
like to thank my long-time friend Tak for his friendship and support especially during my
first couple years in the State.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I would most especially like to thank Beau, Viboon SricharoenchaikuJ, for his love,
understanding, and for his very own way of making things seen a lot easier and lighter. I
also wish to thank him for being so patient and for always been there emotionally and
scientifically throughout the very long days and bad temper that went with those long days.
His kind and encouraging words throughout this time helped to keep me focused on my
goals, and helped renew my own faith in what I could accomplish.
Finally, I would like to express my feeling and gratefulness toward my family who
always been there for me throughout these years. My beloved parents, Mr. Sotnsak and Mrs.
Nipa Atong, always did their best to encourage me since day one. Knowing that my mom
and dad believed in me always help me get through bad days. I would also like to thank
them for their unconditional love, support, and most especially for their urging not to give
up. Also I would like to thank them for being so patient and understanding especially during
these last couple years.
Saving the best for last, I give much love and many thanks to my dearest twin sister,
my very best friend, Dr. Duangdao Atong, for her endless love, support, and understanding
since the day we were bom. Without her, I would never have been here in the first place. I
thank her for being my big sister who always look after me in any ways. I definitely thank
her for always encourage me and believed in me even when I did not believe in myself. Also
I am forever grateful for her patience and the sacrifice that she has made especially during
this last long year of my education. It is the love of her that give me strenght and confidence
to overcome the obstacles I have met. None of this would have been possible without her
love and support. I love her more than words could ever explained.
The financial support for my graduate study was provided by the Royal Thai
Goverment.
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
pagg
ACKNOWLEDGMENTS..................................................................................................ui
LIST OF TABLES.............................................................................................................vii
LIST OF FIGURES........................................................................................................... ix
ABSTRACT.................................................................................................................... xiv
CHAPTER
1 INTRODUCTION..........................................................................................................1
2 BACKGROUND........................................................................................................... 7
2.1 Combustion Synthesis...............................................................................................7
2.1.1 History of Combustion Synthesis........................................................................7
2.1.2 Introduction to SHS............................................................................................9
2.1,2a Terminology..................................................................................................9
2.1.2b Advantages................................................................................................. 11
2.1.2c Types of reactions........................................................................................11
2.1.2d Combustion Products...................................................................................13
2.1.3 Thermodynamics of Combustion Synthesis.......................................................16
2.1.4 Combustion Wave Theory.................................................................................19
2.1.4a Combustion wave structure.......................................................................... 19
2.1.4b Stability of combustion wave...................................................................... 24
2.1.5 Important Reaction Parameters in Combustion Synthesis Reactions.................. 27
2.1.5a Stoichiometric ratio..................................................................................... 27
2.1.5b Green density..............................................................................................28
2.1.5c Sample diameter.......................................................................................... 29
2.1.5d Particle size, particle size distribution, and particle shape............................29
2.1.5e Heating rate and ignition power................................................................... 30
2.1,5f Initial, ignition and combustion temperature................................................31
2.2 Microwave Heating.................................................................................................33
2.2.1 Microwave - Material Interaction..................................................................... 35
2.2.1a Dielectric behavior...................................................................................... 36
2.2.lb Microwave power dissipation...................................................................... 44
2.2.2 Microwave Processing of Materials.................................................................. 48
v
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
2.3 Microwave Ignition and Combustion.......................................................................52
3 EXPERIMENTAL PROCEDURE............................................................................... 65
3.1 Effects of Reaction Parameters Studies....................................................................65
3.1.1 Screening Experiments: Plackett-Burman Design..............................................66
3.1.2 Sequential Experiments : Central Composite Design......................................... 71
3.1.3 Sample Preparation...........................................................................................73
3.1.4 Experimental Setup...........................................................................................74
3.1.5. Sample Characterization.................................................................................. 75
3.1.5a X-ray diffraction analysis (XRD)................................................................ 76
3.1.5b Pycnometry.................................................................................................76
3.1.5c BET............................................................................................................ 76
3.1,5d The Coulter principle.................................................................................. 77
3.1.5e Energy Dispersive X-ray spectroscopy (EDS)............................................. 77
3.1.5f Scanning electron microscopy (SEM).......................................................... 77
3.2. Mechanism Studies.................................................................................................78
4 RESULTS AND DISCUSSIONS................................................................................. 83
4.1 Effects of Reaction Parameters................................................................................ 83
4.1.1 Screening Experiments: Plackett-Burman Design.............................................. 83
4.1.1a Data collection............................................................................................83
4.1. lb Data analysis...............................................................................................90
4.1.2 Sequential Experiments : Central Composite Design......................................... 99
4.1,2a Factor settings and design matrix................................................................99
4.1.2b Data collection.......................................................................................... 100
4.1.2c Data analysis............................................................................................. 120
4.1.3 Summary.........................................................................................................161
4.2 Mechanism Studies................................................................................................163
4.2.1 Microwave heating......................................................................................... 164
4.2.2 Conventional Heating..................................................................................... 189
4.2.3 Summary.................................................................................................. 202
5 SUMMARY AND CONCLUSIONS......................................................................... 206
APPENDIX A HISTORY OF COMBUSTION SYNTHESIS....................................... 211
APPENDIX B CENTRAL COMPOSITE DESIGN...................................................... 214
APPENDIX C PARTICLE SIZE DISTRIBUTION OF THE REACTANTS AND
POWDER PRODUCT........................................................................... 217
LIST OF REFERENCES.................................................................................................222
BIOGRAPHICAL SKETCH...........................................................................................234
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Eage
Table 2.1 Key Characteristics, Features, and Benefits of Microwave Processing [Bin92,
Cla96, Kat92, Sut89]........................................................................................49
Table 2.2 Microwave Processing Challenges [Cla96, Sut89, Sut92]...................................50
Table 3.3 Actual Factor Levels for a Sixteenth-Run Screening Design for the MicrowaveInduced Combustion Synthesis of ALC^-TiC Powder Study............................. 69
Table 3.4 Central Composite Design for 4 Factors.............................................................72
Table 3.5 Powder Characterization....................................................................................74
Table 3.6 All Combinations of Mixtures for Each Composition.......................................... 80
Table 4.1 Ignition and Combustion Behavior of the Sixteenth-runs Plackett-Burman Design
of the AI2O3 - TiC Powders............................................................................. 84
Table 4.2 Characteristics of the Sixteenth-runs Plackett-Burman Design of the AI2 O3 - TiC
Powders...........................................................................................................89
Table 4.3 The Effects of the Reaction Parameters on Each Investigated Response Value ....95
Table 4.4 Relative Significance of Reaction Parameters of AI2O3 -TiC Powders................98
Table 4.5 Factors and Investigated Levels of Thirtieth-Runs CCD...................................... 101
Table 4.6 Actual Factor Levels for a Thirtieth-Run Central Composite Design for the
Microwave-Induced Combustion Synthesis of AkCVTiC Powders Study......... 102
Table 4.7 Compositions Studied in a Thirtieth-Run Central Composite Design.................... 103
Table 4.8 Ignition and Combustion Behavior of theThirtieth-Run CCD Design for MH
Ignited AfeCVTiC Powders...........................
104
Table 4.9 Ignition and Combustion Behavior of die Thirtieth-Run CCD design for MHH
Ignited AUC^-TiC Powders.............................................................................. 105
Table 4.10 Characteristics of the Thirtieth-Run CCD design for MH Ignited AhCh-TiC
Powders........................................................................................................... 110
vu
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.11 Characteristics of the Thirtieth-Run CCD design for MHH Ignited AfcC^-TiC
Powders........................................................................................................... I l l
Table 4.12 Analysis for the Ignition Time Responses for Both MH and MHH Experiments
........................................................................................................................123
Table 4.13 Analysis for the Ignition Temperature Responses for Both MH and MHH
Experiments......................................................................................................124
Table 4.14 Analysis for the Combustion Time Responses for Both MH and MHH
Experiments..................................................................................................... 125
Table 4.15 Analysis for the Combustion Temperature Responses for Both MH and MHH
Experiments..................................................................................................... 126
Table 4.16 Analysis for the Powder Density Responses for Both MH and MHH Experiments
........................................................................................................................127
Table 4.17 Analysis for the Particle Size Responses for Both MH and MHH Experiments ..128
Table 4.18 Analysis for the Specific Surface Area Responses for Both MH and MHH
Experiments......................................................................................................129
Table 4.19 Analysis for the Agglomeration Size Responses for Both MH and MHH
Experiments......................................................................................................130
Table 4.20 Relative Significance of Reaction Parameters on all Investigated Responses for
the MH Experiment.......................................................................................... 133
Table 4.21 Relative Significance of Reaction Parameters on all Investigated Responses for
the MHH Experiment....................................................................................... 134
Table 4.22 Summary of the Regression Models for Predicting the Ignition Behavior and
Characteristic of MH Ignited Ah0 3 -TiC Powders.............................................135
Table 4.23 Summary of the Regression Models for Predicting the Ignition Behavior and
Characteristic of MHH Ignited Ah0 3 -TiC Powders..........................................136
Table 4.24 The amounts of reactants used in each mixtures for four compositions studied... 165
Table 4.25 Chemical Composition Analysis Obtained from XRD for the 3Ti02+3C+4Al
Samples Quenched after Being Heated to Various Temperatures in DTA.......... 202
Table B1 Central Composite Design for 4 factors..............................................................216
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Page
Figure 2.1 A schematic of a SHS reaction...........................................................................10
Figure 2.2 Schematic representation of the adiabatic temperature calculation [Holt86]
16
Figure 2.3 Relationship of H^g/ACp^g and T«d [Adapted from Mun88]........................... 19
Figure 2.4 Combustion wave structure for homogeneous combustion (q is degree of
conversion, <|>is the rate of heat generation) [Mer74].........................................20
Figure 2.5 Combustion wave structure for heterogeneous combustion (q is degree of
conversion, <j>is the rate of heat generation) [Mer74].........................................23
Figure 2.6 Schematic representation of combustion wave propagation: (a) stable, (b)
oscillatory, (c) spin [Adapted from Mun88], and (d) burnout.............................25
Figure 2.7 Schematic representation of the temperature profile associated with the passage of
a combustion front........................................................................................... 31
Figure 2.8 Electromagnetic spectrum and frequencies used in microwave processing [Sut93].
....................................................................................................................... 34
Figure 2.9 Interaction of microwaves with materials [Sut89]...............................................35
Figure 2.10 Schematic of the different polarization mechanisms, (a) Electronic, (b) Atomic or
ionic, (c) High frequency oscillatory dipoles, (d) Low frequency cation dipole, (e)
Interfacial space charge polarization at electrodes, (f) Interfacial polarization at
heterogeneities [Hen90]................................................................................... 37
Figure 2.11 Frequency dependence of the polarization mechanisms m dielectrics, (a)
Contribution to the charging constant (representative values of k’). (b)
Contribution to die loss angle (representative values of tan§) [Hen90]............... 38
Figure 2.12 Vector diagram of charging, loss and total currents in a dielectric [Hen90].......42
Figure 2.13 Qualitative representation of the loss factor as a function of the temperature [Met
83]...................................................................................................................47
Figure 2.14 Illustration of various ignition methods and propagation of the combustion
wavefront [Dal90]............................................................................................ 54
IX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2.IS Temperature versus time relationship when synthesizing combustible materials
with (a) conventional heating and (b) microwave energy [Adapted from Dal90]
(Ts is sustained temperature).............................................................................57
Figure 3.1. Simplified experimental setup for samples run in Radarline multimode
microwave oven............................................................................................... 76
Figure 4.1. Effect of composition on temperature profile of the AI2 O3 - TiC powders
86
Figure 4.2. Effect of heating method and microwave power on temperature profile of the
stoichiometric and 40% excess both A1 and AI2O3 powders.............................. 87
Figure 4.3. X-ray diffraction of the AI2 O3 - TiC powders ignited by 800watts MHH
88
Figure 4.4. Scanning electron micrograph of the stiochiometric 2Al2C>3-3TiC powder ignited
by various heating method................................................................................91
Figure 4.5. Scanning electron micrograph of the 2 AI2O3 - 3TiC - 40% excess AI2O3 powder
ignited by various heating method.....................................................................92
Figure 4.6. Scanning electron micrograph of the 2 AI2O3 - 3TiC - 40% excess A1 powder
ignited by various heating method.....................................................................93
Figure 4.7. Scanning electron micrograph of the 2 AI2O3 - 3TiC - 40% excess both AI2O3 and
A1 powder ignited by various heating method....................................................94
Figure 4.8 Temperature profile of 2 AI2 O3 - 3TiC - 20% excess A1 -20% excess AI2 O3
powders ignited by microwave heating (MH) and microwave hybrid heating
(MHH).............................................................................................................107
Figure 4.9 X-ray diffraction patterns for the MH ignited AhC^-TiC powders: (a) 10% excess
Al, 10% excess AI2O3 , (b) 10% excess Al, 30% excess AI2 O3 , (c) 30% excess Al,
10% excess AI2O3 , and (d) 30% excess Al, 30% excess AI2O3 ..........................108
Figure 4.10 X-ray diffraction patterns of the AhCb-TiC powders ignited by: (a) microwave
heating (MH), (b) microwave hybrid heating (MHH).........................................109
Figure 4.11 Scanning electron micrograph of the MH and MHH ignited 2 AI2O3 -3TiC - 20%
excess Al - 20% excess AI2O3 powders with varying particle size of carbon: (a,
b) 0.017 pm, (c, d) 0.037 pm, and (e, f) 0.075 pm. (bar = 100 pm).................... 112
Figure 4.12 Scanning electron micrograph of the MH and MHH ignited 2 AI2 O3 -3TiC 20%excess Al - 20% excess AI2 O3 powders with varying particle size of carbon:
(a, b) 0.017 pm, (c, d) 0.037 pm, and (e, f) 0.075 pm. (bar = 10 pm)................. 113
Figure 4.13 Scanning electron micrograph of the MH and MHH ignited 2 AI2 O3 - 3TiC 20%excess Al - 20% excess AI2 O3 powders with varying particle size of alumina:
(a, b) 0.49 pm, (c, d) 20-38 pm, and (e, f) 63-74 pm. (bar = 100 pm).................114
x
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 4.14 Scanning electron micrograph of the MH and MHH ignited 2 AI1 O3 -3TiC 20%excess Al - 20% excess AI2O3 powders with varying particle size of alumina:
(a, b) 0.49 pm, (c, d) 20-38 pm, and (e, f) 63-74 pm. (bar = 10 pm)................. 115
Figure 4.15 Scanning electron micrograph of the MH and MHH ignited powders of 2 AI2O3 3TiC -20% excess AI2O3 with varying amount of excess Al: (a, b) 0% (c, d)
20%, and (e, f) 40% (bar = 100 pm)..................................................................116
Figure 4.16 Scanning electron micrograph of the MH and MHH ignited powders of 2 AI2O3 3TiC -20% excess AI2 O3 with varying amount of excess Al: (a, b) 0% (c, d)
20%, and (e, f) 40% (bar = 10 pm)...................................................................117
Figure 4.17 Scanning electron micrograph of the MH and MHH ignited powders of 2 AI2O3 3TiC -20% excess Al with varying amount of excess AI2O3 : (a, b) 0% (c, d)
20%, and (e, 0 40% (bar = 100 pm)................................................................. 118
Figure 4.18 Scanning electron micrograph of the MH and MHH ignited powders of 2 AI2O3 3TiC -20% excess Al with varying amount of excess AI2O3 : (a, b) 0% (c, d)
20%, and (e, f) 40% (bar = 10 pm)....................................................................119
Figure 4.19 Result of EDS analysis of 2 AI2O3 -3TiC -20% excess Al - 40%Al2C>3..............121
Figure 4.20 The BSE image with EDS spot analysis of 2 AI2O3 -3TiC -20% excess Al 40%AI2O3 sample............................................................................................ 122
Figure 4.21 Prediction Profiler of ignition time of Ah0 3 -TiC powders.................................137
Figure 4.22 Prediction Profiler of ignition temperature of A^Os-TiC powders..................... 138
Figure 4.23 Prediction Profiler of combustion time of Al2 0 3 -TiC powders..........................139
Figure 4.24 Prediction Profiler of combustion temperature of AbCb-TiC powders...............140
Figure 4.25 Temperature profile of 2 AI2 O3 -3TiC -20% excess Al -20% excess AI2 O3 with
varying carbon particle size ignited by (a) MH and (b) MHH.............................142
Figure 4.26 Temperature profile of 2 AI2 O3 -3TiC -20% excess Al -20% excess AI2 O3 with
varying alumina particle size ignited by (a) MH and (b) MHH...........................143
Figure 4.27 Temperature profile of 2 AI2 O3 -3TiC -20% excess AI2O3 with varying excess
amount of Al ignited by (a) MH and (b) MHH...................................................146
Figure 4.28 Temperature profile of 2 AI2 O3 -3TiC -20% excess Al with varying excess
amount of AI2O3 ignited by (a) MH and (b) MHH............................................. 147
Figure 4.29 Prediction Profiler of powder density of AhCVTiC powders............................148
Figure 4.30 Prediction Profiler of agglomeration size of AUOs-TiC powders......................150
Figure 4.31 Prediction Profiler of particle size of AhQj-TiC powders................................. 151
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.32 Prediction Profiler of specific surface area of A^Ch-TiC powders..................... 152
Figure 4.33 The effect of the amount of excess AI2 O3 and its particle size on the ignition time
of MH ignited AUOs-TiC powders.................................................................... 153
Figure 4.34 The effect of the amount of excess A1 2 0 3 and carbon particle size on the ignition
temperature of MHH ignited A^Ch-TiC powders.............................................. 154
Figure 4.35 The effect of the amount of excess AI2 O3 and carbon particle size on the
combustion temperature of MHH ignited AhCb-TiC powders............................156
Figure 4.36 The effect of the amount of excess AI2 O3 and its particle size on the combustion
temperature of MHH ignited AhOs-TiC powders.............................................. 157
Figure 4.37 The effect of the amount of excess AI2 O3 and Al on the combustion temperature
of MHH ignited AfeCh-TiC powders................................................................. 158
Figure 4.38 The effect of the amount of excess AI2 O3 and Al on the agglomeration size of
MHH ignited Al2 0 3 -TiC powders..................................................................... 159
Figure 4.39 The effect of the amount of excess AI2 O3 and its particle size on the
agglomeration size of MHH ignited AfeCb-TiC powders....................................160
Figure 4.40 The effect of amount of excess A I 2 O 3 and Al on the final density of MHH ignited
AhCVTiC powders.......................................................................................... 162
Figure 4.41 Temperature profile of the reactants heated by MH.......................................... 167
Figure 4.42 Temperature profile of the reactants heated by MHH....................................... 173
Figure 4.43 Temperature profile of all combination mixtures for the stoichiometric
composition heated by MH............................................................................... 174
Figure 4.44 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by MH............................................................................... 175
Figure 4.45 AG vs. temperature for some possible reactions between three reactants...........177
Figure 4.46 Temperature profile of all combinations mixtures of the stoichiometric
composition heated by MHH.............................................................................179
Figure 4.47 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by MHH.............................................................................180
Figure 4.48 Temperature profile of all combinations mixtures of the composition with excess
Al heated by MH.............................................................................................. 181
Figure 4.49 Temperature profile of all combinations mixtures of the composition with excess
Al heated by MHH............................................................................................182
xii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.50 Temperature profile of all combinations mixtures of the composition with excess
AI2 O3 heated by MH........................................................................................ 184
Figure 4.51 Temperature profile of some combinations mixtures of the composition with
excess A I 2 O 3 heated by MH............................................................................. 185
Figure 4.52 Temperature profile of all combinations mixtures of the composition with excess
A I 2 O 3 heated by MHH......................................................................................187
Figure 4.53 Temperature profile of some combinations mixtures of the composition with
excess A I 2 O 3 heated by MHH..........................................................................188
Figure 4.54 Temperature profile of all combinations mixtures of the composition with excess
Al and AI2 O3 heated by MH............................................................................ 190
Figure 4.55 Temperature profile of some combinations mixtures of the composition with
excess Al and A I 2 O 3 heated by MH................................................................. 191
Figure 4.56 Temperature profile of all combinations mixtures of the composition with excess
Al and AI2 O3 heated by MHH......................................................................... 192
Figure 4.57 Temperature profile of some combinations mixtures of the composition with
excess Al and AI2O3 heated by MHH.............................................................. 193
Figure 4.58 The DTA results for all mixtures of the stoichiometric composition.................. 196
Figure 4.59 The DTA results for all mixtures of the composition with excess Al..................197
Figure 4.60 The DTA results for all mixtures of the composition with excess A I 2 O 3 ............ 198
Figure 4.61 The DTA results for all mixtures of the composition with excess both Al and
AI2 O3 ............................................................................................................. 199
Figure 4.62 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by DTA........................................................................... 200
Figure 4.63 The reaction path within a combustible mixture of Ti0 2 , C, and Al using a)
conventional heating, b) microwave heating.................................................... 205
xiii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
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
MICROWAVE-INDUCED COMBUSTION SYNTHESIS OF Al20 3-TiC POWDER
By
Duangduen Atong
December 2000
Chairman: Dr. David E. Clark
Major Department: Materials Science and Engineering
Microwave heating offers many potential advantages, including synthesizing of
material. This advantage results from the unique features of microwave heating which
include its internal volumetric heating and subsequent inverted temperature profile.
Fundamental knowledge of how reaction parameters influence microwave-ignition behavior
is needed in order to provide information required for synthesizing the desired products.
This study was performed to synthesize Al2 0 3-TiC powders by using microwave
heating. The important reaction parameters were identified using the Plackett-Burman
design including amount of excess Al and
A I2 O 3 ,
particle size of C and A I 2 O 3 , and heating
method. The effect of these reaction parameters on the ignition behavior and characteristics
of the resulting powders was then evaluated by the central composite design.
The combustion synthesis of Al20 3 -TiC powders using microwave heating (MH) and
microwave hybrid heating (MHH) was successfully achieved. The MHH-ignited sample
required a longer time to reach ignition temperature and thus resulted in a lower combustion
temperature. The results showed no significant difference in characteristics of powders
ignited by these two heating methods. The addition of diluents to reactants increased
xiv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ignition time and temperature, while decreasing the combustion temperature. An increased
particle size of AI2 O3 lowered the ignition time and temperature. The density of product
decreased with increasing amounts of Al. Addition of excess
A I2 O 3
to reactants resulted in
decreasing the agglomeration size. The empirical models relating these important parameters
and their interactions to the responses were then developed.
The mechanism governing the combustion reaction of AUCh-TiC powders under
microwave and conventional heating was also investigated. The results suggested that
reaction mechanisms using these two methods were similar. The reaction proceeded in a
three-stage process where the aluminum melted, then die melting Al reacted with the dtania,
and finally the titanium reacted with the carbon to produce the Al2 0 3 -TiC. The only
difference was the way the reactants were heated. In conventional heating, material was
heated by heat transfer process depending on its thermal conductivity. Heating by
microwaves is a function of dielectric properties of the material. Depending on the
microwave absorption of the materials in question, heat generation by microwave heating
occurs internally.
xv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1
INTRODUCTION
Ceramic composites have been used in the cutting tool industry for years [Bor88,
Kin65], The composites’ machining abilities are superior to those of carbides, cast alloys,
and high-speed steels because of the high-temperature deformation resistance of ceramics
[Lee83]. Therefore, ceramic cutting tools can be used at the significantly higher
temperatures generated by faster cutting speeds and deeper depths of cut. A major
improvement in ceramic cutting tools was an AfeCVbased composite containing 1 to S
microns of titanium carbide particles uniformly dispersed throughout the matrix. The
alumina-titanium carbide composites are used in the tooling industry for the machining of
super alloys and turning of cast irons and very hard steels. Another modem application of
this composite is the use as a magnetic head slider substrate, which has stringent
requirements for oxide-carbide homogeneity and microscopic wear resistance. Most
commercial materials are nominally AI2O3 - 3 0 wt% TiC (25 vol%TiC). They exhibit good
electrical conductivity as well as high hardness (20 to 22GPa), good strength (500 to
600MPa), and moderate fracture toughness (4 to 4.5 MPa mI/2).
Currently, commercial AhC^-TiC composites are manufactured primarily by
pressureless sintering or hot pressing of TiC and AI2 O3 powders. They are hot pressed at
temperatures above 1600°C and are machined into their final shapes [Bur98, Wah80]. This
additional finishing of the cutting edges is the major production cost of these tools [Bal88].
The pressureless sintering of AfcC^-TiC composites is produced at temperatures in excess of
1800°C with rapid heating rates and short holding tones. The sintered samples are
subsequently hot isostatic pressed (HIP) for final densification. Although this procedure
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
lends itself to near-net shape fotming, the processing costs are still high because of the need
for HIPing. In addition, the very high heating rate needed to fully density the composites
also generates high thermal stresses that result in fracture of large parts [Bor86, Cut88,
Ish89, Lee88],
Recently, the synthesis of AhCh-TiC composites by the aluminothermic reduction of
TiC>2 in the presence of carbon has drawn much attention as an alternative to conventional
processes [Ada90, Bow91, Bow94, Bow96, Cut85, Cut92, Fen92, Fen95, Kun96a, Kun96b,
Kun96c, Moor94, Pet92, Rab90], The process utilizes the heat released by an exothermic
reaction between reactants to prepare materials. Once ignited, the exothermic energy
generated from a reaction is large enough to bring subsequent portions of the reactants up to
the ignition temperature without adding external energy. In this way, the reaction can selfpropagate through the reactant mixture until it is converted to the product. The use of such a
process to fabricate materials is commonly referred to as self-propagating high temperature
synthesis (SHS) [Mer72] or more simply, combustion synthesis [Holt82]. A number of
possible advantages over conventional materials preparation processes include (a) fast
processing time and simplicity of the process [Mer83, Mer90], (b) self-purification of the
reactant and higher purity of products [Man84], (c) generation of temperatures in excess of
2000°C without the need of expensive furnaces, (d) using a low-cost oxide (e.g. TiC>2 ) as a
reactant instead of high-cost materials (e.g. TiC or metallic TO, (e) the possibility of
obtaining complex or metastable phases [Man84], and (f) the possibility of simultaneous
synthesis and densification of materials [Holt86, Yam85].
While there has been interest in combustion synthesis of materials, the use of
microwave energy to process a wide variety of ceramics and composites also has attracted a
lot of attention. The potential benefits of microwave processing over conventional
processing methods include (a) significant reductions in manufacturing costs due to energy
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
3
saving and shorter processing times, (b) improved or unique microstructures and properties,
(c) improved product uniformity and yields, and (d) synthesis of new materials [Sut89].
Microwave heating is fundamentally different from conventional processes. In
microwave processing, thermal energy induced by or dissipated by the microwaves within
the bulk of materials is the source of heat. This internal and volumetric heating raises the
temperature of the materials such that the interior portions become hotter than the surface,
because the surface loses heat to the cooler surroundings.
This is the reverse of
conventional heating, where heat from an external source is supplied to the exterior surface
and diffuses toward the cooler interior regions. As a result, this internal volumetric heating
mechanism provides several advantages such as heating small and large samples uniformly
and rapidly, reducing the time and temperature processing requirements and improving the
properties of the processed materials [Coz91, De91a, De91b, Jan91]. In addition, the
inverted thermal gradients in microwave heating makes it possible to reduce thermal stresses
that cause cracking during processing; to efficiently remove volatile constituents (gases or
binders) from the interior of thick materials; or conversely, to penetrate required gases into
the hotter interior of material; and to synthesize composites and ceramic materials.
The nature of the microwave power deposition into the materials and subsequent
inverted temperature profile appear to offer advantages to combustion synthesis. More
recently, the union of two technologies: microwave processing and combustion synthesis
has been developed. They have been found to reinforce each other in the fabrication of
composites by allowing for enhanced process control and superior product quality, as
compared to conventional combustion synthesis [Cla91, Dal90]. A study of the synthesis of
AhCh-TiC using microwave energy and conventional heating showed that the conventional
ignited samples required longer processing times as compared to the samples ignited by
microwave heating [Ato98]. The time required to ignite a sample using an isothermal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
furnace was observed to be dependent on the temperature in the furnace. In this process,
ignition times were reduced exponentially from 2 hours at 600°C to 4 minutes at 1000°C
though still longer than I to 3 minutes required for samples ignited by microwave energy.
In addition, microwave-ignited samples had a higher density and more uniform
microstructure. The complete conversion of reactants was achieved in the microwave ignited
sample, while the reactants were still observed in the sample ignited at 1000°C in the
isothermal furnace. This complete conversion of the reactants occurred as a result of the
inverted temperature gradient in microwave heating and may lead to producing a completely
different product morphology [Ahm9l, Yii95, Wil93], Microwave energy also has been
used to synthesize the solid solution of urania and thoria which is used as fuel in the nuclear
industry [AntOO, Cha98], The microwave-processed powders yielded compacts that can be
sintered to relatively high densities compared to the conventionally hot plate-derived
powders.
Even though early data are promising, so far there has not been a systematic
investigation conducted to determine how microwave energy ignites and sustains a
combustion reaction. Those investigations consider such variables as sample mass, sample
density, amount of liquid metal, or mass of susceptor. Interestingly enough, the effect of the
particle size of the reactants has never been considered in any of these studies. In addition,
most of SHS products often are in porous monolith forms, which require further pulverizing
to obtain powder. The use of microwave energy combined with combustion reactions to
synthesize ceramic powders has received little attention [AntOO, Cha98, Par98], These SHS
powders have various applications, which are greatly dependent on their grain sizes.
Submicron and micron-sized single-crystalline powders are used for sintering and hot
pressing. Coarse-grained agglomerate and composite powders are used for gasothermal
sputtering [Mer93a]. Powders of various sizes are used as abrasive and fillers. Therefore, it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
is the purpose of this study to synthesize and process A^C^-TiC powders by using a
combined technology of SHS combustion synthesis and microwave heating.
The objectives of this study are as follows:
•
To perform a comprehensive evaluation of the effects of reaction parameters on ignition
behavior of the microwave-induced combustion synthesis of A^Ch-TiC powder. These
reaction parameters include particle sizes of reactants, compositions of the reactants,
heating method, and microwave power.
•
To examine the effects of the above parameters on the microstructure and characteristics
of the resulting powders, including particle size, surface area, and powder density
•
To identify the important reaction parameters that influence microwave-ignition of the
A^C^-TiC powders
•
To develop an empirical model based on statistical analysis relating the important
reaction parameters to ignition behavior and characteristics of the resulting powders
•
To suggest the mechanisms in synthesizing AkOj-TiC powders using microwave energy
Based on the results of the study and knowledge of combustion synthesis and
microwave heating, this work will provide fundamental knowledge of how reaction
parameters influence microwave-ignition behavior. This should yield insight into the effect
of various reaction parameters on certain properties of the resulting products, which
ultimately may help determine a suitable combmation of reaction parameters for
synthesizing the desired products for a particular application. In addition, an empirical
model will help predict combustion behavior and the resultant product configuration. This
information should be helpful in the future development of combustion synthesis using
microwave energy.
This study is organized into several chapters. Chapter 2 provides a literature review
pertinent to this work. This includes some background on combustion synthesis, a general
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
review of microwave heating fundamentals and microwave/materials interactions, and a
review of microwave-induced combustion synthesis. Chapter 3 describes the materials,
experimental design and procedure, and the characterization techniques used to evaluate
final products. The results and discussion of this investigation are presented in Chapter 4.
Chapter 5 provides the conclusions drawn from this study.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2
BACKGROUND
This chapter has three major independent sections. In Section 2.1, a brief review of
the combustion synthesis is provided. Discussions include thermodynamic considerations,
theory and mechanisms, and some uniqueness and advantages of the combustion synthesis.
The influence of processing parameters on reaction progress and structure formation is
discussed in detail. An historical account of the development of combustion synthesis is
presented.
Section 2.2 of this chapter is an overview of the microwave heating fundamentals.
Detailed discussions include the theoretical aspects of interactions between microwaves and
materials, unique characteristics of microwave heating, its advantages, and the history of
microwave processing.
Section 2.3 presents the union of two technologies: combustion synthesis and
microwave processing, in the field of ceramics. The fundamentals discussed in this section
are relevant to the goals of this study and will be referred to throughout this dissertation.
2.1 Combustion Synthesis
2.1.1 History of Combustion Synthesis
Combustion synthesis of inorganic materials is a technology that has been in
existence for more than a century. The elements of combustion synthesis were understood
early in the 19th century by Berzelius and refined at the end of that century by Moissan
[Hla91]. In light of this, however, Goldschmidt, a German metallurgist, should receive credit
for discovering the self-propagating phenomena in 1885 [Hla91]. He fully described the
7
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
physical principles of self-propagating reactions and, based on his work, aluminothermic
metallurgy emerged. Goldschmidt’s first thermite reaction was
Fe2Oj + 2AI -> Al20 3 +2Fe -851.4&/I mol
(2.1)
Hematite was reduced by aluminum, forming an alumina slag and iron metal. This
reaction was found to expel large quantities of heat. This generated heat propagates a
reaction front, forming the product at the other end.
In 1959, Walton and Poulos [Wal59, Wal89], American scientists, used the
Goldschmidt self-propagating principles of aluminothermic reactions to synthesize a variety
of cermets. An example of a thermite reaction to form a metal composite as proposed by
Walton is
Cr20 3 + 2AI
2Cr + Al20 3 - 536.0kJ t mol
(2.2)
While research in the United States slowed significantly soon afterward, the Soviet
Union made substantial progress in combustion synthesis during the mid sixties through the
late seventies. In 1967, A.G. Merzhanov et al. [Mer71] discovered the phenomenon of selfpropagating high-temperature synthesis (SHS) during the study of the combustion of a
titanium-boron mixture. The product produced in this first experiment was unexpectedly
hard and a gasless combustion reaction was achieved. This was the beginning of a rapidly
expanding research effort into the formation of refractory materials by exothermic reaction.
When the product is a solid at the combustion temperatures, then die term “solid flame” is
applied to the reaction. The SHS products cover a wide variety of materials, ranging from
the abrasive powders of TiC to the superconducting oxides of YBa2Cu3<>7. In 1987, the
Institute of Structural Macrokinetics of the USSR Academy of Sciences was established in
Chernogolovka to serve as the Soviet SHS research center with Merzhanov as a director.
Since then, more than 300 compounds have been synthesized by the SHS process. More
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
than 30 different institutes and universities in Russia and the former USSR have been
continually engaged in SHS research and development. This results in a large number of
publications present in the literature, with recent review publications being available
elsewhere [Fra85, Mer89, Moor95a, Moor95b, Mun88].
Since the early 1980s, SHS-related research has become multinational. The
publication of the early Russian work by J. Crider [Cri82] stimulated the interest of
researchers in several countries. In the United States, the combustion synthesis research
began primarily with the Defense Advanced Research Projects Agency (DARPA)
program in 1984 (Gab85). In Japan, the study of combustion synthesis was primarily
initiated by Koizumi and Miyamoto in 1984. Since then a substantial and continuously
growing research effort has been invested in this field [Hol87, Mer93b, Mun89, Pam97,
Pat93, Ric91, Sub92, Var92, Wan93, Yi90]. This is accented by the genesis of a journal
entitled “International Journal of Self-Propagating High-Temperature Synthesis,” which
has been published since 1992 with Merzhanov as the general editor. Five international
symposia on SHS were held in 1991 (Alma-Ata, Kazakhstan), 1993 (Honolulu, USA),
1995 (Wuchan, China), 1997 (Spain), and 1999 (Moscow, Russia). Details of the
historical development of combustion technology are summarized in Appendix A.
2.1.2 Introduction to SHS
2.1.2a Terminology
Self-propagating high temperature synthesis is the term being used to describe a
process in which initial reagents (powder or pellet form), when ignited in the presence of
either air or an inert atmosphere, spontaneously transform into products because of the
exothermic heat of reaction. Other terminologies used for describing the SHS process are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
gasless combustion, combustion synthesis, self-propagating combustion, or self-propagating
exothermic reactions. Various heat sources used to ignite samples are, for instance, electric
arc, electrically heated Mo filament, a magnesium ribbon, laser pulse, another highly
exothermic reaction, rapid heating in a furnace, or microwave energy.
Combustion reactions can be conducted either by a self-propagating mode or by a
thermal explosion mode. In the self-propagating mode, the reaction is initiated at one end of
the sample and self-propagates in the form of a combustion wave with a velocity of about
0.1 to 15 cm/s through the reactants, leaving the reacted product behind [Holt87],
Figure 2.1 shows a schematic of a SHS reaction. A heated coil raises the temperature
of the reactants, A+B, at one face of the sample to a thermodynamically favorable state to
form the product AB. The formation of AB liberates heat, which allows the reaction front to
propagate to the other end of the sample. The reaction front ideally converts all the reactants
in to the product [Cri82]. In die thermal explosion mode, the sample is heated uniformly and
the reaction occurs simultaneously through the whole sample. This mode is particularly
appropriate for weakly exothermic systems, which require preheating before ignition
[Var92]. The maximum temperature of the thermal explosion is higher than the
corresponding self-propagating regime because the entire sample is heated to the ignition
temperature [Holt91a, Holt91b].
Heating dement
Q
Reaction Zone
Propagation
AB
*
Heat
Product
Reaction Zone
A+B
Reactants
Figure 2.1 A schematic of a SHS reaction
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
2.1.2b Advantages
The most obvious advantages of the SHS process are as follows:
•
The generation of a high reaction temperature (800 to 3500°C) can vaporize volatile
contaminants and, therefore, produce high purity products.
•
The self-sustaining exothermic reaction produces thermal energy, which provides
enough heat to sustain the reaction of the unreacted reactants; therefore an expensive
high-temperature furnace is not required and energy consumption can be reduced.
• The relatively short reaction time - in the order of seconds instead of hours or days in
comparison with the normal sintering, because of the highly exothermic nature of the
reaction - results in low operating and processing costs.
• The high thermal gradients (up to 107deg/cm) and rapid cooling rates can give rise to
new non-equilibrium or metastable phases, which may be more sinterable.
2.1.2c Tvoes of reactions
A variety of compounds, intermetallics, and composites are produced by SHS
reactions. The simplest combustion synthesis reaction involves an interaction between
elements only, for example, Ti+C to form TiC or Mo+2Si to produce MoSi2 . The next level
of complexity is the reaction among elements) and compound(s), which is represented by
the well-known “thermite” reaction. There are two types of thermite reactions. The first one
involves the reduction of an oxide to the element. For example:
3Fe30< + 8/4/ -» AA120 3 + 9Fe - 3347.7kJ / mol
(2.3)
The second type, which is the type of interest in this study, involves the reduction of
an oxide to the element, which subsequently reacts with another element to form a required
compound. This type has drawn interest because it uses cheaper oxide reactants compared
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
with expensive elemental reactants to produce ceramic composites. An example for this type
of reaction is
37702 + 3C + 4AI -»• 377C + 2 A I Q -1076.6kJ / mol
(2.4)
The SHS reactions also can occur between compounds. For example, BaO + TiOi
react to form BaTiCb, an important electronic material. Many other complex oxides, e.g.
PbTiCb, MnSiCb, and BaSiCh, can be synthesized in a similar fashion.
The combustion of materials occurs at several interfaces; solid-solid, solid-liquid,
liquid-liquid, solid-gas, and liquid-gas. The type of interface present during the combustion
is determined mainly by the adiabatic temperature, Tad- This is the temperature to which the
material is raised as a consequence of the evolution of heat due to the chemical reaction,
with an assumption of no heat loss.
The formation of carbides and borides usually occurs predominately at a solid-solid
or solid-liquid interface. Complete solid-state combustion occurs when Tad is less than the
melting temperature of all of the reactants leading to the lowest combustion velocity. If the
T^ lies between the melting point of the reactants, then the interface is a solid-liquid
interface. The molten component spreads at a high rate in the compact, resulting in the
highest velocity of combustion. The latter is the case for the formation of the Ah0 3 -TiC
composite. When Tad is greater than the melting points, but lower than the boiling points, of
the reactants, reactions occur in die liquid phase. Many intermetallics, e.g. NiAl, and a few
silicides form by this reaction. Lasdy, the solid-gas interlace occurs when one of the
components is in the gas phase. This is most common in the formation of nitrides, hydrides,
sulphides, selenides, and phosphides.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
2.1.2d Combustion Products
Materials produced using the SHS method have been used in the following
applications [Esl89]:
•
Abrasive, cutting tools, polishing powders, e.g. TiC, cemented carbides, and
carbonitrides
•
Resistive heating elements, e.g. MoSi2
•
High temperature intermetallic compounds, e.g. nickel aluminides
•
Shape memory alloys (SMA), e.g. TiNi
•
Steel processing additives, e.g. nitrided ferroalloys
•
Electrodes for electrolysis of corrosive media, e.g. TiN, TiBi
•
Coatings for contaminant of liquid metals and corrosive media, e.g. products of
aluminum and iron oxide thermite reactions
•
Powders for further ceramic processing, e.g.
•
Thin film and coatings, e.g. MoSi2 , TiB2
•
Functionally-graded materials (FGM), e.g. TiC+Ni
•
Composite materials, e.g. TiC+AfeCh, TiC+AfeOj+Al
•
Materials with specific magnetic, electric, or physical properties, e.g. BaTiC>3 ,
S13N 4
YBa2Cu307.x
These SHS produced materials can be in powder or porous monolith form. The
powder form can be used “as-is” (e.g. abrasive materials) or can be processed subsequently
to form products of different shapes and densities. Porous monolithic products, e.g. 50% of
theoretical density, have been used as filters or preforms for liquid metal infiltration in the
production of ceramic-metal composites. The source of high porosity can be rationalized to
two groups: intrinsic and extrinsic. The high exothermicity associated with SHS reactions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
implies tighter atomic bonding in the products, compared with reactants, and thus gives rise
to higher density products. The density difference leads to a volume change and results in
intrinsic porosity. Extrinsic porosity is generated as a result of outgassing or evaporation of
gases due to high reaction temperatures and the porosity already present in the green
reactant sample [Moor95b, Ric85a]. To obtain highly dense and perhaps near-net-shaped
products, three approaches are used. The first involves the application of pressure either
during the combustion or immediately after it is completed. The second uses an excess of
liquid metal to infiltrate the pores in the ceramic matrix. The last takes advantage of the
molten state of the products to form cast bodies. These densification techniques are
described briefly as follows:
Hot pressing. In most cases the samples were placed in graphite dies under pressure
and ignited. The timing of the application of the load is critical in obtaining a crack-free,
high-density product. If the load is applied too early, gases cannot escape from the sample
and this results in increased porosity. If the consolidating load is applied too late, the
products are no longer in their plastic range and the opportunity to complete consolidation is
lost [Dun91, Hol87].
Shock wave consolidation. A hot porous body produced by SHS is densified while
hot by the action of a pressure wave generated by the detonation of a high explosive. This
process is termed SHS/DC, a combining of SHS reactions with dynamic compaction
(explosive). The TiC and TB 2 were fabricated using this method [Kec90a, Kec90b].
However a major problem was cracking caused by such high impacts on the products. An
alternative to explosive compaction was to use pressure generated by a high-speed projectile
gas gun to density the product [Holt87],
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
Hot isostatic pressing (HtPingl. Holt sealed the Ti+2B mixture in a tantalum can and
ignited the reaction under an argon pressure of 101 MPa. The product was densified to 93%
of theoretical density [Holt87],
High pressure self combustion sintering (HPCSi. Yamada, Miyamoto, and Koizumi
developed a process of consolidation under very high pressure. They inserted a reactant
pellet into a pyrophylite cube cell and subjected this cell to a pressure of 3 GPa. Dense
materials of TiC, Si, and AhOj-TiC have been fabricated using this approach [Ada85,
Koi90, Miy84, Yam86, Yam87],
Gas-pressure combustion sintering. This was an alternative process to HPCS and was
developed by Miyamoto [Miy90a, Miy90b], The press- formed reactants are vacuum-sealed
in borosilicate glass containers. These containers then are placed in an ignition agent in a
graphite-crucible, which is placed in a high-pressure chamber. This process is initiated by
heating the pressure chamber to make the glass container soft and 101 MPa argon pressure is
applied. The ignition agent then is ignited which in turn ignites the reactant pellets
converting them into dense products within a few minutes. The TiB2 , TiC, AUOs-TiC, T®2Ni and TiC-Ni have been compacted to near theoretical densities using this technique.
Hot rolling. The powder mixture is loaded in a metal tube with a covering of graphite
paper and a thin alumina-based felt material. The metal tubes containing reactant powder
were cold rolled to 60 to 70% theoretical density. Though high densities could be obtained,
such high density green compacts, either cannot be ignited, or if ignited, are soon
extinguished because of their high thermal conductivity [Ric86, Ric90],
In-situ liquid metal infiltration. Feng et al. [Fen92, Moor95b] deliberately generated
an excess of liquid metal in-situ with the combustion synthesis reaction and used this excess
liquid metal simultaneously infiltrate the pores in the ceramic matrix, which also is produced
within the overall SHS reaction system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 .1.3 Thermodynamics of Combustion Synthesis
Once ignited, extremely high temperatures can be achieved in very short times (ca. 2
to 3 seconds) because of the highly exothermic nature of the reaction. It is reasonable, at this
point, to assume that a thermally isolated exothermic system exists because there is very
little time for the heat to dissipate to the surroundings. This results in the maximum
temperature to which the product is raised supposedly being the adiabatic temperature, Tad.
Because of the assumption of adiabatic conditions, for energy to be conserved, the
enthalpy of the reactants at the initial temperature (T0) must be equal to the enthalpy of the
end products at the final temperature ( T a d ) (Figure 2 . 2 ) . This leads to
'£ln<T«),-n(TM-Q
>•1
(2-5)
where H ( T a d ) is the enthalpy of product i at T « d , H(T0) is the enthalpy of product i at T0 , Q
is the heat released during the reaction, and d is the total number of product phases [Holt86].
«
€
s
u
Temperature
Figure 2.2 Schematic representation of the adiabatic temperature calculation [Holt86].
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
17
If Reaction (2.S) takes place under standard conditions of temperature (T0=298K)
and pressure (P=latm), then Q is the negative of the standard heat of formation of product, AH°Q98 - However most exothermic reactions are initiated at a temperature above room
temperature (Tig). As a result, die value of Q should be modified by the following equation:
Q=
jA C pdT
(2.6)
29S
where ACp is the difference between the specific heats of the products and reactants at
constant pressure. For most many materials, however, the Neumann-Kopp rule is obeyed
(i.e. ACp=0), so that Q= -AH°p9s, and equation (2.5) then becomes
r- AH} 29* = J ACp(product)dT
(2.7)
298
where Tad < Tmp (product). If T«a = Tmp (product), then
- A
= ^ACp(product)dT+vAHm
(2.8)
298
where v is the theoretical fraction of molten product, and AHm is die heat of fusion of the
product. Finally, if Tad>Tmp then
- AH °f 29i
Tw
=J
r <*
AC p (product, solid)dT + AH m + J AC p (product, liquid)dT (2.9)
298
By using the above equations and the required thermodynamic data obtained from
the literature [Kub79], it is possible to calculate the Tad for the reactions. In practice,
computer programs (D.Halverson and CEdward, personal communication) become almost a
necessity for the calculation of Tad when considering complex reactions between multiple
compounds that react to form many product phases [Holt91a].
R e p ro d u c e d with permission o f the copyright owner. Further reproduction prohibited without permission.
In case of the A^Ch-TiC composites, the heat generated by an exothermic reaction
can be given as
-A H } ^ = J CpOTiC + 2A I& )dT + A//„ (2 A I& )
29*
( 2 . 10)
2323
where Cp(3 TiC+2 Al2 0 3 ) is the specific heat of product. The enthalpy of fusion of alumina,
AHm(2 Ah0 3 ), is included in the equation as Tad is greater than the melting point of A I 2 O 3
(2323 K). According to thermodynamic data, the calculated value of Tad is 2390K.
In general, however, the experimentally observed combustion temperature, Tc, is
frequently lower than Tad owing to heat losses. Nevertheless, Tad provides a useful estimate
of the reaction temperature. It also provides an indication of whether or not the synthesis of
a given material can be accomplished by a self-propagating method.
It has been empirically suggested that combustion reactions will not become selfsustaining unless Tad £1800K [Mer75, Nov75], Munir [Mun88] observed experimentally that
materials with a ratio of H ^s/A C p^s being less than 2xl03 are not considered to produce
self-sustaining combustion without the addition of energy from external heat source. This
rule is limited to reactions that do not form a liquid phase during combustion
(Tad<Tmp).
Figure 2.3 shows the relationship between these two parameters, H°Q98/ACPi298
v s . T*d.
These guidelines are not strict rules however. Combustion temperatures as low as 800°C in
the formation of hydrides were found to be self-sustainable [Dol78]. In any case, it is still of
interest to experimentally determine whether a particular reaction is sufficiently exothermic
to sustain continuous combustion.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
7000
'2
'2
▼
0
1000
2000
o
Figure 2.3 Relationship of H°q98/ACp,298 and Tad [Adapted from Mun88].
2.1.4 Combustion Wave Theory
2.1.4a Combustion wave structure
The solid-solid reactant combustion systems can be classified into two types based
on the nature of reaction occurring in the combustion zone. In the first type, complete
conversion of the reaction takes place in a narrow combustion zone. This condition is
usually attained when the reactants are thoroughly dispersed as a homogeneous mixture. In
the second type, the heterogeneous system, the reaction in the combustion zone does not
achieve frill conversion, and the reaction continues to occur after the passage of the
combustion wave. An example of this system is one having the reactants diffusing through
the solid reaction product to reach the reaction site. This solid reaction product represents a
highly activated barrier between the reactants. The continuation of the reaction requires the
diffusion of the reactant(s) through the solid product, a process which is typically associated
with a high activation energy and thus occurs at a slow rate.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
20
2.1.4a.l Homogeneous system
In a homogeneous system [Mer74], complete conversion of the reactants occurs in a
narrow combustion zone. The reactants are intimately and thoroughly dispersed as very fine
particles. In this case, mass diffusion is negligible. Mass diffusivity, D, is much less than
thermal diffusivity, a, and thus the mass transfer equation becomes unimportant and is hence
neglected. Combustion occurs under adiabatic conditions in which there is no heat loss to
the ambient environment. Combustion occurs in a steady state and the combustion wave
propagates at a constant rate.
Figure 2.4 illustrates a combustion wave structure for homogeneous combustion. The
combustion wave is moving from right to left The region ahead of the wave is the heat-up
zone in which the reactants are heated from an initial temperature T0 to
Tp,
the temperature
at which product formation begins. In this zone no heat is evolved (rate of heat evolution,
<p=0) and no product is formed (degree of conversion, q=0). The region in which Tp
increases to the maximum combustion temperature, Tc, is the reaction zone. In this zone <{>
becomes greater than zero and full conversion of reactants occurs (q increases from 0 to 1)
in a narrow zone, 5w.
Wave
propagation
Figure 2.4 Combustion wave structure for homogeneous combustion (q is degree of
conversion, (ji is the rate of heat generation) [Mer74].
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
21
At any point in the sample a heat balance equation can be given by
p
ot
=
oX
r
r
(2.11)
Heat accumulation = heat diffusion + heat generation - convective heat loss - radiative heat
loss
where Cp is heatcapacity (Jg'*K'1), p is the density of the product (gem'3), T is the
temperature (K), t is time (s), k is the thermal conductivity (Wcm*lKTl), x is the axial
distance coordinate, Q is heat of reaction (J g'1), <j»is the reaction rate (K s'1), a is the heat
transfer coefficient (Wcm'2K‘l), r is the sample radius (cm), e is the emissivity coefficient
(K), c0 is the Stefan-Boltzman constant (Wcm^K-4).
For a homogeneous reaction, the function <j>(T, q) can be express by
=$ =^
dt
dx
=
exp( - £ / R rx 1- n)-
(2 .1 2 )
where q is reacted fraction (K), U is wave velocity (cm s*1), K, is the pre-exponential
constant (s'1), n is the order of the reaction (no unit), R is the gas constant (J mole'1 K'1) and
E is the activation energy (J mole'1).
In general the heat losses are assumed to be negligible so that equation (2.11)
becomes
C ,/> ^ = * 0 + e M r , 7 )
(2.13)
c , p ^ = ± ^ - + Q p K aexp(-E /R m -nr
(2.14)
The average value of Cp over temperature range T0 to Tc and the value of the thermal
conductivity at Tc are generally used as effective values. Khaikin and Merzhanov [Kha66]
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
22
derived the following solution (combustion velocity) for the expression in equation (2.14) as
follows:
C RT1
U 2 = m a - Z — ^ - K Qe M -E /R T c)
Q E
(2.15)
where f(n) is a function of the reaction order, n, and a is the thermal diffusivity of the
product (a ~ k/pCp).
From experimentally determined values of U and Tc, a plot of ln(U/Tc) vs. 1/TCis
then constructed and the apparent activation energy for the process can be calculated. The
variation of Tc and subsequently U can be affected by adding an inert diluent (i.e. final
product) to lower Tc or by preheating the reactants to raise Tc. Calculated activation energies
are compared with literature values for various processes in the chemical system being
studied to propose a mechanism for the combustion reaction.
2.1.4a.2 Heterogeneous system
For a heterogeneous system, the reactants are not intimately in contact. In order for
the reaction to occur, the reactants must be transported to each other and diffusion in the
condensed phase must be considered. Diffusion processes can be very slow that the rate of
conversion of reactants into products can be much slower compared to the rate of heat
propagation. Therefore, the reaction in the combustion zone does not achieve full
conversion, and the reaction continues to occur after the passage of the combustion wave.
As a result, the sample may continue to glow well beyond the passage of the wave. A
modified combustion wave structure of this combustion process is shown in Figure 2.5. In
addition to two zones described earlier (heat up and reaction zone), this profile contains one
extra zone called the after-bum or bum-out zone. Within this zone, the degree of conversion
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
increases from nr to 1 and the rate of heat generation, <t>, remains greater than zero over an
extended distance [Zen80, Zen81],
Front
Figure 2.S Combustion wave structure for heterogeneous combustion (q is degree of
conversion, 4 is the rate of heat generation) [Mer74],
In heterogeneous systems, the heat source may not be distributed uniformly
throughout the substance but is localized in the region adjacent to the contact surface of the
reactants. Thus the combustion propagation may not be one-dimensional. A general
formulation of the problem of heterogeneous combustion systems, therefore, consists of
multi-dimensional nonlinear equations of thermal conduction and mass diffusion of
reactants. Solving these equations entail tremendous difficulties. One of the difficulties is
finding the form of the function <KT, q), rate of heat generation, for heterogeneous systems.
Zenin solved for reacted fraction, q, under heterogeneous conditions [Zen80,
Zen81],
A g a in ,
the general heat balance equation in which heat losses are neglected
(equation 2.13) is utilized. Applying the boundary conditions
x = -o c
T = T0
q =0
5T/5x = 0
x = + oc
T = TC
q=1
<3T/dx = 0
and integrating with respect to x, equation 2.13 becomes
k ‘f - C pp U (T -T 0) + QpUr1 = 0
ax
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(2.16)
24
The variation of the thermal conductivity, k, with r\ is k = n(ki - ki) + ki, where k2
and ki are thermal conductivity of products and reactants, respectively. Then substitution
this expression into equation (2.16) yields
Cpp U (T -T 0) ~ k {
1700 = ------------ ~ ----------------------(k2- k x) - j- + QpU
ax
(2.17)
By utilizing the formula: <^{T,rj)=dr}/dt, finally the rate of heat generation (<j>) was
determined [Zen81]. The rate of heat generation of heterogeneous combustion can be
expressed as
tj) = CQ rfp exp(-m rj)exp(-E / RT)
(2.18)
where the parameter m and p are defined depending on whether the reaction rate follows a
linear (m=p=0), parabolic (m=0, p=l), cubic (m=0, p=2), or exponential law (m>0, p=0).
The formal derivation of the expression found in equations (2.17) and (2.18) is beyond the
scope of this study. The original work can be found in [Zen 80, Zen81].
2.1.4b Stability of combustion wave
The combustion wave propagation can be either stable or unstable. In stable or
steady state combustion, the wave propagates at a uniform velocity through the reactants.
The wavefront velocity is mainly controlled by heat generation and heat loss. Any alteration,
which results in less production of the heat of reaction or increases the rate of heat loss, can
cause the wave propagation to become unstable. These alterations are ones such as the
addition of an inert component or diluent, the use of large particle size reactants, or
decreasing the preheating. Under unstable conditions, the wave displacement is not uniform
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
with time. This typically results in oscillatory, spin, or repeated combustion, and bum out.
These combustion modes are schematically presented in Figure 2.6.
Heating source
Product •<
Reactants
.AB_
AB
T
-
I
A+B
A+B
(a)
(b)
V
V
ht
h
»4
AB
*i
«*
h
I
-----^
A+B
(c)
A+B
(d)
Figure 2.6 Schematic representation of combustion wave propagation: (a) stable, (b)
oscillatory, (c) spin [Adapted from Mun88], and (d) burnout.
In the oscillatory mode, the wave propagates in successions of cyclic rapid and slow
jumps. The final product has a layered structure, which can be easily divided into discs
[Bor74]. The layered structure resulting from oscillation motion in the SHS reaction 3Ti02
+3C +4A1 -+3TiC +2 AI2O3 has been found by Moore et al. [Moor95a]. Holt et al., however,
have reported the spin combustion in the TiC>2-Al-C cylindrical sample [Holt91a]. In the
spin mode, the wave propagates in the form of a spiral encircling the sample at one or more
reaction spots. This combustion affects only a narrow surface layer of ca. 2mm, while the
bulk of the sample remains unreacted during the passage of the reaction wavefront [Fil75],
This was observed during the interaction between metal compacts and nitrogen gas.
However, recent observations have shown that spin combustion can take place throughout
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the whole sample [Mak79], Repeated combustion occurs when the ignited wavefront passes
a given point and a secondary combustion wave(s) follows at a lower rate of motion. The
secondary wavefront reacts the unreacted material leftover from the primary one. It should
be noted that the secondary wavefront can proceed in parallel to or against the direction of
the initial combustion front. Burnout is exactly what it implies. The combustion wavefront
extinguishes itself before all the material can be reacted. This burnout comes from
insufficient heat generation, which can lead to the transformation from a steady state to a
non-steady state mode and ultimately lead to the extinction of the combustion process.
Weakly exothermic reactions are prime candidates for extinction. These reactions are
difficult to initiate and sustain the reaction front.
To ignite such reactions, two methods can be used. The first method is referred to as
“thermal explosion” in which the reactant is heated in a furnace until the combustion takes
place spontaneously. This technique has been used to synthesize intermetallic compounds
such as Ni-Al [Phi85] or Cu-Al [Wan87], The second method is referred to as the “chemical
oven” in which the weakly exothermic reaction mixture is encapsulated in another highly
exothermic one. For example, in the formation of B4C, a mixture of B and C (Tad for B4C is
1000K) can be ignited by placing it inside a mixture of Ti and B (Tad = 3190K) and initiating
the latter reaction through the standard combustion synthesis method [Mun88],
Usually, for steady state combustion, both reactants and products are in the solid
state, i.e., the combustion temperature is less than the lowest eutectic temperature on the
phase diagram for the binary reactants. Complex multi-component solid-solid reactions and
many gas-solid combustion reactions appear in the unstable mode. The boundary between
the stable and unstable regimes is defined by the value of the parameter, a [Shk71],
DT
C T
a = — ^-(9.1 —£-2! - 2.5)
E
Q
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
(2.19)
where Cp is heat capacity of product (Jg‘K'1), Q is heat of reaction (Jg*1) , R is the gas
constant (J mole'1 K'1), T„d is the adiabatic temperature (K) and E is activation energy (J
mole1). If a > 1, combustion occurs in die stable mode and if a< l, the combustion is
unstable. The nature of unstable combustion is also determined by the value of a. The more
negative a becomes, the more pronounced are the oscillations.
2 \ .5 Important Reaction Parameters in Combustion Synthesis Reactions
There are a number of reaction parameters that affect SHS reaction; for example,
stoichiometry (including the use of diluents or inert reactants), green density, reactant
preform diameter, reactant particle size and distribution, and combustion conditions (e.g.
ignition power, initial temperature). These variables control both the combustion and
ignition temperature, percent conversion of the materials, wavefront velocity, and mode of
combustion. Therefore they have a significant effect on the microstructure and properties of
the final product. In order to control the process and to establish the optimum reaction
parameters for synthesizing a material, a fundamental understanding of the governing
reaction mechanism must be obtained. The effect of these parameters on SHS reactions are
summarized below.
2,1,5a Stoichiometric ratio
Non-stoichiometric mixtures are mostly used to control the process, e.g. making up
the loss of any species due to volatilization during combustion, and making the reaction less
violent. In general, any deviation from the stoichiometric ratio, e.g. addition of a diluent
(product itself), excess of a reactant or inert filler, will normally decrease the heat generation
rate and result in a reduction in the combustion temperature (Tc) and reaction wavefront
velocity. Holt and Munir [Holt86] found that the addition of 10% TiC to a mixture of Ti + C
caused a decrease in combustion temperature from 2720 to 2518K. The effect of dilution on
with permission of the copyright owner. Further reproduction prohibited without pe rm is sio n .
28
the conversion depends vary from one system to another. For example, in the case of 2Ti+ty
—►2TiN, TiN dilution lowers the Tc below the melting point of the Ti metal, and can thereby
increase the permeability of Ni, which leads to the increase of the degree of conversion of Ti
to TiN [Esl90]. In contrast, in the case of NbN, 2Nb+N2 -* 2NbN, melting of Nb is not a
consideration, so NbN dilution causes a monotonic decrease in Tc and hence of conversion
[Moor95a],
2.1.5b Green density
An increase in green density usually increases the amount of material reacted and
raises the combustion temperature. In the solid-solid reactions, the effect of green density on
the propagation rate is attributed to a balance between having high enough particle contact
to aid reaction but not too much to lead to excessive heat loss from the reaction zone as a
result of the increased thermal conductivity. At low density, the heat in the combustion front
is not effectively transferred to the pre-combustion zone and this can lead to unstable
combustion or extinction of the reaction. Similar results are obtained at high densities due to
rapid heat transfer from the combustion zone. An optimal density of the compact is desired
to obtain steady state combustion. In the case of solid-gas reactions, the combustion wave
velocity decreases with increasing density because of the limited access of the gaseous
reactant to the surface of the solid reactant. In the solid-liquid reactions, one reactant melts
and spreads through the pores by way of capillary forces. At lower density, smaller capillary
forces and the lack of reactant contact lead to incomplete reaction. At higher density, the
liquid is able to spread more quickly and completely, resulting in the reaction being
complete and the obtained product being of low porosity.
Logan and Walton found that in comparison with a pressed pellet, a loose powder
mixture of 3TiO: +3 B2O3 +10A1 was easier to ignite in a heated furnace, i.e. simultaneous
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
combustion mode, but more difficult to ignite by using a resistant heating element in the
propagation mode [Log84],
2.1.5c Sample diameter
Since heat loss during the combustion synthesis significantly influences combustion
temperature, propagation rates and also stability, the geometry of a green reactant sample
becomes an important parameter due to the effect of dimension on heat losses. It has been
reported that the combustion rate increases as the diameter of a sample increases and
remains constant after the diameter reaches a threshold value, which is dependent on the
reaction system [Fra85]. Lower combustion rates at small diameters are a consequence of
high radial heat losses. Consequently, there exists a critical diameter below which the
combustion wave becomes unstable and is finally extinguished.
2.1.5d Particle size, particle size distribution, and particle shape
Fine particles are desirable for a difficult-to-combust system. A decrease in the
particle size of the reactants tends to increase the combustion temperature and the
propagation rate of the combustion wavefront. Smaller particle size and narrower
distribution of reactant particle size increase the amount of reacted materials. When Ti
particle size is 45pm, a narrow combustion zone with high degree of conversion is seen. As
particle size increases, the combustion zone widens and the conversion during after bum
becomes significant [Shki81]. The particle size influences not only the rate of the reactions
but also the nature of the product formed [Yam86]. For example, the results of the
combustion synthesis of Ni-Ti intermetallic indicated that the degree of fineness of the NiTi2
structure increases with decreasing Ni particle size. This may be due to an increased surface
area of oxides on these fine reactant particles resulting in increased nucleation sites [Yi89].
The shape of particles was also found to affect the combustion process. Rice et al. [Ric91]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
showed that the use of foil, sheets or flakes of Ti or C prevented ignition in the Ti+C
TiC
reaction system.
2.1 ,Se Heating rate and ignition power
An effect of heating rate on die SHS reaction is expected when one considers the
kinetics of the reaction. A low heating rate can give rise to a certain degree of diffusion
between the reactant particles prior to initiation of the required exothermic SHS reaction.
This interdiffusion may result in the formation of intermediate products, which can seriously
affect the overall reaction. When the heating rate is increased, the total reaction time
decreases, limiting the formation of intermediate products. However, high heating rates can
produce high thermal gradients and possibly result in products with inhomogeneous
microstructure and composition distribution.
The heating rate of the reactants is dependent on the power input and can
significantly influence the combustion synthesis reaction. Lebrat has reported that, in the
synthesis of M 3AI using tungsten coil, the time required for heating the sample prior to
ignition increases significantly as the ignition power decreases [Leb92]. The ignition power
was controlled by changing die magnitude of the current heating the tungsten coil.
In the combustion of 3Ni-Al using a heated tungsten coil, the product microstructure
was found to vary with the ignition power input [Var92]. By using high ignition power, the
temperature of the top surface of the pellet is raised relatively quickly to the ignition
temperature, while the rest of the pellet remains near room temperature. Therefore, as the
reaction front propagates, more energy is lost by conduction to the cold unreacted part,
leading to an incompletely developed microstructure. When using a lower ignition power,
preheating of the entire sample takes place, and fully reacted homogeneous microstructure is
obtained.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 1 5f Initial ignition and combustion temperature
Figure 2.7 illustrated the temperature profile associated with the passage of a
combustion front. The ignition temperature is the temperature which the reaction ignites,
while the adiabatic temperature is the temperature at which the reaction propagates,
assuming no heat loss. However, realistically as ignition takes place, the sample’s
temperature rises abruptly to the combustion temperature which is lower than the adiabatic
temperature owing to the heat loss to surroundings.
Adiabatic
Temperature
Combustion Temperature
1 < — Ignition Temperature
Initial Temperature
Time
Figure 2.7 Schematic representation of the temperature profile associated with the passage
of a combustion front.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
The combustion temperatures (Tc) affect the reaction mechanism, reaction sequence,
and possibly invoke intermediate reactions that may take place in the SHS reaction system.
Henshaw et al. [Hens83] observed that if Tc lies between Tm(melting point) of the reactants,
rigid materials with fine pores were formed, e.g. TiC, SiC. If Tc>Tm for both reactants and
products, reactions occur in liquid phase and severely distorted solid were formed, e.g. TiNi,
ZrSi. If Tc<Tm for all species, complete solid-state combustion occurs and powdery product
were produced, e.g.
B 4C
and WC. In addition, it has been found that a higher combustion
temperature will give rise to a higher rate of combustion because the kinetic rate of the
reaction is proportional to the reaction temperature.
The combustion temperature is dependent on both ignition and initial temperature.
The larger the difference between the ignition and initial temperature is, the lower the
combustion temperature becomes. This is because more heat energy, generated from the
reacted layers, is consumed in order to increase the temperature of adjacent layer up to the
ignition temperature. Generally, ignition will occur at the moment when the rate of heat
arrival from an external source is equal to the rate of heat generated from the chemical
reaction [Eni68], Therefore, the ignition temperature is dependent not only on the chemical
characteristics of the reactant mixture but also on the energy of the external heat source. It
has been shown that the ignition energy needed to ignite a bulk condensed sample is one to
two orders of magnitudes higher than that for a loose powder system [Moor95a], This
difference is due to the significant heat losses occurred in igniting a low surface area, bulk
condensed reactant, and, therefore, high ignition temperatures are expected.
The initial temperature can be controlled through a pre-heating element, which heats
the reactants to a certain temperature. Initial temperature affects not only combustion
temperature and wave velocity but also the stability of the reaction [Zha91]. Both
combustion temperature and wave velocity increased proportionally with the initial
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
temperature. Decreasing the initial temperature leads to an increase in the amplitude of the
non-steady state combustion and period of time required for establishment of steady state
combustion [Moor95a].
It has been shown that the nature of the product is also dependent on the ignition
temperature. In the combustion of Zr+C in a hydrogen atmosphere, the product was found to
be a hydride when the ignition temperature was low, and a carbide when the ignition
temperature was at its highest [Mar85]. The microstructure of the product was also found to
be influenced by the initial temperature in the combustion synthesis of B4C-TiB2 composites
from their elements [Hal88].
In summary, combustion synthesis offers many potential advantages over
conventional techniques of synthesis. High purity materials can be produced by combustion
synthesis quickly and at low cost. Generally, the combustion process is ignited at the surface
of the material by thermal radiation. This thermal energy can come from a laser beam, a
heating coil, or an isothermal furnace. An alternative heating source to ignite the material is
microwave energy. Microwave energy is unique in its internal heating. This distinct feature
has been used to ignite reactions in the interior of the sample. A relatively uniform
combustion reaction wavefront propagates radially outward until reaches the surface of the
sample. This can lead to more complete conversion of reactants. A through literature review
on microwave heating is needed to provide insight into the technique and it’s potential as an
alternative approach for igniting the combustion synthesis.
2.2 Microwave Heating
Microwaves are coherent, polarized electromagnetic waves that fall between
radiowave and visible light on the electromagnetic radiation spectrum. The frequency of the
microwaves covers a range of 0.3 to 300GHz, which respectively corresponds to a
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
34
wavelength of lm to 1mm. Typical frequencies for materials processing are 915 MHz, 2.45
GHz, 5.8 GHz, and 24.124 GHz. Figure 2.8 shows the electromagnetic spectrum including
the microwave region.
uM
»
(0 -
MP
(M*
100
MHz
m uuuuu.
GHz
UHF
HF
10 m
1m
1 cm
l|u n
Figure 2.8 Electromagnetic spectrum and frequencies used in microwave processing
[Sut93].
Microwaves were first controlled and used during the second world war in radar
systems. The usefulness of microwaves in the heating of materials was first recognized in
1946, shortly after radar equipment was invented. Raytheon introduced the first microwave
oven to the marketplace in 1952 [Dec86], The application of microwaves around that time
was primarily for heating water and the technology was quickly adopted for drying and food
processing. Microwave processing ceramics was found to be feasible by Tinga and Voss in
the tnid-1960’s [Tin68]. In 1975, Sutton observed that in addition to heating water during
drying of high alumina cement, microwaves also heated the ceramic [Sut88]. Since that
time, many investigations on the application of microwaves for material processing have
been made. In the late 1970s and 1980s, the microwave heating and sintering of uranium
oxide [Haa79], barium titanites [Hum80], ferrites [Kra81], glass-ceramics [Mac84], and
aluminas [Mee87a], among others, were investigated. The reasons for the growing interest in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
microwave processing over conventional processing techniques include shorter processing
times, reductions in processing costs, improved material properties, and synthesis of new
materials [Sut89].
2.2.1 Microwave - Material Interaction
Microwaves are a part of the electromagnetic spectrum, which obey the laws of
optics. Microwaves can be transmitted, reflected, or absorbed depending on the dielectric
properties of the materials of interest (Figure 2.9).
A A A A A A A /W
tAAA
Material type
Penetration
TWssPAPgfr
Total
2E£StiE
Nona
(Reflected)
ABSffiBBEB
Partial
to Total
ABSORBER
(Mxad)
(a) Matrix s Iowl oss insulator
(b) Fbar/particlaa/addtfvos =
(abaofbing materials)
Partial
to Total
(Low low insulator)
(Conductor)
(Loaty Insulator)
Figure 2.9 Interaction of microwaves with materials [Sut89],
Materials with low conductivities, such as insulators (e.g., alumina, silica), allow
microwaves to pass through and are considered to be transparent to microwaves. On the
other hand, materials with high conductivities such as metals reflect microwaves and are
considered to be opaque. However, in very small metal particles, microwaves will penetrate
slightly into the surface and may create surface currents, which leads to surface heating of
these materials [Bes91].
Other ceramic materials such as semiconductors, silicon carbide, carbon, and others,
will absorb microwaves, which can be effectively heated from room temperature through the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
interaction of the materials with microwaves. The last category of material is a mixed
absorber, in which the matrix is a low loss insulator that contains an additive, which absorbs
microwave energy. The additive phases will absorb microwave energy far more rapidly than
the matrix and, hence, will be heated selectively and rapidly. When heated to above a critical
temperature, the matrix will begin to absorb and couple more efficiently with microwave
radiation. An example for this type of materials is the composite materials of interest in this
work with titanium oxide behaving as an insulator, aluminum as a conductor, and carbon as
an absorber.
2.2.1a Dielectric behavior
When microwaves penetrate and propagate through a dielectric material, the internal
electric field (E) is generated within a specific volume. This electric field induces
polarization and motion of charges. The resistance to these induced motions, due to inertial,
elastic, and frictional forces, causes losses and attenuates the electric field. As a result, these
losses lead to volumetric heating.
There are primarily two loss mechanisms by which microwaves interact with the
materials. These are polarization and conduction processes (ohmic conduction) [Sut89].
Conduction involves the long-range motion of charge carriers, electrons or ions. As the
charged particles move, a current is induced. In metals or semiconductors, the electrical
charge is carried by electrons, in which case electronic conduction occurs. For ionic
materials, the charge is transported by ions. The ionic conduction losses occur when
electrical charges move and collide with other species in the materials. The conduction
losses dominate at low frequencies. They decreasewith increasing frequency due to a
decrease of the time allowed for transport in the direction of the field.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
37
The situation is quite different in polarizable processes. Instead of resting on the
motion of electrons or ions, polarization involves the short-range displacement of charge
through the rotation and orientation of electric dipoles. Polarization results in losses as a
result of the dipoles resistance to oscillation and/or rotation under an alternating field. These
polarization losses dominate at high(er) frequencies.
2.2.la. 1 Mechanisms n f polarization
There are four primary mechanisms of polarization in ceramics [Hen90]. These
include electronic polarization
( P e) ,
atomic polarization
(P«),
dipole polarization
(P d ),
and
interfacial polarization (P i) (Figure 2.10). Each mechanism has a certain operable frequency
range in the electromagnetic spectrum as shown in Figure 2.11.
cation
Q ^ / v w —^
atom M
atom N
♦
—
O-V W W W W V—0
"—
Na*
Na*
I -------
(6 )
Na*
Na*
**—-Haaroda*
<«>
(O
I
</>
Figure 2.10 Schematic of the different polarization mechanisms, (a) Electronic, (b) Atomic
or ionic, (c) High frequency oscillatory dipoles, (d) Low frequency cation dipole, (e)
Interfacial space charge polarization at electrodes, (f) Interfacial polarization at
heterogeneities [Hen90],
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Inurfacial polarization
Dipole polarization
(low fr*q.) (high frequency)
II •
Atomic (or ionic)
polarization
14 •
tslcctronic
polarization
12 »
42 -
I01
MT
(« )
Log frequency
10Log frequency
Figure 2.11 Frequency dependence of the polarization mechanisms in dielectrics, (a)
Contribution to the charging constant (representative values of k’). (b) Contribution to the
loss angle (representative values of tanfi) [Hen90].
Electronic polarization results from the displacement of the electron cloud in the
atoms relative to the positively charged nuclei. This process acts so fast (10'15second) that it
can follow the ac field and corresponds approximately to the frequency of ultraviolet light.
The optical properties of materials are strongly dependent on this polarization mechanism.
Atomic polarization arises from the displacement of atoms inside the molecule with
respect to each other. This process requires ca. 10*12 to 10'l4second, depending on the bond
strength between the atoms. This polarization gives rise to resonance absorption in the
infrared range.
Dipole polarization, orientation polarization, or ion jump relaxation involves the
perturbation of the thermal motion of ionic or molecular dipoles, producing a net dipolar
orientation in the direction of die applied electric field. The mechanism of the dipole
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
polarization can be divided into two types. The first type is the rotation of die permanent
dipole against the elastic restoring forces about an equilibrium position. The time required
for this process is in the order of 10*lo-10'12second at room temperature (Figure 2.10c). This
mechanism is important for a variety of liquids, gasses, and polar solids. The second
mechanism of dipole polarization involves the rotation of dipoles between two equilibrium
positions (Figure 2.10d). The longer time required for this process is in the range of KT’-IO'6
second, which is due to the appreciable atomic distances involved in the ionic transitions.
The losses related to this mechanism sometimes are referred to as migration losses because
the cations that contribute to polarization are the same as those that contribute to dc
conductivity.
Interfacial or space charge or Maxwell Wagner polarization occurs in heterogeneous
materials when two phases differ from each other in the dielectric constant and conductivity.
The interfaces between phases, impurities or second phases, act as physical barriers that
inhibit charge migration. The accumulation of charge at the barriers results in a localized
polarization. The interfacial polarization is observed over a broad range of frequencies (10°
- 103 Hz).
Of all the possible forms of loss mechanisms, dipole and interfacial polarizations are
perhaps the most important for microwave heating because they can occur over the
frequency range of microwaves [Met83]. As the frequency of the electric field increases, the
rotation of the dipoles cannot follow, and the net polarization in the materials is no longer in
phase with the electric field. In this case, the resistance to dipole motion is equivalent to
large damping effect, resulting in relaxation type absorption. On the other hand, the
electronic and atomic polarizations result in resonance absorption peaks. The dipoles act so
fast that the net polarization observed under an electric field at microwave frequencies is in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
phase with the field, resulting in large restoring forces and small damping effects. As a
result, these two types of polarization do not generally contribute to microwave absorption.
2.2.1a.2 Dielectric constant and losses
The polarization, P (in C/m2), is related to the dielectric constant (referred to as
permittivity, e \ F/m) of a material. This relationship can be established by considering the
total electric displacement field or the electric induction, D (in C/m2).
The electric displacement field (D) in a material is given by
D = eaE + P
(2.20)
where So is dielectric constant of free space (e«, =8.86*10‘l2F/m) and E is the applied electric
field (V/m).
The dielectric constant (s’, F/m) of a material can be expressed as
(2.21)
The dielectric constant is often identified as a measurement relative to the dielectric
constant of free space (&o). The relative dielectric constant (e r) is expressed as
(2.22)
Substitution of equations (2.21) and (2.22) into equation (2.20) yields
e0e'rE = e0E + P
(2.23)
P = e0{ e ,-l)E
(2.24)
Rearrangement gives
where the quantity (Erl) is the susceptibility (x)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
The polarization (P), using the concept of dipole polarization, can be expressed as
P = N/u
(2.24)
where N is number of dipole per unit volume (1/m3) and p is average dipole moment (C-m).
p can be expressed as
fi = qd
(2.26)
where q is charge (C) and 5 is the separation distance (m) of the positive and negative
charge. The polarization is then given by
P = NqS
(2-27)
By making an analogy between equations (2.24) and (2.27)
e0{er -\)E = Nq5
(2.28)
Rearranging this expression yields
(2.29)
This equation shows that material properties influencing the dielectric constants includes
number of dipoles, charge of dipoles, and the separation distance between charges.
The complex interaction of sinusoidal applied voltage with a dielectric material can
be represented by a vectorial representation of two induced currents, the charging current
(Ic)
and the loss current (Ii) (Figure 2.12).
In an ideal dielectric, there would be no free-ion conduction and no loss currents.
Therefore the total current would be only the charging current (Ic), which would lead die
voltage by 90°. In real dielectrics, there are loss currents (It) arising from two sources: dc
conduction losses (!&) and dielectric losses ( l a c ) . The dc conduction losses involve the long-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
range motion of charge carriers. The dielectric losses, referred to as ac conduction losses,
result from the dipoles resistance to oscillation and/or rotation under an ac field as discussed
earlier. Both loss currents are in phase with the applied voltage. The total current (Itoui) in a
real dielectric therefore corresponds to the summation of Ic and Ii (note that Ii = Iac+Idc). As
illustrated in Figure 2.12, the total current leads the voltage by an angle (90°-8), where S is
referred to as the loss angle.
Imaginary
-jRcal
Figure 2.12 Vector diagram of charging, loss and total currents in a dielectric [Hen90].
The concept of charging and losses in a dielectric can also be expressed by using a
complex permittivity, e*:
e* = e'~ie"- sa{er - is,)
(2.30)
where i is (- 1 )1/2. e and e'r characterize the charging behavior (polarizability) of a material in
an electric field and are called the dielectric constant and relative dielectric constant,
respectively, s’ and e"r indicate the material’s ability to store the energy and are called the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
dielectric loss factor and relative dielectric loss factor, respectively [Bru8 8 , Cla96, Kat92,
Met83, Sut89],
The total dielectric losses for a given material result from polarization loses and
conduction losses (previously discussed). These losses can be superimposed on one another,
making it difficult to separate the individual loss from the total losses measured. Therefore
these loss mechanisms are all combined into an effective relative dielectric loss factor,s’eff.
s'# (oj) = e'r(o)) + — = e\ (a>) + e] (m) + e 0(o>) + s' (to) + —
e0o)
e0m
(2.31)
where s"r(a>) is the losses associated with the polarizations, which are frequency dependent
and Odc/eoto.relates to the losses from the dc conductivity. The subscripts in e, a, o, and /
refer to electronic, atomic, orientation, and interfacial, respectively, co is the angular
frequency (co =2 nf) and Ode (fi"‘m'1) is the total effective d.c. conductivity.
The complex dielectric constant is thus given by [Met83]
e* = eQ{et -ie'tff)
(2.32)
In practice however, the loss tangent (tan5), also known as the loss angle or the
dissipation factor, is usually used to describe these losses. The loss tangent is indicative of
the ability of the material to convert absorbed microwave energy into heat and is given by:
ta n J = ^ =—
er 2jtfs0er
(2. 33)
where o represents the conductivity contributed by short range polarization and long range
conduction losses.
The tanS value provides a useful indication of the type of the interactions between
microwave and materials (i.e., microwave-material interactions). Transparent materials with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
very low loss tangents (tan5<0.0l) [Bin93], such as insulators, are poor absorbers and allow
microwaves to pass through. On the other hand, opaque materials with extremely high loss
tangents such as metals, reflect microwaves. Materials with intermediate loss tangents, such
as semiconductors, silicon carbide, carbon, and others, will absorb microwaves which can be
effectively heated from room temperature through the interaction of the materials with
microwaves.
2.2.1b Microwave power dissipation
Microwave heating results from the dissipation of electromagnetic energy inside the
material. The power absorbed by the material (per unit volume, W/m3) is defined as [Met83,
Sut89, Tin8 8 ]
or
£ = <t|£|! = 24e„s, tan £|£ J !
(2.34a)
£
(2.34b)
= <7 |£|: = a fr,£ ;ta n 5 |£ ,|:
where Emu = E J^l. E^s is the root mean square internal electric field (V/m) used to
described the average electric field, Eo is the amplitude of electric field and f is frequency
(Hz). This assumes that the field is uniform throughout the volume and that the material is in
thermal equilibrium.
When a material absorbs microwave energy, it’s temperature rises. The heat required
for increasing the temperature, AT, of a material of mass M can be written as
Q = MCpAT
(2.35)
where Cp is the specific heat of material. The power required to increase the temperature of a
material, P, is Q/t and density, p, is M/V, thus the heating rate for a material irradiated by
microwaves can therefore be expressed as
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
AT
or
ttfe0£r tan<?|E0 | 2
(2.36b)
fC,
Equation (2.36) shows that the important parameters m microwave heating are
frequency, relative dielectric constant, loss tangent, and electric field inside the material.
Microwave power is usually attenuated as it penetrates through an absorbing
material. The depth of penetration, Dp, is a useful parameter for describing this phenomenon.
The penetration depth of microwave energy is defined as the distance into the material at
which the incident power drops to 1/e (36.8%) of the surface value. Mathematically, it is
defined by [Met8 8 ]
-t
(2.37a)
-i
or
(2.37b)
where K, is the incident or free-space wavelength. The important parameters for depth of
penetration are an effective relative dielectric loss factor,e eg- (e"efr = s r tanS), and the free*
space wavelength of microwaves. When the penetration depth is greater than, or within the
range of the sample dimensions, the microwave heat generation is volumetric and uniform
heating is more likely to be achieved. However, in the opposite case, penetration of
microwave energy will be limited and surface heating will occur, making uniform heating
impossible. Microwaves are usually reflected by bulk metals due to free electrons in metals
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
moving in a synchronized motion with the electric fields. However, in very small particles,
microwave will penetrate slightly into surface and may create surface currents, which leads
to surface heating. The depth of penetration for metals is known as skin depth, which can
range from less than 0.1 pm to greater than 30 pm [Von54]. Higher conductive metals have
smaller skin depths. For example, at 2.45GHz, the skin depth of copper and chromium are
1.34 pm and 3.61 pm, respectively [Bes91]. The skin dept, 8 , is given by
5=
1
- = 0 .0 2 9 ^
(2-38)
where f is the microwave frequency, p is the magnetic permeability, a is the electrical
conductivity,p is resistivity of material, and K is the incident wavelength [Hip54, Kat92,
New91].
The relative dielectric constant (s r) and the loss tangent (tanS) are the most widely
used parameters that describe the behavior of a dielectric material under the influence of
microwaves. They affect the power absorbed and the depth of penetration; therefore they
influence the volumetric heating. Stable microwave heating depends on the rate of
microwave power absorption and on the ability of the sample to dissipate the resulting heat.
If the temperature dependence of the power absorption is less than the temperature
dependence of the heat dissipation at the surface of the specimen (plus insulation system),
stable heating should be observed. The rapid rise in dielectric loss factor with the
temperature is the major issue in thermal runaway and temperature non-uniformity.
The typical behavior of the effective loss factor as a function of temperature of a
material when exposed to microwave energy is shown in Figure 2.13 [Met83]. The loss
factor is gradually increased with the increasing temperature until the material’s temperature
reaches the critical temperature (Tc) where rapid increase of the loss factor occurs. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
uncontrollable heating or overheating due to an increase in the effective loss factor as a
function of temperature is referred to as thermal runaway {Met83]. In practice, thermal
runaway does not occur throughout the volume of the sample. The area of the sample that
first exceeds the Tc will continue to heat rapidly at the exclusion of the rest of the sample
and localized (hot spots) thermal runaway can occur. Thermal runaway may be controlled or
avoided by reducing microwave power level, using a variable microwave duty cycle (the
percentage of time that the microwave power is applied over a specific time interval), or
through the use of convective air currents to cool the sample surface [Spo95].
•H
Figure 2.13 Qualitative representation of the loss factor as a function of the temperature
[Met 83],
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
2.2.2 Microwave Processing of Materials
Microwave heating is fundamentally different from conventional radiant heating. In
conventional heating, materials are fired in kilns and furnaces with a radiant heat source,
typically gas burners or electric resistance elements. Heat is transferred from the heat source
to the surface of the material by a combination of convection (low temperature) and
radiation (high temperature) and then it passes to the cooler center of the material by thermal
conduction. Since most ceramics are good thermal insulators with a thermal conductivity
normally less than lW/mk, heat transfer can be very slow and die surface of the sample is
typically at a higher temperature than the center during heating.
In contrast to radiant heating, depending on the microwave absorption of the
materials, heat generation by microwave heating (MH) occurs internally throughout the bulk
of the materials, rather than originating from the external heating sources (assuming Dp is
large, see equation 2.37). The sample is generally at a higher temperature than the ambient
temperature of the surroundings, resulting in heat being lost from the surface of the sample
by conduction and radiation to the surroundings. Consequendy, the temperature gradient
within the sample is established in which the center is at higher temperature than the surface.
This is obviously the reverse of the temperature gradient generated by conventional heating.
Because of the nature of the power deposition into the material, microwaves possess
several characteristics that provide unique features that are not available m the conventional
processing of materials. Some of the key characteristics are penetrating radiation,
controllable electric-field distributions, rapid heating, selective heating (differential
absorption) of materials, and self-limiting heating. These characteristics, either singly or in
combination, represent opportunities and benefits not available from conventional heating.
Table 2.1 summarizes the features and benefits that associated with each of the key
microwave characteristics.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
49
Table 2.1 Key Characteristics, Features, and Benefits of Microwave Processing [Bin92,
Cla96, Kat92, Sut891__________________________________________________
Feature
Benefits
Characteristic
Penetrating
radiation, direct
coupling of
microwave energy
• Materials heat internally
• Reversed thermal
gradients (AT)
• Lower surface
temperatures
• Instantaneous
power/temperature
response
• Low thermal mass
Field distributions
can be controlled
Dielectric losses
accelerate rapidly
above Tcm
• Applicator can be
remote from power
source
• High energy
concentration in selected
region
• Optimize frequency and
power level for a
material with a given
size and shape
• Very rapid heating
Differential
coupling of
materials
• Selective heating based
on the dielectric
properties of each phase,
additive, or constituent
Self-limiting
• Selective heating ceases
after certain processes
have been completed
•
•
•
•
Potential to heat large samples uniformly
AT favors chemical vapor infiltration
Reduced skin effect on drying
Removal of binders and gases without
cracking
• Improved product quality and yields
• Materials and composite synthesis,
combustion synthesis
• Automation, precise temperature control
• Rapid response to microwave power
change; pulsed power
• Heat in clean environment, controlled
atmosphere or pressure
• Precise heating of selected regions
(brazing, welding, plasma generation,
fiber drawing)
• Process automation, flexibility, energy
saving
• Increase sintering and diffusion rate due
to high electromagnetic fields
• Rapid processing (2-S0x faster)
• Capability to heat materials in excess of
2000°C
• Ability to heat microwave “transparent”
materials above To*
• Rapidly densify materials with minimal
grain growth
• Reduce processing time, energy, costs
• Heating of microwave transparent
materials via additives, fugitive phases
• Hybrid heating (active containers)
• Synthesis new materials and
microstructures
• Selective zone heating (joining, brazing,
sealing)
• Controlled chemical reactions, oxidation,
reduction; use of micowave transparent
containers
• Drying, curing, annealing; matrix
infiltration
• Below critical temperature, drying and
curing are self-regurating
• Completion of certain phase change is
self-regurating
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
The inverted temperature gradient developed by microwave heating offers several
unique advantages, for example, rapid volumetric heating without overheating the surface,
especially with low thermal conductivity materials; reduced surface degradation during
drying; removal of binders or gases from the interior of sample without cracking; and
fabrication of new and unique structures of ceramics and composites.
The rapidity of microwave processing can inhibit grain growth and thus result in
products having more uniform, fine-grained microstructure. This very rapid processing also
saves energy and money. The ability to selectively heat specific phases can also be useful
for synthesis of new materials and microstructures as well as for applications such as
joining. Microwave hybrid heating is another example of where selective heating has been
used to significant advantage. In self-limiting absorption, microwave heating will cease once
the source of differential absorption has been removed or has been altered during a phase
change in the material during processing. This principle is also used in hybrid heating,
where susceptors are used initially to hybrid-heat low-loss materials from room temperature.
With these benefits, however, microwave processing also introduce many challenges
that need to be overcome, such as the reluctance to abandon proven technologies or the
compatibility of the microwave process with the rest of the process line (Table 2.2).
In some cases, a more even temperature profile is required. The first technique to
flatten out the temperature profile is referred to as “casketing” [Hol90], Samples are encased
within microwave transparent insulation, which absorbs minimal microwave energy and
prevent heat loss from the surface of the samples, generating a more uniform temperature
profile. A second technique to control temperature gradient is microwave hybrid heating
(MHH) [De90]. It can be achieved either by using a conventional heat source such as a gas
or electric furnace in combination with microwaves, or through the use of an external
susceptor that couples with microwaves. The susceptor will absorb microwave energy and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
re-radiate the heat back to the surface of the samples, reducing the temperature gradient. In
addition, the susceptors are used to aid in microwave heating of the materials with low
dielectric losses and low thermal conductivities, which are difficult to heat under the
influence of microwave at room temperature. Once a poorly microwave absorbing sample is
heated by the susceptor (with a high tanS) to its critical temperature, Tc, it will absorb
microwave effectively and cause self-heating (See Figure 2.13).
Table 2.2 Microwave Processing Challenges [Cla96, Sut89, Sut92]
Challenges
Characteristic
Penetrating radiation
•
•
•
•
•
•
Microwave transparent materials difficult to heat
Hot spots, cracking
Large AT in low thermal conductivity materials, and
non-uniform heating
Controlling internal temperature
Arcing, plasma
Require new equipment, designs special reaction vessels
Field distribution can
be controlled
•
•
Equipment more costly and complex
Required specialized equipment
Dielectric losses
accelerate rapidly
above Ton
•
•
•
Hot spot, arcing
Non uniform temperature
Control of thermal runaway
Differential coupling
of materials
•
•
Reactions with unwanted impurities
Contamination with insulation or other phases
Self-limiting
•
•
Undesired decoupling during heating in certain products
Difficult to maintain temperature
The use of a “picket fence” hybrid heating arrangement is an example of one types of
hybrid heating. SiC rods were inserted into the insulation that surrounded the sample to
assist heating [Jan92]. Unfortunately the use of these SiC rods may raise some questions
such as how many rods or how far from the sample to place the rods. Too few rods might
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
52
not heat sample efficiently and too many rods may also result in less efficient heating if all
of the energy is absorbed by the rods. Another SiC susceptor design is the use of an
insulating material, zirconia or alumina refractory cement, lined on the interior with SiC
granules [De91a, De91b, Moo91], More recently, a new type of susceptor has been
developed [Cla94, Coz95]. This susceptor consists of a castable alumina cement, with a
relatively low tanS, combined with SiC particles with a high tanS. SiC is mixed into the
cement before being set into shape and therefore, it can be formed into various ideal shape
and size for each particular sample. It can also be made such that the inner shape resembles
that of the sample. Thus, the temperature gradient due to differences in the proximity of the
sample to the susceptor can be avoided. This particular design for MHH will be used to
process some of the samples in this work.
2.3 Microwave Ignition and Combustion
Combustion synthesis and microwave processing have many attributes for material
processing. As discussed earlier, combustion synthesis, commonly called self-propagating
high temperature synthesis (SHS), offers many potential advantages over conventional
techniques of synthesis, including relatively simple equipment, shorter processing times,
lower energy requirements, higher product purities, and the possibility to synthesize
metastable phases. Likewise, the potential benefits brought about by the use of microwave
energy to process ceramic materials may include the reduction in manufacturing costs due to
energy saving and shorter processing times, improved product yield, improved or unique
microstructures, and synthesis of new materials [Sut89].
Generally, the SHS process is ignited at the surface of the material by thermal
radiation. This thermal energy can come from a laser beam [Dim89] or a heating coil
situated close to die sample surface [Mer83]. Alternatively, the entire sample can be heated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
to the ignition temperature in an isothermal furnace [Wan90]. In these types of combustion
synthesis methods, die surface temperature is raised to the ignition temperature (Tig) where
the materials react exothermically. The heat generated at the surface allows the adjacent
materials to react and a combustion wavefront propagates across the sample converting
reactants to products. Most impurities volatilize during the reaction, result in a highly porous
sample with a great amount of expansion.
A completely different approach for igniting the SHS process occurs when
microwaves are used as the heating source. Microwave heating is fundamentally different
from more conventional radiant element techniques. The nature of the power deposition into
the materials and subsequent inverted temperature profile appear to offer advantages to SHS
process. Microwaves tend to heat the entire sample nearly uniformly due to its internal
heating mechanisms. The surface of the sample radiates thermal energy leading to an
inverted temperature gradient where the interior of the sample is hotter than the surface. In
this approach, the sample ignites in the center and a combustion wavefront propagates
outward in a radial manner, which ensures complete combustion of the material. Figure 2.14
shows the temperature profiles resulting from the wavefront propagation of the SHS process,
which are ignited by different methods. [Dal90].
In conventional local ignition SHS (e.g. heating cod), the adiabatic combustion
temperature
(Tad)
is calculated by assuming that the heat produced from the exothermic
reaction is used to heat up the products to Tad while no energy transfer to the surroundings.
In this process, the heat of reaction can be determined from
= j AC p(product)dT
29S
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(2.7)
Samp*
Spm plp
L A S E R 10 INlTtON
METAL COIL HEA TING FILAM ENT
Radai
Sam p*
A t I, t h a W a v a a C r o a a O v a r
C O N V E N T I O N A L O V E N IO IN IT I O N
Sam pM
M I C R O W A V E IOI NI TI ON
(MICOM)
W avafront
Pro p ag atio n
Figure 2.14 Illustration of various ignition methods and propagation of the combustion
wavefront Pal90],
For combustion synthesis with microwave energy, the entire sample is heated. The
volumetric heating with microwave energy would raise the reactant temperature from its
initial temperature (usually 298K) to die ignition temperature (Tig). At this temperature, the
reaction will take place and the heat produced from the combustion will further raise the
product temperature to the final (adiabatic) temperature (Tad). The adiabatic combustion
temperature using microwave ignition can be calculated by considering the enthalpy of
reaction (-AH°p9 s) and the energy supplied by die microwave heating, AHmw
A/ / ; >298 + AHm = J &Cp(product)dT
(2-39)
29*
where
^11
~ J ACp(reactant)dT
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(2.40)
55
Equation (2.39) predicts a higher adiabatic combustion temperature using microwave
combustion due to the additional energy input from volumetric microwave heating.
In the conventional volume ignition SHS (e.g. conventional oven), the adiabatic
combustion temperature is also higher than that of the conventional local ignition SHS
[Bow91] and can be determined in the similar fashion as the microwave combustion case
(equation 2.40). The difference between these two ignition methods is that internal ignition
takes place in microwave combustion while the ignition in the conventional volume ignition
occurs at the sample surface.
Clark et al. have demonstrated die potential of microwave heating to support SHS
[Dal90, Cla91], Various materials have been synthesized using microwave energy, for
examples, TiC, SiC, B4 C, TiB2 , TisSi3 , AIN, and A^Cb-TiC. The reactants were compacted
to green densities of ca. 60-70% and ignited in a simple kitchen microwave oven (700watts,
2.45GHz). The SHS reactions initiated in the center of the green pellet and propagated
outwardly towards the surface. This was thought to be because of the internal heating
mechanism. Heat was generated within the sample itself by the interaction of microwaves
and the materials.
Clark et al. have also reported that high bulk density samples can be ignited by
microwave energy. In conventional ignition, the propagation of the combustion wavefront is
strongly dependent on thermal conductivity and the density of the compact. At high
compaction densities, excessive heat loss from the reaction zone, as a result of the increased
thermal conductivity, can lead to unstable combustion or extinction of the reaction. Rice et
al. [Ric85b] reported that TiC compacts with densities greater than 80% failed to ignite by
conventional ignition methods. Microwave energy can be used to overcome this limitation.
With microwaves, energy is absorbed continuously within the sample and this ensures that
the ignition temperature (Tig) is sustained. As the temperature of the sample increases, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
absorption of microwave energy by the materials increases [Dal90]. Higher absorption leads
to increased dissipation of microwave energy, resulting in a corresponding increase in
temperature. The center of the sample reaches the ignition temperature first and reacts
leading to an even higher combustion temperature (Tc). The high temperature in the interior
of the sample forces propagation of a relatively uniform radial combustion wavefront to the
exterior of the sample, which ensures complete combustion of the material [Ahm91, Cla91].
With microwave ignition and combustion, the dependence of the reaction on the thermal
conductivity and the density of the compact is greatly reduced as compared to samples
ignited with conventional methods.
Figure 2.15 illustrates the relationship of temperature versus time of the sample
during conventional and microwave heating. The capability of sustaining higher
temperatures when synthesizing combustible materials using microwave energy is indicated
in this figure (Figure 2.15b). After the reaction is over, the microwave power can be left on
to densify the sample further because at higher temperature it can absorb more microwaves
to help maintain this temperature (Ts). In conventional local ignition, this is not possible.
Because once the combustion has taken place, the temperature decreases rapidly. It also has
been reported that diffusion rates are higher in the presence of microwave energy [Jan89].
Therefore, with more controlled temperatures and enhanced diffusion rates, products may be
more dense and homogeneous than those obtained by conventional combustion.
The use of microwave energy for ignition and combustion also aids in releasing of
gases from the interior of materials. This is due to internal heating and the combustion
wavefront propagating from the inside outwards which drive the gases out. The gases are not
trapped inside as they are with conventional methods.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
b) Mtooweve Hasting
■""•Combuetion Tamparstura
a) Convarlional Hasting
Com(nation Temperature
s
IUJ
■Cooing
I-
TIME
Ignition
Temperatu*
Mtooaawa
rOn
TIME
^
Mtameeve
Power Off
Figure 2.15 Temperature versus time relationship when synthesizing combustible materials
with (a) conventional heating and (b) microwave energy [Adapted from Dal90] (Ts is
sustained temperature).
In conventional combustion, the combustion reaction is usually violent (or no
propagation at high densities). This results in a low-density sample with a large amount of
expansion. However, it has been demonstrated that the reaction rate can be controlled with
microwave energy. As shown in equation (2.36), the increase in temperature (AT) depends
on the density (p) of the sample. Increasing the density of the sample, which corresponds to
increasing thermal conductivity, results in reducing the heating rate. Therefore, the two
factors, high compact density and high thermal conductivity which limit or inhibit the
wavefront propagation, and the reduced microwave heating rates provide control of the
propagation of the combustion wavefront [Ahm91]. The control allows for a gradual release
of the volatiles, resulting in little or no expansion in the product. This offers great potential
for the fabrication of dense monoliths.
The rate of the propagation of the wavefront can also be controlled by the incident
microwave power. Turning the power off can terminate the propagation. Pulsing the incident
power by altering the duty cycle will give even more precise control on the velocity of
propagation of the combustion wavefront.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Radial wavefront propagation is another factor contributing to controlled
propagation. As the wavefront propagates radially outward, more surface area is
encountered, which absorbs the heat generated at the smaller adjacent surface. In other
words, the released energy during combustion must heat a larger area of material to the
combustion temperature. In planar wavefront propagation, a sample ignites at one surface on
a finite area and the combustion propagates to adjacent reactants having the same finite area.
Thus, radial wavefront propagation is slower, allowing for more control than conventional
planar wavefront propagation.
Diluent addition is an important parameter used to controlled combustion in
conventional synthesis. Excess of a reactant, or product, or addition of an inert filler will
normally decrease the heat generation rate and result in a reduction in the combustion
temperature (Tc) and reaction wavefront velocity. The reduction in heat will make the
reaction less violent. However, since control is possible with microwave energy, addition of
a diluent may not be required in many material systems. In some cases, diluents may absorb
microwave energy [Kat85] and increase temperature of the sample. Monoliths may not form
and compacts may show expansion.
Ignition times for samples of various masses and densities, synthesized using
microwave energy were also investigated [Dal90]. The time to ignite a sample was observed
to be dependent on the mass and density of the samples. Low density and/or larger mass
samples were ignited in shorter time as compared to samples having smaller mass or higher
compact densities. Approximately 3-5 minutes were required for a compact of TiCh+Al+C
reactant mixture weighing 3-4 g to reach ignition temperatures in a 700 watts microwave
oven. Samples weighing about 7-8 g required less than 30 seconds to reach the ignition
temperature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
In addition to removal of gases from the interior of the sample, another application to
which the internal heating and inverted thermal gradients developed by microwave heating
has been successfully applied is the penetration of reactive gases into the hotter interior
portions of materials and then condensation into solid matter prior to the filling of voids or
pores at the surface. Kiggans et al. [Kig91] showed that nitridation of silicon proceeds from
the inside of a powder compact to the surface under microwave processing. In this way, the
material was converted from silicon to Si3 N4 more completely than with conventional
processing. In addition, nitridation began at a lower temperature, 1200°C for microwave
compared to 1250°C for conventional annealing. Thomas et al. [Tho92] have also reported
the use of microwave heating for the processing of reaction-bonded SijN.» (RBSN). Their
results indicated that the fracture toughness of the microwave heated RBSN is higher than
that of the conventionally heated one with the same density. Likewise, many investigators
have used microwave energy to synthesize yttrium-barium-copper-oxide (YBCO)
superconductors [Ahm8 8 , Bag8 8 , Bin93, CIa8 8 ].
Generally, ceramic superconductors suffer from the major disadvantage that only
surface layers have the correct oxygen stoichiometry and are, therefore, superconducting.
This occurs because conventional sintering results in the surface densifying before the
interior, cutting off the latter from the oxygen-rich atmosphere required to achieve the
correct oxygen stoichiometry. This creates a shell of superconducting phases surrounding a
central core, which is non superconducting [Eki87]. Ahmad et al. [Ahm8 8 ] has
demonstrated that YBa2 Cu3 0 7 .x superconducting pellets can be prepared using microwave
energy. Samples were prepared from powders of Y 2 O 3 , BaCOj, and CuO. The microwaveprocessed pellets exhibited a more refined microstructure, lower porosity, improved oxygen
content, and higher superconductivity transition temperature (93°K versus 90°K) over
conventionally processed samples. This is attributed to the coupling of CuO with microwave
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
60
energy. Binner et al. [Bin93] also have shown that the use of microwave energy resulted in a
significantly
more uniform oxygen content throughout disk shaped superconductors by
generating an inverse temperature profile during sintering. Besides these, conventional
combustion synthesis (SHS) and Microwave-assisted combustion synthesis (MACS) were
also developed for the synthesis of thallium-containing superconductors using Cu metal
powder as a fuel [Bay94]. Bayya et al. reported that both types of combustion synthesis
reduce the reaction time, which in turn reduces the thallium loss due to volatilization.
However, the MACS technique produced a higher degree of reaction due to the fact that the
material was heated uniformly for a longer period of time (about 100s in MACS as
compared to a few seconds at the tip of the SHS combustion front).
Kumar et al. [Kum91] have demonstrated that microwave heating is a feasible
technique for the synthesis of nanophase materials. Ultra-fine ^-silicon carbide was
synthesized from amorphous silica and carbon in a microwave oven. These investigators
also performed comparative conventional experiments. They found that microwave
synthesized powders were 30-200 nm and conventional powders were 50-450 nm in size.
Kozuka and Mackenzie [Koz91] have synthesized SiC, TiC, NbC, and TaC from carbon and
their respective oxides using microwave heating. They also observed whisker formation,
which is used to toughen ceramic composites, during the synthesis of silicon carbide.
In 1993, Willert-Porada et al. [Wil93a] synthesized AkOj-TiC composite powder
starting from TiC>2 precursor compounds by microwave-pyrolysis of the precursor in the
presence of Al powder. The alcoholate, Ti(OCjH7 )4 (also known as TTP) and
acetylacetonato-alcoholate, Ti(0 2 C5H7 )2(OC3 H7 )2 (also known as TACP) were used as the
metallorganic precursor for Ti(>2 . The liquid precursor was mixed with the stoichiometric
amount of Al powder (particle size <45 pm) to yield Al and TiC>2 in a 4:3-ratio. The MWpyrolysis was conducted under Ar atmosphere in a rotary-evaporator type system. After
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TACP decomposed into a TixCyOz-polymer, subsequent annealing of the Al-loaded polymer
mixture was performed at 400°C and 800°C in a vacuum furnace prior to combustion and
sintering. When TACP was reacted with Al powder using microwave heating, after
annealing at 400°C, a TiC peak was detected together with TiC>2 signals. Further annealing
of the Al-loaded mixture at 800°C gave the first indication of y-alumina and Ti, while the
TiC signal disappeared. However, after a 1600°C microwave sintering in a multimode
2.45GHz/2.5 kW cavity, the XRD indicated a-AkCVTiC only. In contrast to TACP, which
absorbed microwave extremely well, pyrolysis of TTP/A1 mixtures using microwave heating
was impossible. Arcing and local overheating of the reactor made this precursor impractical
for microwave-pyrolysis, probably due to its low microwave-absorption ability. WillertPorada concluded that pressureless sintering of combustible mixtures of TiCVAl-C was
achieved at temperatures as low as 1600°C. It was thought that the skin effects of fine metal
dispersions combined with strong dipolar relaxation mechanisms were operative in mixtures
of Al powders with metallorganics. The strong dipolar relaxation of metallorganic
compounds should be the dominant dissipation path when they were decomposed by
pyrolysis using microwave heating. Willert-Porada also found that without additional
processing, no densification occurs upon microwave- or conventional sintering of the
precursor-derived AhCVTiC composite powder. Only after wet milling with additional
AI2O3 and MW-sintering, was a fairly dense ceramic was obtained. The microstructure of
this MW-sintering of the precursor-derived AkCb-TiC composite powder is similar to that
obtained from MW-sintering of a commercial TiOi-Al-C mixture powders. However, in
conventional oven-sintering of Ti0 2 -Al-C, a less dense ceramic was obtained. It was thought
that the formation of TiC prevents further densification.
In 1995, Yiin et al., [Yii95] demonstrated the ability to ignite the combustion
synthesis process 3 TiC>2 +3 C+(4 +x)Al -> 3 TiC+2 Al2C>3+xAl with microwave energy using
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
62
slow and fast heating rates. A higher starting power of -75 watts was required to ignite
sample in the case of fast heating compared to -50 watts starting power needed in the case
of slow heating. The process was studied for the case of x=0 and 4 with and without the
application of uniaxial pressure, in the range of 200-1400 psi. For x=0, the ignition
temperature, as measured at the side of the sample surface, was lower (300°C) for fast
heating compared to slower heating (650°C). There was no report for ignition temperature in
the case of x=4. In the case of x=0, the samples processed with fast and slow microwave
heating rates reached higher densities than those processed conventionally using the hot wire
technique. For x=0, all the slow heating rate cases produced an incomplete reaction, while a
uniform microstructure with 75% theoretical density was achieved with a fast heating rate.
The application of a uniaxial force during the SHS process produced 85% theoretically
dense homogeneous products for both the microwave (fast heating) and conventional
techniques. SEM micrographs clearly indicated differences in the formation of whiskers in
the products obtained from microwave and conventional technique. In the conventional case,
the excess amount of liquid Al and gases produced during SHS reaction results in the
formation of
A I2 O 3 -A I
hollow whiskers with bulbous heads. This behavior suggested a
vapor-liquid-solid mechanism of whisker formation. In the microwave case, the sample
possessed solid whiskers that were much smaller in size and did not have the bulbous head
feature. This was thought to be because microwaves ignited the sample internally and the
process quickly reached a much higher combustion temperature. The aluminum metal and
even alumina may vaporize due to this higher combustion temperature and such a vaporsolid reaction may result in the formation of solid A I 2 O 3 whiskers.
In summary, the above literature review reveals that there have been a few studies on
microwave-induced combustion synthesis of AUCh-TiC, used as a cutting tool and wear
resistance material [Dal90, Cla91, Wil93, Yii95]. Although these initial data were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
promising, so far little effort has been undertaken to systematically investigate mechanisms
of how microwave energy ignites and sustains a combustion reaction. In addition, most of
the previous work focused on the production of SHS products in porous monolith forms.
These studies take into account such variables as sample mass, sample density, reactant
precursor, or amount of liquid metal.
Interestingly enough, the effect of the reactants particle size has never been taken
into account in any of these studies. Microwave heating offers the benefit of very high
heating rates. When the heating rate is high, the total reaction time decreases, limiting the
formation of intermediate products. The high heating rate is mainly based on the skin effect
of fine metal dispersions which, depending upon the particle size and shape as well as the
spatial distribution of the metal, can display very high dielectric losses [Lor91], A reaction
rate control should be possible by adjusting the size and the spatial distribution of the metal
particles. The size of carbon should also have an influence on the combustion reaction
because of its high microwave-dissipation ability. Contact area between particles is also
important. When the carbon serves as a continuous phase, it may be possible for carbon to
act as a barrier to titania-aluminum particle contact, which is necessary for the initial
reaction to titania and aluminum [Bow94], However, if the dtania is instead a continuous
phase, the aluminum-titania particle contact is greater which may lead to the increased
combustion wave velocity.
Other important parameters used to control combustion in conventional synthesis
may also be used in microwave-induced combustion, for example, the addition of a diluent
and ignition power. The addition of diluent to the reactant powders will reduce the total heat
produced per unit volume. The exothermic energy will be released only by the volume
fraction of the reactants. The reduction in heat will make the reaction less violent. For
ignition power consideration, when using a lower ignition power, preheating of the entire
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
sample takes place, and a fully reacted homogeneous microstructure may be obtained.
Alternatively, the use of a susceptor to aid in microwave heating as hybrid heating will also
allow the sample to be heated uniformly.
Therefore it is the purpose of this work to synthesize and process AhOj-TiC powders
by using the combined technology of SHS combustion synthesis and microwave heating.
The first objective is to study the effects of reaction parameters, including compositions of
the reactants, particle sizes of reactants, heating method, and microwave power, on ignition
behavior. The second objective is to examine the effects of reaction parameters on the
characteristics of the resulting powders. Thirdly, it is desired to determine the important
reaction parameters that influence the microwave-ignition of the AkOs-TiC powders and to
develop an empirical model relating these parameters to ignition behavior and characteristics
of resulting powders. Lastly, this research will suggest mechanisms in synthesizing AI2 O3 TiC powders using microwave energy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
EXPERIMENTAL PROCEDURE
In this work, Al2 0 3 -TiC powders were synthesized by the SHS method using
microwave energy. The experimental procedure consisted of two stages. The first stage was
to study the relationships between raw materials, combustion process and characteristics of
the resulting powders. This was carried out using an experimental design approach. The
second stage involved studies on the mechanism of the microwave-combustion reaction.
3.1 Effects of Reaction Parameters Studies
AfeOs-TiC powders were synthesized by the SHS method using microwave energy
from a mixture of Ti0 2 , C, Al and A I 2 O 3 reactants. The reaction involved was as follows
3Ti02 + 3C + (4 + x)Al + yAl20 3 -> 37YC + (2 + y)Al20 3 +xAl
(3.1)
A series of experiments were conducted to study the effect of various factors on the
ignition time, ignition temperature, combustion time, combustion temperature, and
characteristics of resulting powders. The factors of interest were the addition of excess
reactant (Al), the addition of reaction product
(A I2 O 3 ),
titania particle size, carbon particle
size, aluminum particle size, alumina particle size, heating method, and microwave power.
Assuming only two levels of each factors of interest were to be evaluated, the total number
of test runs included in the full factorial experiment would be 2 s = 256 runs, which was
considered too large of a test matrix for practical considerations. To practically reduce the
total number of experiments without losing credibility of the results obtained, the PlackettBurman design, which is widely used in screening experiments [Pla46, Sch8 8 , Mas89], was
employed to achieve this goal.
65
R e p ro d u c e d with permission o f the copyright owner. Further reproduction prohibited without permission.
66
3.1.1 Screening Experiments; Plackett-Burman Design
Screening experiments are conducted in order to identify a small number of
dominant factors for subsequent more extensive investigation. Plackett and Burman suggest
the number of test runs in then designs to be a multiple of 4, i.e., n = 8 , 12, 16,...,100. In
each of these cases, n-1 factors can be studied. For example, a sixteenth-runs design can be
used to study up to IS factors. To allow for an estimate of experimental error, it is
recommended that at least three dummy factors be included in the experiment. In this work,
there are
8
factors of interest with an addition of 7 dummy factors. Thus the sixteenth-run
screening experiment was conducted and satisfied all the recommended requirements. The
chosen factor and investigated levels are shown in Table 3.1. A minus sign (-1) denotes one
level of a factor and a plus sign (+ 1 ) denotes the other level of the factor.
To build a sixteenth-run design, the design generator, consisting of the following IS
values (i.e., + + + + - + - + +
was used as the first row of a sixteenth-run.
Subsequent rows were obtained by taking the sign in the first position of the previous row
and placing it last in the current row, and then sliding all other signs forward one position.
After all fifteenth rows were complete, a final row of minus signs was added. The complete
sixteenth-run design for this work is shown in Table 3.2.
To minimize the possibility of bias effect, the experiment test sequence and die
assignment of the factors to the columns were randomized. The actual factor levels for a
sixteenth-run screening design for the microwave-induced combustion synthesis of A I 2 O 3 TiC composites study are shown in Table3.3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
Table 3.1 Factors and Investigated Levels
Factor
(-D
(+1 )
A- Amount of excess Al
0
wt%
40 wt%
B- Amount of excess AI2O3
0
wt%
40 wt%
C- Titania particle size
0.9-1. 6 pm
1-5 pm
D- Carbon particle size
0.017 pm
0.075 pm
E- Aluminum particle size
3-4.5 pm
-2 0 0 mesh
F- Alumina particle size
0.49 pm
40-45 pm
microwave heating (MH)
microwave hybrid heating
(MHH)
800 watts
1600 watts
I- Dummy factor
-I
+1
J- Dummy factor
-1
+1
K- Dummy factor
-1
+1
L- Dummy factor
-1
+1
M- Dummy factor
-1
+1
N- Dummy factor
-1
+1
0- Dummy factor
-1
+1
G- Heating method
H- Microwave power
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Table 3.2 Sixteenth-Run Screening Design
Factors
Run
A
B
c
D
E
F
G
H
I
J
K
L
M
N
0
1
+1
+1
+1
+1
-1
+1
-1
+1
+1
-1
-1
+1
-1
-1
-1
2
+1
+1
+1
+1
-1
+1
+1
-1
-1
+1
-I
-I
-1
+1
3
+1
4-1
-1
+1
-1
+1
+1
-1
-1
+1
-1
-I
+1
+1
4
+1
-1
+1
-1
+1
+1
-1
-1
+1
-1
-1
+1
+1
+1
5
-1
+1
-1
+1
+1
-1
-1
+1
-1
-1
-1
+1
+1
+1
+1
6
+1
-1
+1
+1
-I
-1
+1
-I
-1
-1
+1
+1
+1
+1
-1
7
-1
+1
+1
-1
-1
+1
-1
-1
-1
+1
+1
+1
+1
-1
+1
8
+1
+1
-1
+1
-1
-1
-1
+1
+1
+1
+1
-I
+1
-1
9
+1
-1
-1
-1
-1
-1
+1
+1
+1
+1
+1
-t
+1
10
-1
-1
-1
+1
+1
+1
+1
-1
-1
+1
+1
11
-1
+1
-1
-I
-1
+1
+1
+1
+1
-1
+1
+1
+1
-1
12
+1
-1
-1
-1
+1
+1
+1
+1
-1
+1
-1
+1
+1
-1
-1
13
-1
-1
+1
+1
+1
+1
-1
+1
-1
+1
+1
-I
-1
+1
14
-1
+1
+1
+1
+1
-1
+1
-1
+1
+1
-1
-1
+1
-1
15
-1
+1
+1
+1
-1
+1
-1
+1
+1
-1
+1
-I
-1
16
-1
+1
+1
+1
+1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-1
69
Table 3.3 Actual Factor Levels for a Sixteenth-Run Screening Design for the Microwave'
Induced Combustion Synthesis of AkC^-TiC Powder Study______________________
Factor
Run
Wt% excess
Panicle size of reactant (nm)
Heating
method
MW
power
(w)
Dummy
d
d
1
2
d
3
1600
-I
I
I
MHH
800
-1
1
-1
0
MHH
800
-1
-1
1
3 -4 .5
40-45
MH
1600
1
0.017
3 -4 .5
40-45
MH
800
-1
0.9-1.6
0.075
3 -4 .5
0
MH
1600
40
1-5
0.017
- 2 0 0 mesh
0.49
MHH
1600
40
40
0.9-1.6
0.017
- 2 0 0 mesh
0.49
MH
9
0
0
0.9-1. 6
0.075
- 2 0 0 mesh
0
10
0
40
1-5
0.075
•2 0 0 mesh
11
0
0
0.9-1.6
0.017
12
0
0
1-5
13
0
40
14
0
40
15
40
16
40
d
5
d
4
6
d
7
1
-1
-I
I
1
1
1
1
-1
-I
1
-I
-1
1
1
1
1
-1
1
I
1
1
1
-1
1
-1
-1
1
-1
-1
I
800
I
1
1
1
-1
1
-1
MHH
800
1
-1
1
1
-I
-1
I
0.49
MHH
800
1
1
-I
-1
1
3 -4 .5
0
MH
800
-1
-1
-1
-1
0.017
3 -4 .5
0
MHH
1600
1
I
1
-1
I
1
0.9-1.6
0.075
- 2 0 0 mesh
0.49
MH
1600
-I
-1
1
1
1
1
0.9-1. 6
0.017
3 -4 .5
40-45
MHH
1600
1
1
-1
1
1
-1
0
1-5
0.017
-2 0 0 meah
0
MH
800
I
-1
-I
I
1
1
0
0.9-1.6
0.017
- 2 0 0 mesh
0
MHH
1600
-1
1
1
-1
-1
Al
AI2 O3
TiO,
C
Al
A 1A
1
0
0
1-5
0.075
-2 0 0 mesh
0
MH
2
40
40
0.9-1.6
0.075
3 -4 .5
40-45
3
40
0
1-5
0.075
3 -4 .5
4
40
40
1-5
0.075
5
0
40
1-5
6
40
0
7
40
8
-1
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
d
-1
70
The data on the ignition time, ignition temperature, combustion time, combustion
temperature, and also characteristics of powders (particle size, surface area, and density)
were collected. Both graphical and statistical analyses were performed to assess the
significance of factors and to reduce a large number of factors to a manageable subset of
important factors. The following is the method used to determine the significance of the
factors. The first step is to find the effect of a factor on the response:
N/2
N j2
where Ei is Effect of factor i, R is Response or Result, and N is total test run. Next, the
“Dummy Effects” are used to estimate the variance:
Veff=
n
(3.3)
where Veff is Variance, Ed is Effects shown by a dummy factor, and n is number of dummy
factors. Then the standard error of the factor will be calculated from the following
relationship:
S.E.eff = jV e ff
(3.4)
The significance of each factor can then be determined by using the t-Test in which
the number of degrees of freedom for entering the tabulated values of t equals the number of
dummy factors.
Ei
t - ———
S £ .e ff
(3.5)
Once important factors were identified, more extensive investigation involving only
these dominant factors were conducted. The remaining (less important) factors were set at
levels which were based on, e.g., convenience, economics, or providing less variability in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
the response. Since the qualitative factor (heating method) is of interest, regardless of its
significance, two experiments for each level of this factor (MH and MHH) were conducted
in parallel to each other.
3.1.2 Sequential Experiments: Central Composite Design
The experimental designs for sequential experiments could be, for example, a full
factorial design or a central composite design (CCD). A full factorial experiment includes all
possible factor-level combination in the experimental design. A CCD is an alternative to the
factorial design and particularly useful for conducting experiments in a sequential manner
[Mas89], Efficiency is achieved by reducing the number of factor-level combinations from
what would be required using frill or fractional factorial design. For instance, suppose 4
important factors, each at 5 levels, are to be investigated. The total test runs for the full
factorial experiment would be 5x5x5x5 = 625, while only 30 runs is required for CCD.
In this work, the central composite design (CCD) was chosen for conducting
sequential experiments. The CCD is feasible to nm 5 levels of each factor. The procedure
for constructing CCD is given in AppendixB [Mas891. An example of the four-factor (A, B,
C, and D) CCD with five levels of each factor (-2, -1,0,1,2) is shown in Table 3.4.
Once the data were collected, graphical and statistical analysis were performed.
Graphical analysis provided a visual understanding of the importance of each effect towards
responses. A statistical method, Analysis of Variance (ANOVA), was used to provide a
quantitative measure of confidence. Empirical models relating the important factors to
responses (e.g. combustion temperature, particle size) were developed. This provides the
information required for adjusting the setting of the important factors in order to obtain the
desired products.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Table 3.4 Central Composite Design for 4 Factors
Coded Factor Levels
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
A
B
-1
-1
-1
-1
-1
-1
-1
-I
c
-I
-1
+1
+1
D
-1
+1
-1
+1
-1
-1
-1
-1
+1
+1
+1
+1
+1
+1
+1
+1
0
0
0
0
0
0
-2
+2
0
+1
+1
+1
+1
-1
-1
-1
-1
+1
+1
+1
+t
0
0
0
0
0
0
0
0
-2
-1
-1
+1
+1
-1
-1
+1
+1
-1
-1
+1
+1
0
0
0
0
0
0
0
0
0
-1
+1
-1
+1
-1
+1
-1
+1
-I
+1
-1
+1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
+2
0
0
0
0
0
-2
+2
0
0
0
0
0
-2
+2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
3.1.3 Sample Preparation
Various particle sizes of reactant powder of TiC>2 ,
C ,
Al and
A I2 O 3
were mixed in
various ratios to examine the effect of reactant particle size and the effect of diluents (excess
reactant (Al) and reaction products
(A I2 O 3 ))
on the reaction process. The reaction studied
was as follows:
m 0 2 + 3C + (4 + x)Al + yAl20 3 -> 3TiC + (2 + y)Al20 3 +xAl
(3.1)
Four compositions to be studied in the screening experiment were stoichiometric,
40% excess Al, 40% excess AI2O3, and 40% excess both Al and AI2O3 where (x,y) were
(0,0), (9.48, 0), (0, 2.51), and (28.43, 7.52), respectively. The reagent powders were mixed
and pulverized in a porcelain mortar and then sieved through a 35 mesh screen. Then the
powders were dried in an oven at 120°C for 2 hours prior to ignition and characterization.
The source and characteristics of these reactant powders are given in Table 3.5. The
particle size of the reactants were chosen based on the reactant powder mixture (i.e.
10
pm
Al, 0.7 pm rutile Ti(> 2 and -0.02pm Carbon black) commonly used elsewhere [Rab90,
Bow94, Bow96, Cho94, Cho95] as a standard. However, it should be noted that the reactants
used to synthesize AhCb-TiC are actually varied among the researchers.The particle size of
44pm for all the starting powders was studied by Feng et al [Fen92] and Kunrath et al.
[Kun96]. Petrie et al. [Pet92] used anatase TiC>2 with particle size of less than 2 pm, <20 pm
graphite, and -325 mesh Al, while Yoon et al [Yoo95] studied 1pm rutile TiC>2 , 20 pm Al,
and 5 pm graphite.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
Table 3.5 Powder Characterization
Reactants
Particle Size (pm)
Purity (%)
Source
Rutile TiOa
0.9-1.6
99.80%
Alfa Aesar Products
1.0-5.0
99.95%
Acros Organics
0.017
99.10%
Cabot Corporation
0.075
>99.00%
Cabot Corporation
99.00%
Alfa Aesar Products
99.00%
Aldrich Chemical Co., Inc.
0.49
99.80%
Alcoa
40-45
99.90%
Alfa Aesar Products
C
Al
3.0-4.5
-200 mesh
A120 3
3.1.4 Experimental Setup
The samples were ignited using two heating methods; microwave heating (MH) and
microwave hybrid heating (MHH). For MH, the 5g powder mixture was filled in a fused
quartz crucible and placed inside the insulation, porous silica refractory brick, which was
then positioned at the center of the microwave oven. While for the MHH case, the crucible
filled with powders was encased in a susceptor, which was then enclosed in the insulation.
The microwave susceptor was fabricated from a mixture of 20 wt% a-SiC (1000pm
diameter; Standard Sand and Silica Co.), 20 wt% a-SiC (85pm diameter; Norton Company),
and 60 wt% calcium aluminate cement (Alfrax66; Carborundum, Refractories Division).
The susceptor was first mixed in dry form and then deionized water was added to the
mixture until sufficient flowability occurred. The wetted mixture was then poured into a
plastic mold, where it was then dried for 24 hours. The hardened susceptor was then
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
removed from the mold and dried in an oven at 100 °C for 24 hours. The susceptor was then
ready to be used.
The microwave-combustion was carried out in a Raytheon Radarline Multimode
Microwave Oven Model QMP 2101B-6 operating at 2.45GHz. This microwave utilized up
to 8 magnetrons of 800 watts each for a maximum of 6.4 kW. Inside the microwave cavity
were 8 mode stirrers that were used to provide a more uniform microwave field.
The refractory housing was placed in the center of the microwave oven cavity under
an At atmosphere (1000 sccm/min). The samples were ignited under microwave power of
either 800 or 1600 watts operating at 100% microwave power duty cycle. An inconelshielded K-type thermocouple (Omega Engineering, Inc., XCIB-K-1-3-L) was placed inside
the fused quartz crucible filled with powder to determine ignition temperature
(T ig).
Its
working temperature range was 0-l250°C. The thermocouple was encased in a MgO
insulated fabric with an inconel metal sheath to prevent arcing of the fine thermocouple
wires. Dining the runs, the thermocouple sheath was grounded to the stainless-steel
microwave cavity to reduce the electric field intensification at the curved tip of the
thermocouple, which resulted in more accurate temperature readings.
The combustion temperature (Tc) was monitored by a two-color infrared pyrometer
(Series30 Ratio Pyrometer, Capintec, Inc.) with a working temperature range from 800 to
2300°C. Both thermocouples and radiation pyrometers have been used extensively for
temperature measurement in microwave field [Bes91, Bin93, Dan93, Fan93, Gre95, Jan91,
Jan92, Wil93, Yii95,]. Figure 3.1 illustrates the microwave heating setup for this work.
3.1.5. Sample Characterization
The powders produced were characterized using a variety of techniques.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
Temperature
controller
,asa) IR-2 color
pyrometer
Tempersjture display
n
Control
Panel
S Refractory
§ Housing
®
3
s c / ai2o 3 |
\
Susceptorg Alumina
tube
In g y lg jg r^ g
Microwave Oven Cavity
Figure 3.1. Simplified experimental setup for samples run in Radarline multimode
microwave oven.
3.1.5a X-rav diffraction analysis fXRDl
The powders produced were analyzed by x-ray diffraction to identify the chemical
compositions present. The x-ray analysis was performed in a Philips APD3720
diffractometer equipped with a Cu x-ray tube operating at 40 kV and 20 mA.
3.1.5b Pvcnometrv
A helium ultrapycnometer (Quantachrome Model 1000) was used to measure the
powder density. The pycnometers measured the true volume of solid materials by employing
Archimedes’ principle of fluid (gas) displacement. The volume of the powder is equal to the
volume of gas it displaces. The helium gas was used since it would penetrate surface flaws
down to about one Angstrom, thereby enabling the measurement of powder volumes with
great accuracy. Density was determined by calculating the mass to volume ratio.
3.1.5c BET
The surface area of the powder was determined using Brunauer-Emmett-Teller
(BET) method (Quantasorb Quanta-Chrome Model Nova 1200) with nitrogen as the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
adsorbate. As more gas molecules are introduced into the system, the adsorbate molecules
tend to form a thin layer that covers the entire adsorbent (powders) surface. Based on BET
theory, the number of molecules required to cover the adsorbent surface with a monolayer of
adsorbed molecules (Nm) can be estimated. Multiplying Nm by the cross-sectional area of an
adsorbate molecule yields the powder’s surface area. The specific surface area (the ratio of
the total surface area and the mass of the powder) can provide an estimate of the equivalent
spherical particle diameter of a powder. This relationship can be expressed as:
Sw= 6/pD
(3.6)
where Sw is the specific surface area, p is the density of a particle, and D is the equivalent
spherical diameter.
3.1 .Sd The Coulter Principle
The particle size and size distribution of the powders were determined using the
Coulter (Model LS230) method in a water medium. The LS series use laser light with a
wavelength of 750 nm to size particles by light diffraction. The laser beam passes through
the sample cell where particles suspended in liquid scatter the incident light in characteristic
patterns, which depend on their sizes.
3.1.5e Energy Dispersive X-rav spectroscopy (EDS)
Energy dispersive spectroscopy was performed on powders in order to identify the
elements within the composite powders. The analysis was performed using the JEOL JSM6400 Scanning electron microscope with an Oxford EDS analysis system.
3.1.5f Scanning electron microscopy (SEM)
Scanning electron microscopy (JEOL Model JSM-6400) was conducted on powders
to examine the particle and agglomerate morphology (size, shape and texture).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
3 2. Mechanism Studies
The key to controlling the final product configuration is to understand the
combustion reaction mechanism. The particle-particle interactions occurring during a
combustion reaction were studied in order to understand mechanisms and help to predict the
combustion behavior and resultant product configuration. In this study, four starting
materials (TiOi, C, Al, A I 2 O 3 ) were used in the reactions. To better understand die complex
interaction of the four reactants on a particle-particle level as the reaction proceeds, it is
necessary to understand how each component behaves with respect to itself and in
combination with the other reactants.
The experimental approach was to document and observe the behavior of single
components, two components, three components, and four component interactions during
and after reaction initiation. The temperatures of the reactants when subjected to microwave
energy as a function of time were recorded. The total heating time was determined by the
total time required for the completion of the reaction. The recorded data of the individual
reactant provided an insight into the heating characteristic of these reactants when subjected
to microwave energy. It also explained, to some extent, the evolution of the temperature as a
function of time of die system from the beginning of the heating to the moment just before
the ignition occurred. Along with the recorded temperature profiles of the two components,
it should be possible to identify which two components should react and which should not. It
should also indicate which combination reacts first for the same period of heating time. Xray diffraction analysis (XRD) was used to identify the chemical compositions and phase
present of the resulting powders.
The combustion synthesis reaction for mechanism studies was based on equation
(3.1):
37702 + 3C+ (4+ x)Al +yAl20 3 -> 3TiC+(2+ y)Al20 3 +xAl
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(3.1)
79
Four compositions chosen for this study were (1) x=0, y=0, (2) x=0, yAhCh, (3) xAl,
y=0, and (4) xAl, yAbOj. These represent the stoichiometric composition, composition with
the excess amount of the product phase (AI2 O3), composition with the excess amount of the
metal reactant phase (Al), and composition with excess amounts of both the metal reactant
and product phase, respectively. The other process parameters including the x and y values,
particle sizes of each component, and microwave power were selected based on findings
from the first part of this work. These results are discussed in Chapter 4. Powder samples
were heated by both heating methods (MH and MHH). Table 3.6 is the summary of the
combinations of mixtures that were heated for each composition.
Since most of the previous work focussed on SHS reactions ignited by conventional
heating, it was also of interest in this study to investigate whether or not the governing
reaction mechanism observed under the microwave heating was the same as that observed
under conventional heating. Thus far, few experimental reports have been dedicated to the
discussion of the mechanism governing the combustion reaction of A^Os-TiC in the
conventional heating [Bow94, Cho95], In 1994, Bowen et al. reported the use of differential
thermal analysis (DTA) to identify the reaction mechanism. The reactant powders used in
this work were TiC>2 (mean particle size 0.7pm), Al (10pm), and carbon black (-0,02pm).
DTA for TO 2 -C, 4A1-3C, 3TiC>2-4Al and 3TiC>2-4AI-3C systems were performed at a
heating rate of 10 degmin*1 under an argon atmosphere. Bowen et al. suggested that the
reaction proceeded in a two-stage process. First, the aluminum phase melted and reacted
with dtania to form alumina at 900°C. The reduced-titania then reacted with the carbon
atoms to produce TiC. These yielded a final product of AhCb-TiC composite. In 1995, the
same reaction mechanism of TiOi-Al-C was confirmed by Yoon Choi et al. Powders of TiC>2
(lpm), Al (20pm), and graphite (5pm) were used in this study. DTA for all possible
interactions of the reactants was performed at a heating rate of 40 deg min*1.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
80
Table 3.6 All Combinations of Mixtures for Each Composition.
Composition 1
Composition 2
Composition 3
Composition 4
3 T i0 2 +3C + 4Al
-> 3TiC + 2 AI2O 3
3TiOj + 3C + (4+x)Al
-»3TiC + 2 AI1O3 +xAl
3T i0 2 + 3 C + 4 A l + yAl203
3TiC + (2+y)Al20 3
3 T i0 2 + 3C + (4+x)Al +
yAltrCh -* 3TiC +
(2+y)AI20 3 +xAl
3Ti02
[3Ti02]*
[3Ti02]*
[3Ti02]*
3C
PC]*
[3C]*
[3C]*
4A1
(4+x)Al
[4A1]*
[(4+x)Al]**
3Ti02+3C
[3TO+3C]*
yAl20 3
[yAUOj]***
3Ti02+ 4Al
3Ti02+■(4+x)Al
[3Ti02+ 3C]*
[3Ti02+3C]*
3C + 4A1
3C + (4+x)Al
[3Ti02+ 4A1]*
[3Ti02+ (4+x)Al]**
3Ti02+3C +
4A1
3Ti02+ 3C +
(4+x)Al
3Ti02+ yAl203
[3Ti02+ yAl20 3]***
[3C + 4A1]*
[3C + (4+x)Al]*»
3C + yAl20 3
[3C + yAl20 3]***
4A1 + yAl20 3
(4+x)Al + yAl20 3
[3Ti02+ 3C + 4A1]*
[3Ti02+3C +
(4+x)Al]»*
3Ti02+ 3C + yAl2C>3
[3Ti02+ 3C +
yAl20 3]***
3Ti02 + 4A1 +
yA1203
3Ti02 + (4+x)Al +
yA1203
3C + 4A1 + yAl20 3
3C + (4+x)Al +
yAi2o 3
3Ti02+ 3C + 4A1 +
yAl20 3
3Ti02+ 3C + (4+x)Al
+ yAl20 3
Total run = 4
Total run = 4
Total run = 8
Total run = 7
[ ]*> [ ]**> [ ]*** denotes the same combination as listed in the first column (composition
1), second column (composition 2), and third column (composition 3), respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
In the present study, the mechanism governing the combustion reaction of AhOj-TiC
powder under conventional heating was investigated, using reactants in the form of powder
summarized in Table 3.6. The result obtained was then compared with those cited from the
literature, and that obtained under the influence of microwave heating. The combustion
reaction mechanism under conventional heating was explored by carrying out DTA
experiments on every combination of the reactants to determine the reactivity of one pair
relative to the others, and on the full mixture of the reactants. The DTA trace of the full
mixture helped to identify the ignition temperature of the reaction. Comparisons among the
DTA traces of the interactions for a given pair of the reactants with that of the full mixture
helped determine which two reactants were likely to initiate the SHS reaction. The
compositions studied were the same as that investigated under microwave heating (i.e.,
composition 1, 2, 3, and 4) for comparison. The possible interactions that were studied for
each composition are also summarized in Table 3.6. The powder mixtures were analyzed in
a DTA (Harrop Industries Model ST-736) at a heating rate of 10 deg min'1 within a
temperature range of 25°C to 1200°C under an argon atmosphere (the same as that used for
all combustion experiments). X-ray diffraction analysis (XRD) was also performed to
identify the chemical compositions and phase present of the resulting powders.
In order to analyze the consecutive reaction steps in either one of these compositions,
the full mixture of the reactants was heated in the DTA at a heating rate of 10 deg-min'1. The
power was turned off immediately after achieving a desired temperature. This temperature is
an immediate temperature right before or after any peak temperatures observed from the
previous DTA trace of the full mixture. The powder mixture was then cooled to room
temperature at a rate of about 30°C-min*1 under a constant flow of argon gas, and was
subsequently analyzed for phase composition using X-ray diffraction (XRD). With a
combination of the results obtained from DTA and XRD techniques, the mechanism
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
governing the combustion reaction of AhCb-TiC powder could then be identified. These
results were then compared to that obtained for the case of microwave heating to determine
if there is any difference in reaction mechanisms under die influence of different heating
sources.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
RESULTS AND DISCUSSIONS
The results of this study were divided into two sections. The first section has focused
on the results of the effects of the various reaction parameters on the ignition -combustion
behavior and characteristics of the resulting composite powders ignited under microwave
energy. The second section has focused on the mechanisms of the microwave-combustion
reaction.
4.1 Effects of Reaction Parameters
4.1.1 Screening Experiments: Plackett-Bunnan Design
4.1.1a Data collection
A 16run Plackett-Burman design was developed to analyze the effects of reaction
parameter on each response value, determine the significance of each reaction parameter,
and identify a small number of dominant factors for subsequent more extensive
investigation. The reaction parameters of interest included amount of excess A1 and AI2O3 ,
particle size of each reactant, heating method, and microwave power. The responses
considered were ignition time and temperature, combustion time and temperature and
characteristics of resulting powders.
The combustion synthesis of
A I2 O 3
- TiC powders using MH and MHH was
achieved successfully. The powder product formed in a fused quartz crucible was circular
from the center. The temperature versus time data was recorded. Table 4.1 summarizes the
results of ignition time (tig), ignition temperature (Tig), combustion time (t«) and combustion
temperature (Tc) of the sixteen-runs Plackett-Burman design.
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
Particle size of reactant (nm)
Heating
method
MW
power
(w)
tig
Tig
(sec)
(°C)
tc
(sec)
1 -1
Wl%
excess
Run
^
Table 4.1 Ignition and Combustion Behavior of the Sixteen-runs Plackett-Burman Design of
t h e A I 2 O 3 - TiC Powders
Parameter
Response*
Al
AI2O 3
TiCF
C
Al
1
0
0
I.0-5.0
0.075
•2 0 0 mesh
0
MH
1600
30
160
48
2248
2
40
40
0.9-1.6
0.075
3 -4 .5
40-45
MHH
800
630
938
660
1 1 2 2
3
40
0
1.0-5.0
0.075
3 -4 .5
0
MHH
800
400
689
425
2257
4
40
40
1.0-5.0
0.075
3 -4 .5
40-45
MH
1600
N/A
N/A
N/A
N/A
5
0
40
1.0-5.0
0.017
3 -4 .5
40-45
MH
800
30
79
40
2257
6
40
0
0.9-1.6
0.075
3 -4 .5
0
MH
1600
60
309
68
2226
7
40
40
1.0-5.0
0.017
•2 0 0 tnesh
0.49
MHH
1600
240
969
300
2215
8
40
40
0.9-1.6
0.017
-2 0 0 mesh
0.49
MH
800
N/A
N/A
N/A
N/A
9
0
0
0.9-1.6
0.075
-2 0 0 mesh
0
MHH
800
2 0 0
354
206
1827
to
0
40
1.0-5.0
0.075
-2 0 0 mesh
0.49
MHH
800
570
8 6 6
590
2054
11
0
0
0.9-1.6
0.017
3 -4 .5
0
MH
800
40
69
50
1981
12
0
0
1.0-5.0
0.017
3 -4 .5
0
MHH
1600
30
142
65
2257
13
0
40
0.9-1.6
0.075
•2 0 0 mesh
0.49
MH
1600
140
311
150
2173
14
0
40
0.9-1.6
0.017
3 -4 .5
40-45
MHH
1600
60
231
70
1796
15
40
0
1.0-5.0
0.017
- 2 0 0 mesh
0
MH
800
50
81
57
2257
16
40
0
0.9-1.6
0.017
-2 0 0 mesh
0
MHH
1600
60
302
70
2223
A1A
* The tig, Tig, tc, and Tc represents ignition time, ignition temperature, combustion time, and
combustion temperature, respectively. N/A means that samples were not ignited.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
The effect of composition on ignition behavior is shown in Figure 4.1 and 4.2. It was
observed that ignition time increased as die amounts of the diluents increased. This was due
to the decreased heat generation rate. In the case of 40% excess of both Al and
A I2 O 3 ,
samples could not be ignited by microwave energy alone but were ignited by using the
MHH technique. For all compositions, approximately 0.30 to 2 minutes were required for
800w-MH processed samples to reach ignition temperatures while 800W-MHH processed
samples required longer times, typically 3.20 to 10.30 minutes. This is because of strong
microwave energy absorption at 2.45GHz by the susceptor, which resulted in less efficient
heating of the sample. By increasing the microwave power from 800W to 1600W in the
MHH technique, however, the ignition tone could be decreased to the same level as those
required for 800w-MH ignited samples.
The x-ray diffraction patterns of synthesized powders produced by 800W-MHH for
various compositions are illustrated in Figure 4.3. XRD of stoichiometric and 40%excess
A I2 O 3
samples detected only TiC and
excess both
A I2 O 3
A I2 O 3 .
In cases of 40%excess Al sample and 40%
and Al sample, x-ray patterns show not only TiC and A I 2 O 3 but also an
Al signal. As expected, the proportion of the reaction product changed with begining
stoichiometry. An increase of A I 2 O 3 incorporation resulted in an increase of alumina peak
intensity in the final product. Excess Al in the beginning mixture resulted in a higher weight
fraction in the reaction products of Al and smaller amounts of A I 2 O 3 and TiC.
The characteristics of synthesized AI2 O3 - TiC powders are given in Table 4.2.
Pycnometry analysis showed that the densities for all powders were in the range of 3.354.25 g/cm3. The specific surface areas and the particle sizes were in the range of 0.6-4.08
m2/g and 0.37-2.4 microns, respectively. The particle size distribution of powders is given in
Appendix C.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
86
8M-MH
800-MHH
2500
Q 2000
2000
g
1500
8. iooo
H
500
0
200l
400
TkDC(KC)-H
0
ItaM(MC)
1600-MH
2500
1 tb
£
a
a
£
1
Q
T
v
^
c
2000
g 1500
1
—
1000
A
s.
'
•
•
1
0
•---•
1
200
Thnc(iec)
. Stoichiometric
. 40% acesiAI.
■ ■
*
1 ■i
(00
0
. 40%exces«A12O3
400
200
Tknc(Kc)
600
40%cxcaiAl:40%excmAI203
Figure 4.1. Effect of composition on temperature profile of die AI2 O3 - TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
2500
Stoichiometric
2000
8a 1500
e
& 1000
••••
E
500
0
0
100
200
Time (sec)
2500
Q 2000
C
£ 1500
2
% 1000
E
£ 500
40% excess Al snd
40% excess AljO>
.
j
|
1
t
t
<
!
/ • • *
i f l 2r : 1
1
500
1000
Time (sec)
1600W, MHH
300
800W, MHH —
1500
1600W, MH -*-800W ,M H
Figure 4.2. Effect of heating method and microwave power on temperature profile of the
stoichiometric and 40% excess both Al and AI2O3 powders.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
88
1 * AljOj 2>TK
I
3 a Al
40*1
l I M 1 aI**; t L ' L i ■ A A A . t i . i i f i ; ? >
h
3
^
*
'40%bCxcaH
2 1 2
£
«
)l—
jl b L i f t j L
2
40«* CxcaM
l
I i l 2l
C _ J l U
10
2
0
3
0
1
J L L J a _ y v J ^ ^
>
!ii----I
I~
*1 * 1 3
j L _ J U
2
I
_«!
■»
4
■■■■—— ,
0
5
,
0
8
0
,
7
0
I
8
0
I
9
0
2Th«ta [dag]
Figure 4.3. X-ray diffraction of the AI2 O3 - TiC powders ignited by 800watts MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
Wt%
excess
r
0.99
1.46
55.14
3.45
3.35
0.52
34.58
800
3.61
1.15
1.44
66.44
MH
1600
N/A
N/A
N/A
N/A
40-45
MH
800
3.93
4.08
0.37
55.14
3 -4 .5
0
MH
1600
3.45
0.99
1.76
105.9
0.017
•200mesh
0.49
MHH
1600
3.39
2.73
0.65
41.67
0.9-1.6
0.017
-200mesh
0.49
MH
800
N/A
N/A
N/A
N/A
0
0.9-1.6
0.075
-200mesh
0
MHH
800
4.14
0.61
2.39
60.52
0
40
1.0-5.0
0.075
-200mesh
0.49
MHH
800
3.94
0.70
2.18
60.52
11
0
0
0.9-1.6
0.017
3 -4 .5
0
MH
800
4.25
0.95
1.48
60.52
12
0
0
1.0-5.0
0.017
3 -4 .5
0
MHH
1600
4.18
0.75
1.92
66.44
13
0
40
0.9-1.6
0.075
-200mesh
0.49
MH
1600
3.63
1.34
1.23
60.52
14
0
40
0.9-1.6
0.017
3 -4 .5
40-45
MHH
1600
3.95
4.07
0.37
45.75
15
40
0
1.0-5.0
0.017
•200mesh
0
MH
800
3.35
IJ33
1.34
66.44
16
40
0
0.9-1.6
0.017
-200mesh
0
MHH
1600
3.39
1.23
1.43
66.44
Heating
method
MW
power
(w)
C
Al
AljOj
1.0-5.0
0.075
•200tnesh
0
MH
1600
4.13
40
0.9-1.6
0.075
3 -4 .5
40-45
MHH
800
40
0
1.0-5.0
0.075
3 -4 .5
0
MHH
4
40
40
1.0-5.0
0.075
3 -4 .5
40-45
5
0
40
1.0-5.0
0.017
3 -4 .5
6
40
0
0.9-1.6
0.075
7
40
40
1.0-5.0
8
40
40
9
0
10
AljOj
1
0
0
2
40
3
m
I
Al
8
1
Run
Density (g/cc)
■SI
Agglomeration
Size (pm)
Particle size of reactant (pm)
Specific Surface
Area (tn2/g)
Table 4.2 Characteristics of the Sixteen-runs Plackett-Burman Design of the AI2O3 - TiC
Powders
Parameter
Response*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
The corresponding microstructures of these composite powders are shown in Figure
4.4 - 4.7. Results of the agglomeration size obtained from the Coulter method were
confirmed, and more details of the irregular morphology are illustrated. Stoichiometric
powders were fragmented and angular shaped, while others, especially 40%excess of either
AI2 O3 or Al particles, are more rounded.
4.1.1b Data analysis
After all the response data including ignition time and temperature, combustion time
and temperature, and also characteristics of powders were obtained, the effects of the
reaction parameters on each response value were analyzed using equation (3.2). The results
are presented in Table 4.3.
From the table, the plus sign in front of the number means that the addition of that
particular parameter increases the corresponding response value, while the minus sign
means that the response value is decreased. For example, by increasing the amount of Al and
A I2 O 3 ,
the ignition time of the sample increases while combustion temperature decreases. In
addition, the higher the absolute value, the greater the effect of that particular parameter on
the investigated response value.
The relationship between die effect of the parameters and the investigated responses
were then determined using the following linear model of the Plackett*Burman design:
y=B,
+
+• ••+s . jr.)
(4,i)
where y is the investigated response, B0 is the average of all the observations, Bi is the effect
of parameters, and Xi is the value of the parameters. Equation (4.2H4.9) represented the
prediction equations of the effects of the reaction parameters on each investigated response
value.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
800W MH
1600W MH
800W MHH
1600W MHH
Figure 4.4. Scanning electron micrograph of the stiochiometric 2 Ali0 3 -3 TiC powder
ignited by various heating method. Note: Scale shown in the left and right column are
100 pm and 10 pm, respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
800W MH
1600W
MH
800W MHH
1600W
MHH
Figure 4.5. Scanning electron micrograph of the 2 AI1O3 - 3TiC - 40% excess AI2 O3
powder ignited by various heating method. Note: Scale shown in the left and right
column are 1 0 0 pm and 1 0 pm, respectively.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
93
800W MH
1600W MH
800W MHH
1600W MHH
Figure 4.6. Scanning electron micrograph of the 2 AI2O3 - 3TiC - 40% excess Al powder
ignited by various heating method. Note: Scale shown in the left and right column are
1 0 0 pm and 1 0 pm, respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
800W MH
1600W MH
800W MHH
1600W MHH
Figure 4.7. Scanning electron micrograph of the 2 AI2 O3 - 3TiC - 40% excess both AI2 O 3
and AI powder ignited by various heating method. Note: Scale shown in the left and right
column are 1 0 0 pm and 1 0 pm, respectively.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
95
Table 4.3 The Effects of the Reaction Parameters on Each Investigated Response Value
Effect of Parameter
Response
W t% of
Excess
Al
Wt% of
Excess
AI2O3
Ti0 2
Particle
Size
C
Particle
Size
Al
Particle
Size
AI2O3
Particle
Size
Ignition Time
+42.50
+100.00
+20.00
+190.00
+5.00
Ignition
Temperature
+134.55
+161.00
+59.00
+219.23
Combustion
Time
+45.13
+102.63
+31.38
Combustion
Temperature
-536.63
-707.38
Powder
Density
-1.44
Particle Size
Heating
Method
MW
Power
-57.50
+230.00
-162.50
+73.30
-224.48
+435.20
-81.55
+186.88
+5.38
-67.50
+246.63
-157.13
+274.63
-134.88
+137.63
-316.75
+326.13
+172.88
-1.03
+0.04
-0.01
-0.11
+0.10
+0.91
-0.07
-0.53
-0.99
+0.02
+0.42
+0.35
-0.70
+0.41
-0.11
Agglomeration
Size
-10.40
-31.21
-2.81
+5.15
-2.94
-6.81
+4.84
+4.71
Specific
Surface Area
-0.34
+1.03
-0.10
-0.75
-0.80
+1.68
+0.61
-0.01
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
96
( 42.5Al% +\0toAl2O3%
+ 20TiO2particlesize+190Cparticlesize
IgnitionTime = 158.75 + -^
+ 5AlParticleSize - 57.5Al20 3ParticleSize
^+ 230Heatingmethod -162.5MWpower
(4.2)
f 134.55/4/%+16L4/j03%
>
+ 59. \Ti02particlesize+219.23Cparticlesize
IgnitionTemp ~ 343.76+-^
(4.3)
+ 73.3AlParticleSize - 224.475Al20 3ParticleSize
+435.2Heatingmethod-&\.55MWpower
,
'45.13,4/%+102.63Al20 3%
N
+ 31.3&Ti02particlesize + 186.88Cparticlesize
(4.4)
Combustion Time = 174.94+—
2 + 5.3%AlParticleSize - 67,5Al20 3ParticleSize
[_+246.63Heatingmethod -157.13A/Jf^ower ,
(-536.63Al%-707.3&Al2O3%
+ 274.63Ti02particlesize-\34.SSCparticlesize
(4.5)
CombustionTemp ~ 1805.80+y
+137.63AlParticleSize - 316.75Al20 3ParticleSize
+326.13Heatingmethod +172.%8MWpower
(-1.44i4/%-l.O3i4/,0j%
+ 0.147702particlesize -0.0 XCparticlesize
PowderDensity = 3.30+
-0.11 AlParticleSize+ 0A0Al2O3ParticleSize
+ 0.9\Heatingmethod - 0.07MWpower
(4.6)
r-0.53^/% -0.99^/20 3%
ParticleSize = 1.16 +
42
-
+ 0.02TiO2particlesize+ 0A2Cparticlesize
+0.35AlParticleSize-0.70Al20 3ParticleSize
[_+0.4\Heatingmethod - 0.1 \MWpower
(4.7)
'-10.40/4/% -31.2 U /20 3%
^
i
- 2.8 YTi02particlesize+ S.ISCparticlesize
(4.8)
AgglomerationSize = 52.88+i
- 2:94AlParticleSize - 6.8\Al20 3ParticleSize
l^+4MHeatingmethod+4.7 IMWpower
J
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
97
r-0 M A l% + \m A l2O,%
1 - 0.lQTiO2particlesize - O.ISCparticlesize
SpecificSurfaceArea = 1.52 +
(4.9)
2 - 0.80AlParticleSize+1.68AI20 3ParticleSize
Q.6\Heatingmethod -0.0 IAIWpower
J
For example, from equation (4.6), it can be seen that by increasing amount of excess
Al and/or excess AI2 O3, the density of the resulting powder decreased. Equation (4.2) and
(4.4) shows that combustion time and ignition time are both increased by increasing the
amount of excess Al and/or AI2O3 , increasing particle size of carbon, decreasing particle size
of alumina, and decreasing microwave power. Particle size of the powder decreased with the
amount of excess Al and/or AI2O3 and particle size of AI2 O3 , but increased with the
increased particle size of carbon and aluminum (equation 4.7).It should be noted that,
however, these relationships are only preliminary results. The actual relationships were
studied in more depth and their results are presented in the following section (central
composite design).
The significance of each reaction parameter on all the investigated responses was
then determined by using equation (3.2>(3.5). The higher the significance level, the greater
the certainty that the parameter is an important factor contributing toward the investigated
response value. Significance values of the reaction parameters are presented in Table 4.4. A
dash represents a significance value below 80%, and thus the level of the experimental error
prohibited a differentiation. Thus the factor could not be determined to be significant.
From Table 4.4, the significance of reaction parameters varied from one response to
another. It was clear that the density of the composite powder depended on the amount of
excess Al and AI2 O3 as much as heating method. Almost every parameter, except particle
size of TiC>2 and microwave power, was significant toward particle size of composite
powder. Agglomeration size and surface area of the powders tended to depend on amount of
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Table 4.4 Relative Significance of Reaction Parameters of AI2O 3 -TiC Powders
Significance o f Parameter
Response
Wt% of
Excess
Al
Wt%
of
Excess
AI2O 3
TiOj
Particle
Size
80.0%
90.0%95.0%
-
80.0%90.0%
80.0%90.0%
Combustion
Time
80.0%
Combustion
Temperature
Al
Particle
Size
AI2O3
Particle
Size
Heating
Method
MW
Power
9999.5%
-
80-90%
99.599.95%
97.599%
-
90.0%95.0%
-
90.0%95.0%
99.5%99.95%
-
90.0%95.0%
-
99.0%99.5%
-
80.0%90.0%
99.5%99.95%
97.5%99%
80.0%90.0%
90.0%95.0%
-
-
-
80.0%
80.0%
-
Powder
Density
95.0%97.5%
90.0%95.0%
-
-
-
-
90.0%95.0%
-
Particle Size
97.5%99.0%
99.5%99.95%
-
95.0%97.5%
90.0%95.0%
99.5%99.95%
95.0%97.5%
-
80.0%
97.5%99.0%
-
-
-
-
-
90.0%95.0%
-
80.0%90.0%
97.5%99.0%
80.0%90.0%
-
Ignition Time
Ignition
Temperature
Agglomeration
Size
Specific
Surface Area
C
Particle
Size
-
80.0%90.0%
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
99
diluents and particle size of A I 2 O 3 more than any other factors. In addition to amount and
particle size of A I 2 O 3 , the particle size of C, heating method, as well as microwave power
showed a significant role on the ignition behavior of the powders. It was found that heating
method was significant in almost every response, while microwave power had a significant
effect on only ignition and combustion time. As shown in equation (4.2), the time required
prior to ignition increased as the ignition power decreased.
When the significance of all parameters on every response was considered together,
the important reaction parameters for subsequent extensive investigation were identified.
Those parameters included amount of excess Al and A I 2 O 3 , particle size of C and A I 2 O 3 , and
also heating method.
In this research, the qualitative factor, heating method, is of interest. Therefore, two
experiments for each level of this factor (MH and MHH) were then conducted in parallel to
each other.
4.1.2 Sequential Experiments ; Central Composite Design
4.1,2a Factor settings and design matrix
The central composite design (CCD) was chosen because it reduces considerably the
number of experimentation points that would otherwise be required if the full factorial
design were used. Based on the results of the sixteen-run Plackett-Burman design, the most
significant factors affecting the investigated responses were amount of excess Al and AI2 O3 ,
particle size of C and AI2O3 and also heating method (MH and MHH). Two experimental
sets for MH and MHH were conducted m parallel. Thus there were only four parameters to
be investigated in each set of experiment. The number of experimental runs required for four
parameters was 30, as shown previously in Table 3.4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
Each reaction parameter was evaluated at five different levels. The five levels for
each of the four parameters are shown in Table 4.5. Other (less significant) factor levels
were set at fixed levels based on the literature and economics. The particle size of Ti0 2 and
Al were chosen at 2-3 pm (99% ruble, Alfa Aesar) and 10 pm (99.7%, Alfa Aesar),
respectively. The samples were ignited under microwave power of 800 watts operating at
100
% duty cycle.
The actual factor levels for a thirty-run CCD design for the microwave-induced
combustion synthesis of Ah0 3 -TiC powders study are shown in Table 4.6. This design was
used for both experimental sets, MH and MHH. As seen from Table 4.6, there were nine
compositions to be studied in the CCD experiment. These compositions are summarized in
Table 4.7.
4.1.2b Data collection
To minimize the possibility of bias effects, the experiment test sequence and the
assignment of the factors to the columns were randomized. The data on the ignition time,
ignition temperature, combustion time, combustion temperature, and also characteristics of
powders (particle size, surface area, and density) were collected.
Table 4.8 and 4.9 summarizes the results of ignition time (tig), ignition temperature
(Tig), combustion time (tc) and combustion temperature (Tc) of the thirty-run CCD design for
MH and MHH experiment, respectively. A clear difference in ignition and combustion time
was observed for these two ignition methods. It was found that approximately 7sec to
1,50min were required for MH ignited samples to reach ignition temperatures while MHH
ignited samples required longer times, 20sec to
6
min. The MHH samples ignited at higher
temperature than MH samples and resulted in lower combustion temperature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Table 4.5 Factors and Investigated Levels of Thirty-Runs CCD
Level
Factor
C Particle
Size
-2
-I
0.017 11m
0.027 pm
(99.1%)
(>99%)
0
1
2
0.037 pm
0.042 pm
0.075 pm
(>99%)
(99.9%)
(>99%)
(Alfa Aesar)
(Cabot
Corporation)
20-38 pm
44-53 pm
63-74 pm
(Cabot
(Cabot
(Cabot
Corporation) Corporation). Corporation)
A I2 O 3
pm
0.49 pm
<10
(99.8%)
(99.7%)
(99.9%)
(99.9%)
(>99%)
(Alcoa)
(Aldrich
Chemical
Co., Inc.
(Alfa Aesar)
(Alfa Aesar)
(Fisher)
Particle Size
Amount of
excess Al
0
%
10
%
20
%
30%
40%
Amount of
excess
0
%
10
%
20
%
30%
40%
A I2 O 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.6 Actual Factor Levels for a Thirty-Run Central Composite Design for the
Microwave-Induced Combustion Synthesis of AlaCVTiC Powders Study_______
Factor Levels
Run
C Particle Size (tun)
A120 3 Particle Size (ym)
Wt% Excess Al
Wt% Excess AI2O 3
1
0.027
<10
10
10
2
0.027
<10
10
30
3
0.027
<10
30
10
4
0.027
<10
30
30
5
0.027
44-53
10
10
6
0.027
44-53
10
30
7
0.027
44-53
30
10
8
0.027
44-53
30
30
9
0.042
<10
10
10
10
0.042
< 10
10
30
11
0.042
<10
30
10
12
0.042
<10
30
30
13
0.042
44-53
10
10
14
0.042
44-53
10
30
15
0.042
44-53
30
10
16
0.042
44-53
30
30
17
0.037
20-38
20
20
18
0.037
20-38
20
20
19
0.037
20-38
20
20
20
0.037
20-38
20
20
21
0.037
20-38
20
20
22
0.037
20-38
20
20
23
0.017
20-38
20
20
24
0.075
20-38
20
20
25
0.037
0.49
20
20
26
0.037
63-74
20
20
27
0.037
20-38
0
20
28
0.037
20-38
40
20
29
0.037
20-38
20
0
30
0.037
20-38
20
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
Table 4.7 Compositions Studied in a Thirty-Run Central Composite Design___________
System
m 02+ 3C + (4 + x)Al + y A ip ,
-> 377C + (2 + y)AlzOi + xAl
Theoretical Composition of
Product
Reactant
Composition
Wt%
Excess
Al
Wt%
Excess
A120 3
X
y
Wt%
3TiC
Wt%
(2+y)Al20 3
Wt%
xAl
1
10
10
1.77
0.47
37
53
10
2
10
30
2.37
1.88
28
62
10
3
30
10
7.11
0.63
28
42
30
4
30
30
10.66
2.82
19
51
30
5
20
20
4.74
1.25
28
52
20
6
0
20
0.00
0.94
37
63
0
7
40
20
14.22
1.88
19
41
40
8
20
0
3.55
0.00
37
43
20
9
20
40
7.11
3.76
19
61
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
Table 4.8 Ignition and Combustion Behavior of the Thirty-Run CCD Design for MH
Ignited AkCh-TiC Powders_________________ ___________________________
Factor
Run
Response
C Size
AliOj Size
Wt%
excess Al
Wt% excess
AI2O 3
ti( (sec)
Ti.C’C)
to (sec)
T 0 (°C)
1
0.027
<10
10
10
10
130
17
1904
2
0.027
<10
10
30
50
175
61
2201
3
0.027
<10
30
10
25
228
32
1925
4
0.027
<10
30
30
110
284
117
1905
5
0.027
44-53
10
10
20
107
29
1933
6
0.027
44-53
10
30
30
132
34
2143
7
0.027
44-53
30
10
35
144
40
2168
8
0.027
44-53
30
30
80
171
85
1792
9
0.042
<10
10
10
7
124
8
2005
10
0.042
<10
10
30
35
162
39
2120
11
0.042
<10
30
10
20
148
30
2152
12
0.042
<10
30
30
70
194
72
1862
13
0.042
44-53
10
10
10
56
15
2026
14
0.042
44-53
10
30
20
107
23
2056
15
0.042
44-53
30
10
30
113
36
2236
16
0.042
44-53
30
30
48
152
58
1701
17
0.037
20-38
20
20
50
147
59
1908
18
0.037
20-38
20
20
47
150
56
1936
19
0.037
20-38
20
20
49
141
58
1953
20
0.037
20-38
20
20
50
149
59
1925
21
0.037
20-38
20
20
49
145
59
1941
22
0.037
20-38
20
20
48
142
56
1912
23
0.017
20-38
20
20
80
177
84
2040
24
0.075
20-38
20
20
70
155
79
2225
25
0.037
0 .4 9
20
20
70
169
73
2154
26
0.037
63-74
20
20
40
133
49
2207
27
0.037
20-38
0
20
30
120
38
1988
28
0.037
20-38
40
20
70
260
84
1858
29
0.037
20-38
20
0
30
116
34
2113
30
0.037
20-38
20
40
80
537
110
1661
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Table 4.9 Ignition and Combustion Behavior of the Thirty-Run CCD design for MHH
Ignited AbOrTiC Powders_________________ ____________________________
Factor
Run
Response
C Size
A120 ] Size
Wt%
excess Al
Wt% excess
Ala0 3
ti, (sec)
Ti,(°C)
to (sec)
Tc (°C)
1
0.027
<10
10
10
70
120
79
1694
2
0.027
<10
10
30
160
323
162
2039
3
0.027
<10
30
10
150
207
157
1779
4
0.027
<10
30
30
3 60
585
362
1690
5
0.027
44-53
10
10
60
107
63
1735
6
0.027
44-53
10
30
100
195
103
1879
7
0.027
44-53
30
10
80
156
89
1857
8
0.027
44-53
30
30
270
506
279
1480
9
0.042
<10
10
10
30
161
36
1765
10
0.042
<10
10
30
70
218
77
1943
11
0.042
<10
30
10
90
280
92
1964
12
0.042
<10
30
30
160
339
166
1577
13
0.042
44-53
10
to
20
153
13
1902
14
0.042
44-53
10
30
60
159
72
1793
15
0.042
44-53
30
10
70
187
76
2071
16
0.042
44-53
30
30
150
333
157
1478
17
0.037
20-38
20
20
160
380
163
1730
18
0.037
20-38
20
20
157
385
161
1724
19
0.037
20-38
20
20
161
389
164
1726
20
0.037
20-38
20
20
162
382
165
1729
21
0.037
20-38
20
20
158
379
160
1727
22
0.037
20-38
20
20
161
384
165
1721
23
0.017
20-38
20
20
230
443
236
1937
24
0.075
20-38
20
20
220
450
224
1700
25
0.037
0 .4 9
20
20
210
428
214
1947
26
0.037
63-74
20
20
no
244
116
2090
27
0.037
20-38
0
20
90
301
100
1830
28
0.037
20-38
40
20
260
460
261
1587
29
0.037
20-38
20
0
120
309
130
1893
30
0.037
20-38
20
40
320
603
329
1633
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
The difference in ignition time and temperature observed for MH and MHH methods
is explained by the fact that the susceptor used for the MHH experiments exhibited strong
microwave energy absorption at 2.4S GHz, which partially shields the sample from the
microwave field. This resulted in less efficient heating of the sample. This difference can
also be illustrated graphically in Figure 4.8.
The confirmation of the product phases was made by X-ray diffraction analysis.
XRD analysis of MH ignited A I 2 O 3 - TiC powder with various compositions is provided in
Figure 4.9. Figure 4.10 presents a comparison of the x-ray diffraction pattern between MH
and MHH ignited powders.
The characteristics of synthesized AliCh-TiC powders of the thirtieth-run CCD
design for MH and MHH experiments are shown in Table 4.10 and 4.11. The results showed
no significant difference in the properties of powders ignited by these two methods.
Pycnometry analysis showed that the densities for MH and MHH ignited powders were in
the range of 3.37-4.10 g/cm3 and 3.4-4.20 g/cm3, respectively. The specific surface areas and
the particle sizes for MH ignited powders were in the range of 0.77-3.05 m2/g and 0.55-2.11
microns, and were 0.90-3.76 m2/g and 0.45-1.77 microns for MHH samples. The particle
size distribution of powders are given in Appendix C.
The corresponding microstructures of these MH and MHH ignited powders are given
in Figure 4.11-4.18. SEM micrographs of ignited powders with three different carbon and
AI2O3 particle sizes can be seen in Figure 4.11-4.12 and 4.13-4.14, respectively. Figure 4.154.16 and 4.17-4.18 show SEM micrographs of the MH and MHH ignited powders with
various amount of Al and AI2 O3 , respectively. The whiskers observed in 2 AI2 O3 - 3TiC 20wt% excess Al - 40wt% excess AI2O3 sample were found to be compounds of Al and Ti
by X-ray diffraction analysis and Energy dispersive spectroscopy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
2000
O 1500
s
a
MH
2o 1000
MHH
a
E
*
0
0
50
100
150
200
250
Time (sec)
Figure 4.8 Temperature profile of 2 AI2O3 - 3TiC - 20% excess Al -20% excess AI2O3
powders ignited by microwave heating (MH) and microwave hybrid heating (MHH).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
l» A l,O j
30% Excess
A1& 30%
Excess ALO,
30% Excess
AI& 10%
Excess AUOj
s
I
10% Excess
A1 & 10%
Excess AM)
10
3 = A1
Li
.• I
j
1I
JL
2
10% Excess
Al&30%
Excess Al^O,
2 = TiC
jL
12
13 111
1
III i. Hii
,13
13 111
[
2l3
30
50
2 Thsta [dag]
70
I 3 1 1 1
BO
Figure 4.9 X-ray diffraction patterns for the MH ignited AhCb-TiC powders: (a) 10%
excess Al, 10% excess AI2 O3 , (b) 10% excess Al, 30% excess AI2O3 , (c) 30% excess Al,
10% excess AI2 O3, and (d) 30% excess Al, 30% excess AI2 O3 .
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
109
(a)MH
20H
A^O,
10
30
50
70
90
2TM * Idas]
(b)MHH
2-TiC
i- A J ,0 ,
3-
Al
2
20%Exccu
A i& m
a
Cxccn ALO, I
lij
1 1
ll
1
\\
t i l
i
l l
iti
1
:
r . i'i
i
id l i.iL
i . . r
1
Q
3
Q
IS
3
Q
4
0
S
V
1
i i
i . "
. i i ’i . i '
7
D
B
H
IIM a lla a ]
Figure 4.10 X-ray diffraction patterns of the A^Ch-TiC powders ignited by: (a) microwave
heating (MH), (b) microwave hybrid heating (MHH).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
110
Table 4.10 Characteristics of the Thirty-Run CCD design for MH Ignited AIiCb-TiC
Powders
Response
A1;0} Size
W«%
excess Al
Wt% excess
AI2O 3
Particle
Size (pm)
Agglomer
ation Size
(pm)
1
0.027
<10
10
10
3.80
1.37
1.15
55.14
2
0.027
<10
10
30
3.76
1.49
1.07
41.67
3
0.027
<10
30
10
3.51
1.50
1.14
50.23
4
0.027
<10
30
30
3.49
1.75
0.98
34.58
5
0.027
44-53
10
10
3.94
1.18
1.29
60.52
6
0.027
44-53
10
30
3.89
1.78
0.87
55.14
7
0.027
44-53
30
10
3.58
2.55
0 .6 6
60.52
8
0.027
44-53
30
30
3.50
2 .6 6
0.64
55.14
9
0.042
<10
10
10
3.89
1.34
1.15
55.14
10
0.042
<10
10
30
3.84
1.59
0.98
55.14
11
0.042
<10
30
10
3.51
1.83
0.93
60.52
12
0.042
<10
30
30
3.50
2.38
0.72
50.23
13
0.042
44-53
10
10
3.92
0.99
1.54
66.44
14
0.042
44-53
10
30
3.84
1.16
1.35
60.52
15
0.042
44-53
30
10
3.59
2.07
0.81
80.08
16
0.042
44-53
30
30
3.58
2.93
0.57
55.14
17
0.037
20-38
20
20
3.74
1.24
1.29
60.52
18
0.037
20-38
20
20
3.76
1.13
1.41
55.14
19
0.037
20-38
20
20
3.75
1.26
1.27
50.23
20
0.037
20-38
20
20
3.73
1.25
1.29
50.23
21
0.037
20-38
20
20
3.77
1.04
1.53
50.23
22
0.037
20-38
20
20
3.78
1.03
1.54
50.23
23
0.017
20-38
20
20
3.59
3.05
0.55
37.96
24
0.075
20-38
20
20
3.73
0.99
1.62
60.52
25
0.037
0.49
20
20
3.70
0.77
2.11
31.50
26
0.037
63-74
20
20
3.61
0.95
1.75
66.44
27
0.037
20-38
0
20
4.09
0.99
1.48
37.96
28
0.037
20-38
40
20
3.37
2.25
0.79
37.96
29
0.037
20-38
20
0
3.75
0.99
1.62
66.44
30
0.037
20-38
20
40
3.68
1.87
0.87
37.96
Specific
Surface
Area
fm 2/a)
C Size
Density
(g/ec)
Run
Factor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
Table 4.11 Characteristics of the Thirty-Run CCD design for MHH Ignited AfcC^-TiC
Powders
_____
Factor
C Size
AliOj Size
Wl%
excess Al
Wt% excess
AI2O 3
Specific
Surface
Area
Particle
Size (pm)
Agglomer
ation Size
(tun)
1
0.027
<10
10
10
3.96
1.24
1.22
50.23
10
30
3.87
1.47
1.06
37.96
3.61
1.51
1.10
55.14
2
0.027
<10
Density
(g/cc)
Run
Response
3
0.027
<10
30
10
4
0.027
<10
30
30
3.56
1.86
0.91
34.58
1.44
5
0.027
44-53
10
10
3.91
1.07
55.14
30
3.90
1.46
1.05
55.14
10
3.65
1.71
0.96
60.52
0.61
55.14
6
0.027
44-53
10
7
0.027
44-53
30
8
0.027
44-53
30
30
3.58
2.77
9
0.042
<10
10
10
3.94
1.56
0.98
60.52
10
0.042
<10
10
30
3.83
2.06
0.76
55.14
11
0.042
<10
30
10
3.54
1.86
0.91
66.44
12
0.042
<10
30
30
3.52
2.34
0.73
37.96
1.24
66.44
13
0.042
44-53
10
10
3.90
1.24
14
0.042
44-53
10
30
3.84
1.25
1.25
55.14
15
0.042
44-53
30
10
3.60
1.84
0.91
87.90
16
0.042
44-53
30
30
3.53
3.76
0.45
55.14
17
0.037
20-38
20
20
3.82
0.95
1.65
55.14
18
0.037
20-38
20
20
3.81
1.01
1.56
55.14
19
0.037
20-38
20
20
3.79
0.99
1.60
55.14
20
0.037
20-38
20
20
3.81
1.20
1.31
55.14
21
0.037
20-38
20
20
3.79
1.04
1.52
50.23
20
3.80
1.05
1.45
55.14
22
0.037
20-38
20
23
0.017
20-38
20
20
3.91
0.95
1.62
37.96
24
0.075
20-38
20
20
3.76
2.42
0.66
60.52
25
0.037
0.49
20
20
3.82
0.89
1:77
34.58
26
0.037
63-74
20
20
3.69
1.17
1.39
60.52
27
0.037
20-38
0
20
4.20
1.25
1.14
37.96
28
0.037
20-38
40
20
3.40
2.74
0.64
37.96
29
0.037
20-38
20
0
3.91
0.93
1.65
60.52
30
0.037
20-38
20
40
3.78
2.15
0.74
37.96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
MH
MHH
Figure 4.11 Scanning electron micrograph of the MH and MHH ignited 2AIi0 3 -3TiC 20% excess Al - 20% excess AI2 O3 powders with varying particle size of carbon: (a, b)
0.017 pm, (c, d) 0.037 pm, and (e, 0 0.075 pm. (bar = 100 pm).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
113
MH
MHH
Figure 4.12 Scanning electron micrograph of the MH and MHH ignited 2 AI2 O3 -3TiC 20%excess Al - 20% excess A I 2 O 3 powders with varying particle size of carbon: (a, b)
0.017 pm, (c, d) 0.037 pm, and (e, f) 0.075 pm. (bar= 10 pm).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
MH
MHH
Figure 4.13 Scanning electron micrograph of the MH and MHH ignited 2 AI1O3 - 3TiC 20%excess Al - 20% excess A I 2 O 3 powders with varying particle size of alumina: (a, b)
0.49 pm, (c, d) 20-38 pm, and (e, 0 63-74 pm. (bar = 100 pm).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
115
MH
MHH
Figure 4.14 Scanning electron micrograph of the MH and MHH ignited 2 AI2 O3 -3TiC 20%excess Al - 20% excess A I 2 O 3 powders with varying particle size of alumina: (a, b)
0.49 pm, (c, d) 20-38 pm, and (e, f) 63-74 pm. (bar = 10 pm).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
MH
MHH
Figure 4.15 Scanning electron micrograph of the MH and MHH ignited powders of
2 AI2 O3 -3TiC -20% excess A I 2 O 3 with varying amount of excess Al: (a, b) 0% (c, d)
20%, and (e, 0 40% (bar = 100 pm).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
MH
MHH
Figure 4.16 Scanning electron micrograph of the MH and MHH ignited powders of
AI1 O3 -3TiC -20% excess AI2 O3 with varying amount of excess Al: (a, b) 0% (c, d)
20%, and (e, 0 40% (bar = 10 pm).
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
MH
MHH
Figure 4.17 Scanning electron micrograph of the MH and MHH ignited powders of
2 AI2 O3 -3TiC -20% excess Al with varying amount of excess AI2 O3 : (a, b) 0% (c, d)
20%, and (e, 0 40% (bar = 100 pm).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.18 Scanning electron micrograph of the MH and MHH ignited powders of
2 AI: 0 3 -3TiC -20% excess Al with varying amount of excess AI2 O3 : (a, b) 0% (c, d)
20%, and (e, 0 40% (bar = 10 pm).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
120
The Energy dispersive spectroscopy (EDS) of 2 AI2O3 - 3TiC - 20wt% excess Al 40wt% excess
A I2 O 3
are shown in Figure 4.19. The analysis showed the presence of carbon,
oxygen, aluminum and titanium. The backscattered electron image (BSE) with EDS spot
analysis for the light (area 1) and dark (area 2) area is shown in Figure 4.20. Both light and
dark area have high oxygen and aluminum content. The composition for light and dark area
are 15.86%C, 40.09%0, 31.29%A1, l2.41%Ti, 0.35%Ar, and 16.69%C, 52.22%0,
26.25%A1,4.7%Ti, 0.15%Ar, respectively. The experimental error is approximately ±0.5%.
4.1,2c Data analysis
After all the response data were obtained, the JMP Start Statistics software was used
to perform a multiple regression analysis for each of the responses measured. The regression
variables were the four reaction parameters including particle size of carbon and alumina
and amount of aluminum and alumina. Analysis of Variance (ANOVA) linear regression
was performed, and significance levels for each linear (e.g. xi,
X22, . . . ) ,
and crossproduct (e.g.
X1X2, X 1 X 3 ,...)
X 2 ,...) ,
quadratic (e.g. xi2,
factors were thus determined. The analyses for
all the responses, including ignition time and temperature, combustion time and temperature,
and also characteristics of powders, are summarized in Table 4.12*4.19.
Since a goal of this dissertation was to characterize the relationship between
responses and a set of factors of interest, constructing a model that described the response
over the applicable ranges of the factors was the next step. The fitted model was referred to
as a response surface because the response can be graphed as a surface. The response
surface can be explored to determine important characteristics such as optimum operating
conditions, or relevant tradeoffs when there are multiple responses [Mas89].
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
121
eps
2000-
1300-
1000-
0
2
4
8
8
E n a r g y p tV )
Figure 4.19 Result of EDS analysis of 2 AI2 O3 -3TiC -20% excess Al - 40%Al2Oj.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 4.20 The BSE image with EDS spot analysis of 2 AI2 O3 -3TiC -20% excess
40%Al2C>3 sample.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
Table 4.12 Analysis for the Ignition Time Responses for Both MH and MHH Experiments
MH
Ignition Time (sec)
MHH
Source
Effect of
Parameter
Prob>|t|
Effect of
Parameter
Prob>|tj
Average of all data
48.83
<0.0001
159.83
<0.0001
XI
C particle size (1)
-5.83
0.0699
-25.83
0.0294
X2
AI2 O3 particle size (1)
-4.75
0.1285
-20.00
0.0820
X3
Wt% excess Al (I)
13.17
0.0005
45.83
0.0007
X4
Wt% excess AI2 O3 (1)
16.08
<0.0001
48.33
0.0004
X1*X1
C particle size (q)
3.00
0.2945
3.58
0.7261
X2*X1
Interaction of C and
AI2 O3 particle size
0.38
0.9188
11.25
0.4055
X2*X2
AI2O3 particle size (q)
-2.00
0.4801
-12.67
0.2262
X3*X1
Interaction of Wt%
excess Al and C particle
size
-2.75
0.4587
-11.25
0.4055
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
-0.63
0.8651
-6.25
0.6412
X3*X3
Wt% excess Al (q)
-3.25
0.2576
-8.92
0.3884
X4*X1
Interaction of Wt%
excess AI2 O3 and C
particle size
-4.63
0.2203
-18.75
0.1741
X4*X2
Interaction of Wt%
excess AI2O3 and AI2 O3
particle size
-7.50
0.0557
-3.75
0.7793
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
6.88
0.0767
21.25
0.1267
X4*X4
Wt% excess AI2 O3 (q)
(I) - linear term, (q) - quadratic term
-2.00
0.4801
2.33
0.8193
Term
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Table 4.13 Analysis for the Ignition Temperature Responses for Both MH and MHH
Experiments____________________ ___________________ _________________
Ignition Temperature (°C)
MH
MHH
Source
Effect of
Parameter
Average of all data
145.67
<0 . 0 0 0 1
383.17
<0 . 0 0 0 1
XI
C particle size (1)
-14.92
0.2706
-14.80
0.4751
X2
AI2 O3 particle size (1)
-22.25
0.1086
-33.54
0.1174
X3
Wt% excess Al (1)
30.08
0.0357
61.45
0.0082
X4
Wt% excess AI2O3 (1)
48.75
0.0020
78.13
0.0015
X1*X1
C particle size (q)
-4.35
0.7261
-9.65
0.6166
X2*X1
Interaction of C and
AI2 O3 particle size
4.00
0.8056
6.56
0.7943
X2*X2
AI2O3 particle size (q)
-8.23
0.5102
-37.28
0.0671
X3*X1
Interaction of Wt%
excess Al and C particle
size
-7.88
0.6291
-16.31
0.5195
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
-5.25
0.7469
-1.31
0.9584
X3*X3
Wt% excess Al (q)
1.65
0.8945
-26.16
0.1863
X4*X1
Interaction of Wt%
excess AI2 O3 and C
particle size
1.25
0.9386
-46.93
0.0771
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
-2.63
0.8716
-6.70
0.7905
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
0.50
0.9754
36.19
0.1640
35.77
0.0103
-7.28
0.7053
Term
Wt% excess AI2O3 (q)
X4*X4
(1) - linear term, (q) - quadratic term
Prob>|tl
Effect of
Parameter
Prob>|t|
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
Table 4.14 Analysis for the Combustion Time Responses for Both MH and MHH
Experiments
MH
MHH
Combustion Time (sec)
Source
Effect of
Parameter
Average of all data
57.83
<0 . 0 0 0 1
163.00
<0 . 0 0 0 1
XI
C particle size (1)
-6 . 0 0
0.0964
-26.20
0.0297
X2
AI2 O3 particle size (1)
-4.33
0.2196
-19.80
0.0898
X3
Wt% excess Al (1)
14.00
0.0009
45.63
0.0008
X4
Wt% excess AI2O3 (1)
18.08
<0.0001
48.80
0.0004
X1*X1
C particle size (q)
1.69
0.6016
3.86
0.7103
X2*X1
Interaction of C and
AI2 O3 particle size
1.28
0.7445
10.81
0.4311
X2*X2
AI2O3 particle size (q)
-3.44
0.2944
-12.40
0.2437
X3*X1
Interaction of Wt%
excess Al and C particle
size
-1.38
0.7445
-11.69
0.3956
X3*X2
Interaction of Wt%
excess Al and AI2 O3
particle size
-0.50
0.9055
-4.56
0.7375
X3*X3
Wt% excess Al (q)
-3.44
0.2944
-8.51
0.4175
X44‘Xl
Interaction of Wt%
excess AI2 O3 and C
particle size
-4.75
0.2695
-16.44
0.2377
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
-7.63
0.0855
-2.06
0.8794
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
6.63
0.1306
20.44
0.1470
Wt% excess AI2 O3 (q)
X4*X4
(1) - linear term, (q) - quadratic term
-0.70
0.8309
3.74
0.7192
Term
Prob>|t|
Effect of
Parameter
Prob>|t|
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Table 4.15 Analysis for the Combustion Temperature Responses for Both MH and MHH
Experiments
Combustion Temperature (°C)
MH
MHH
Term
Source
Average of all data
Effect of
Parameter
Prob>|t|
Effect of
Parameter
Prob>|t|
1929.17
<0 . 0 0 0 1
1726.17
<0 . 0 0 0 1
XI
C particle size (1)
23.20
0.0917
-5.58
0.6696
X2
AI2 O3 particle size (I)
3.63
0.7822
1.25
0.9237
X3
Wt% excess Al (1)
-37.80
0.0103
-55.83
0.0006
X4
Wt% excess AI2O3 0)
-61.38
0.0003
-58.67
0.0004
X1*X1
C particle size (q)
47.05
0.0014
16.96
0.1781
X2*X1
Interaction of C and
AI2 O3 particle size
-13.81
0.3950
15.34
0.3434
X2*X2
AI2O3 particle size (q)
59.05
0.0002
66.96
<0.0001
X3*X1
Interaction of Wt%
excess Al and C particle
size
8.44
0.6005
14.25
0.3788
X3*X2
Interaction of Wt%
excess Al and AI2 O3
particle size
7.81
0.6276
0.5
0.9750
X3*X3
Wt% excess Al (q)
-5.32
0.6649
-10.54
0.3936
X4*X1
Interaction of Wt%
excess AI2O3 and C
particle size
-49.44
0.0068
-58.38
0.0021
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
-48.31
0.0079
-61.38
0.0014
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
-117.06
<0.0001
-125.25
<0.0001
-14.32
0.2530
3.08
0.8007
X4*X4
Wt% excess AkCb (q)
(1) - linear term, (q) - quadratic term
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
Table 4.16 Analysis for the Powder Density Responses for Both MH and MHH Experiments
Powder Density (g/cc)
MH
MHH
Source
Effect of
Parameter
Prob>|t|
Effect of
Parameter
Prob>|t|
Average of all data
3.760
<0.0001
3.800
<0.0001
XI
C particle size (1)
0.020
0.0473
-0.030
0.0416
X2
AI2 O3 particle size (I)
0.020
0.1259
-0.010
0.5404
X3
Wt% excess Al (1)
-0.170
<0.0001
-0.170
<0.0001
X4
Wt% excess AI2O3 0)
-0.020
0.0473
-0.030
0.0211
X1*X1
C particle size (q)
-0.020
0.0192
-0.005
0.6881
X2*Xl
Interaction of C and
AI2O3 particle size
-0.010
0.3916
0.000
1.0000
X2*X2
AI2O3 particle size (q)
-0.020
0.0144
-0.020
0.0443
X3*X1
Interaction of Wt%
excess Al and C particle
size
0.000
1.0000
-0.005
0.7378
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
-0.003
0.7454
0.010
0.4548
X3*X3
Wt% excess Al (q)
-0.005
0.5565
-0.010
0.2522
X4*X1
Interaction of Wt%
excess AI2 O3 and C
particle size
0.003
0.8284
-0.002
0.8669
X4*X2
Interaction of Wt%
excess AI2O3 and AI2 O3
particle size
-0.006
0.5895
0.004
0.8016
X4*X3
Interaction of Wt%
excess AI2 O3 and wt%
excess Al
0.006
0.5895
0.004
0.8016
Wt% excess AI2 O3 (q)
X4*X4
(1) - linear term, (q) - quadratic term
-0.009
0.3172
-0.002
0.8549
Term
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
Table 4.17 Analysis for the Particle Size Responses for Both MH and MHH Experiments
MH
MHH
Particle Size (pm)
Term
Source
Effect of
Parameter
Average of all data
1.40
XI
C particle size (1)
X2
Prob>|t|
Effect of
Parameter
Prob>|t|
<0 . 0 0 0 1
1.52
<0 . 0 0 0 1
0 .1 0
0.1072
-0 .il
0.0338
AI2O3 particle size (1)
-0.05
0.4385
-0.03
0.4483
X3
Wt% excess Al (1)
-0.18
0.0072
-0 . 1 2
0.0175
X4
Wt% excess AI2O3 (1)
-0 . 1 2
0.0488
-0.14
0.0096
X1*X1
C particle size (q)
-0.14
0.0248
-0.14
0.0082
X2*X1
Interaction of C and
AI2 O3 particle size
0.09
0.2476
0.07
0.2695
X2*X2
AI2O3 particle size (q)
0.08
0.1839
-0.03
0.5752
X3*X1
Interaction of Wt%
excess Al and C particle
size
-0.06
0.3800
-0.03
0.6667
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
-0 . 1 1
0.1368
-0.08
0.1808
X3*X3
Wt% excess Al (q)
-0 . 1 2
0.0389
-0 . 2 0
0.0005
X4*X1
Interaction of Wt%
excess AI2 O3 and C
particle size
-0 . 0 1
0.9106
-0.01
0.8911
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
-0 . 0 2
0.8292
-0.004
0.9412
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
0 .0 1
0.8427
-0.05
0.4106
-0.10
0.0993
-0.12
0.0154
X4*X4
Wt% excess AI2 O3 (q)
(1) - linear term, (q) - quadratic term
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
Table 4.18 Analysis for the Specific Surface Area Responses for Both MH and MHH
Experiments____________________ ___________________ __________________
MH
Specific Surface Area (m2/g)
MHH
Source
Effect of
Parameter
Average of all data
1.16
XI
C particle size (1)
-0.17
0.0472
0 .2 2
0.0031
X2
AI2 O3 particle size (1)
0 .1 0
0.2207
0.09
0.1849
X3
Wt% excess Al (1)
0.39
0 .0 0 0 2
0.37
<0 . 0 0 0 1
X4
Wt% excess AI2O3 (1)
0.19
0.0267
0.30
0.0004
Xl*Xl
C particle size (q)
0.27
0.0027
0.21
0.0037
X2*X1
Interaction of C and
AI2 O3 particle size
-0.13
0.2065
-0.06
0.4233
X2*X2
AI2 O3 particle size (q)
-0.02
0.7499
0.04
0.4993
X3*Xl
Interaction of Wt%
excess Al and C particle
size
0.09
0.3524
0.09
0.2647
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
0.21
0.0431
0.22
0.0147
X3*X3
Wt% excess Al (q)
0.17
0.0407
0.28
0.0003
X4*X1
Interaction of Wt%
excess AI2 O3 and C
particle size
0.05
0.6360
0.08
0.3337
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
0.04
0.7186
0.09
0.2647
X4*X3
Interaction of Wt%
excess AI2 O3 andwt%
excess Al
0.04
0.6906
0.19
0.0277
0.12
0.1309
0.17
0.0127
Term
X4*X4
Wt% excess AI2O3 (q)
(1) - linear term, (q) - quadratic term
Prob>|t|
<0 . 0 0 0 1
Effect of
Parameter
1.04
Prob>|t|
<0 . 0 0 0 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
Table 4.19 Analysis for the Agglomeration Size Responses for Both MH and MHH
Experiments____________________ ___________________ ________________
MH
MHH
Agglomeration Size (|im)
Source
Effect of
Parameter
Prob>|t|
Effect of
Parameter
Prob>|t|
Average of all data
52.76
<0.0001
54.32
<0.0001
XI
C particle size (1)
4.81
0.0045
5.25
0.0022
X2
AI2 O3 particle size (1)
6.70
0.0003
6.02
0.0007
X3
Wt% excess Al (1)
-0.14
0.9260
0.71
0.6242
X4
Wt% excess AI2 O3 (1)
-5.75
0.0012
-6.72
0.0003
X1*X1
C particle size (q)
0.61
0.6600
0.32
0.8142
X2*X1
Interaction of C and
AI2 O3 particle size
-0.53
0.7663
-0.22
0.9028
X2*X2
AI2O3 particle size (q)
0.53
0.6957
-0.10
0.9391
X3*X1
Interaction of Wt%
excess Al and C particle
size
1.30
0.4744
0.21
0.9078
X3*X2
Interaction of Wt%
excess Al and AI2O3
particle size
1.24
0.4944
2.29
0.2101
X3*X3
Wt% excess Al (q)
-2.21
0.1214
-2.50
0.0802
X4*X1
Interaction of Wt%
excess A1 2 0 3 and C
particle size
-0.08
0.9647
-2.48
0.1755
X4*X2
Interaction of Wt%
excess AI2 O3 and AI2 O3
particle size
-0.14
0.9387
1.08
0.5461
X4*X3
Interaction of Wt%
excess AI2 O3 and wt%
excess Al
-1.97
0.2826
-3.64
0.0546
X4*X4
Wt% excess AI2 O3 (q)
(1) - linear term, (q) - quadratic term
1.35
0.3344
0.32
0.8142
Term
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
131
Data from central composite design can be efficiently used to fit a response-surface
model. A full quadratic response-surface model in four variables consists of the following
15 terms:
y = 50 + 5, jc, + B 2x 2 + B3x3 + B 4x 4
+ Bnxxx2 + 5,3jc,x3 + B 14x xx 4 + B23x 2x 3 +
B 24x 2x 4
+ B34x3x4
(4.10)
+ Bnx ‘ + Bnx\ + B 33x \ + Bu x;
where y is the investigated response, B0 is the average of all die observations, Bj, B,j,and B„
are the effect of parameters in linear, crossproduct, and quadratic term and Xb Xy, Xu are the
value of the parameters in linear, crossproduct, and quadratic term, respectively.
By using the effect of parameters provided in Table 4.12-4.19, the relationships
between the effect of the parameters and the investigated response were generated. For
example:
Te = 1929.17 + 23.20CParticleSize + 3.63A120 3ParticleSize
- 31.19Wt%excessAl - 6 1.3Wt%Al20 3
-13.81(/4120 3ParticleSize$(2particleSize)
+8M(fPt%excessAl){CparticieSize)
+ l.il(Wt%excessAl){Al20 3particleSize)
- 49M(Wt%excessAl20 3)(CparticleSize)
- 48.3 l{jVi%excessAl20 3\A120 3ParticleSize)
- 1ll.06(fVt%excessAl2O3)(ffrt%excessAl)
+41.0S{CparticleSize)2 +59.05{Al2O3ParticleSize)2
-5.32{m%excessAlf - \4.3ltyt%excessAl20 3f
where Tc (combustion temperature) is in °C and particle size of materials (C or A I 2 O 3 ) is in
microns.
The significance of each parameters can be determined by using die Rule of Thumb
for p(2-tail) values. Thus some parameters in the equation, which are insignificant, could
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
132
possibly be deleted from the model. The Rule of Thumb for p(2-tail) is usually as follows
[Sch88]:
•
prob(2-tail) < 0.05 implies term is significant; place that term in the model
•
prob(2-tail) > 0.10 implies term is insignificant; leave that term out of the
model
• 0.05 <prob(2-tail) < 0.10 is a grey zone where most investigators will
typically decide to place the term in the model
The significance at the 0.5 levels (above 95%) of each parameter on the ignition
behavior and characteristic of MH and MHH ignited powders are presented in Table 4.20
and Table 4.21, respectively. The higher the significance level, the greater the certainty that
the parameter is an important factor contributing toward the investigated response value. A
dash represented the significance value below 90%, which were not considered significant.
The final regression models for predicting all the investigated response values for
MH and MHH experiments are summarized in Table 4.22 and Table 4.23, respectively. For
example,suppose the objective was to use only 10% Al (-1 level) and 10%Al2O3 (-1 level),
and the only available carbon and alumina were the intermediate size (0 level). Then the
ignition time of the sample ignited by MH could be estimated by using the first relationship
shown in Table 4.22. That was 206 seconds (tig = 159.83 - 25.83 (-1) -20 (-1) + 45.83 (0) +
48.33 (0) = 205.66 seconds).
The relationship between the effect of the parameters and the investigated responses
can also be illustrated graphically by the Prediction Profiler plots. The Prediction Profiler
Plots for investigated responses including ignition time and temperature, combustion time
and temperature, are displayed in Figure 4.21-4.24, respectively.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
133
>99.99
>99.99
>99.99
>99.99
>99.99
C particle size (1)
(93.00)
-
(90.36)
(90.83)
95.27
-
95.28
99.55
•
•
-
-
-
-
99.97
AI2 O 3 particle size ( 1 )
-
§
i r i
A gglom eration
Size (pm )
Particle S ize
(pm )
>99.99
C om bustion
T em perature
<°C) .
>99.99
C om bustion
Time (sec)
>99.99
Ignition
T em perature
<°C)
A verage o f ail data
Parameter
Ignition T im e
(sec)
Powder D ensity
(g/cc)
,
Table 4.20 Relative Significance of Reaction Parameters on all Investigated Responses for
the MH Experiment___________________________________________________
Significance (%)
MH
W t% excess A1 (1)
99.95
96.43
99.91
98.97
>99.99
99.28
99.98
-
W t% excess AI2 O 3
>99.90
99.80
>99.90
99.97
95.27
95.12
97.33
99.88
C particle size (q)
-
-
-
99.86
98.08
97.52
99.73
-
Interaction o f C and
AI2 O 3 particle size
-
•
-
-
-
-
•
•
AI2 O 3 particle size
-
-
-
99.98
98.56
-
-
-
Interaction o f W t%
excess A1 and C
particle size
-
-
*
-
-
-
-
*
Interaction o f W t%
excess A1 and AI2 O 3
particle size
-
-
-
-
-
*
95.69
*
W t% excess A1 (q)
-
-
-
-
-
96.11
95.93
-
Interaction o f W t%
excess AI2 O 3 and C
particle size
-
-
•
9932
*
*
•
-
(1)
(q)
Interaction o f W t%
excess Al2 0 3 and
AI 2 O 3 particle size
(94.43)
*
(91.45)
9931
-
-
*
*
Interaction o f W l%
excess AI2 O 3 and
wt% excess A1
(92.33)
-
-
>99.90
-
•
-
•
98.97
-
-
-
(90.07)
-
-
W t% excess AI2 O 3
-
(q)
(1) - linear term, (q) - quadratic term,
( )- grey zone where 0.05 <prob(2-tail) < 0.10 ( 90% significance < 95%)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
>99.99
-
97.03
-
95.84
96.62
99.69
99.78
(91.80)
-
(91.02)
•
-
•
-
99.93
Wt% excess A1 (1)
99.93
99.18
99.92
99.94
>99.99
98.25
>99.99
-
W t% excess AI2 O 3
99.96
99.85
99.96
99.96
97.89
99.04
99.96
99.97
C particle size (1)
97.06
AI2 O 3 particle size (1)
A gglom eration
Size (pm )
>99.99
Specific
Surface Area
(m 2/g)
...
Particle Size
(pm )
>99.99
>9 9 . 9 9
C om bustion
Tem perature
fC )
>99.99
>99.99
C om bustion
Time (sec)
>99.99
Average o f all data
Ignition
Tem perature
m
>99.99
Parameter
Ignition T im e
(sec)
Powder D ensity
(g/cc)
Table 4.21 Relative Significance of Reaction Parameters on all Investigated Responses for
the MHH Experiment__________________________________________________
Significance (%)
MHH
(1)
C particle size (q)
-
•
-
-
-
99.18
99.63
•
Interaction o f C and
AI2 O 3 particle size
-
-
-
-
•
-
-
-
AI2 O 3 particle size
-
(93.29)
-
>99.9
95.57
-
-
-
Interaction ofW t%
excess A1 and C
particle size
-
-
•
-
-
-
-
*
Interaction o f W t%
excess A1 and AI2 O 3
panicle size
-
-
-
-
-
*
98.53
-
W t% excess A1 (q)
-
-
-
-
-
99.95
99.97
(91.98)
Interaction o fW t%
excess AI2 O 3 and C
particle size
-
(92.29)
-
99.79
-
•
-
*
Interaction o f Wt%
excess AI2 O 3 and
AI2 O 3 particle size
-
-
-
99.86
-
*
-
-
Interaction o f Wt%
excess AI2 O 3 and
wt% excess A1
-
-
-
>99.90
-
-
97.23
(94.54)
W t% excess AI2 O 3
-
•
-
-
-
98.46
98.73
-
(q)
(q)
0) - linear term, (q) - quadratic term,
( )- grey zone where 0.05 <prob(2-tail) <0.10, or 90% significance < 95%
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
135
Table 4.22 Summary of the Regression Models for Predicting the Ignition Behavior and
Characteristic of MH Ignited A I 2 O 3 - T 1 C Powders_____________________________
Prediction Equations
MH
Ignition Time
(sec)
t,g = 48.83 -5.83CParticleSize + 13.17Wt%excessAl + l6.08R^r%y4/2O3
Ignition
Temperature
Tlg = 145.67+30.0Wt%excessAl+ 4&.7Wt%excessAl20 3
(°C)
- 7.5(Wt%excessAl20 3\A l20 3ParticleSize)
+ 6.%$(Wt%excessAl20 3fWt%excessAl)
+ 35.77(m%excessAl2Oi )2
Combustion
Time (sec)
tc = 57.83 - (CParticleSize+14Wt%excessAl+1 %.03Wt%Al2O3
- 7 .63(}Vt%excessAl20 3\A l20 3ParticleSize)
Combustion
Temperature
Te = 1929.17 + 23.20CparticleSize
- 37,79Wt%excessAl - 6\.3W t%Al20 3
(°C)
+ 47.05{CparticleSize)2 + 59.0S(Al2O3ParticleSize)2
- 49.44(fVt%excessAl20 3XCparticleSize)
- 48.3 l(jVt%excessAl20 3\A120 3ParticleSize)
- 1 \7 SXtjm«excessAl20 3\Wta/oexcessAl)
Powder
Density (g/cc)
Particle Size
(pm)
Specific
Surface Area
(m2/g)
Agglomeration
Size (pm)
p = 3.76+0.02CparticleSize~0.l7Wt%excessAl-0.02Wt%Al20 }
- 0.02(CparticleSizef -0.02(Al20 3ParticleSizef
P.S. = 1.39-0. lWt%excessAi-QA2fVt%Al20 3
- 0.\4(CparticleSizef -0.\2(Wt%excessAl)2 ~0.l0{jVt%excessAl2O3f
Sp.SA. = 1.16-0.17CparticleSize+ 9.3%Wt%excessAl+0.19Wt%Al20 3
+0.27{pparticleSizef +0.17(Wt%excessAi f
+0.2 l(lVt%excessAl\A l20 3ParticleSize)
AS. = 52.76+4.81CParticleSize + 6.70Al20 3ParticleSize
- 5.75Wt%excessAl20 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4 .23 Summary of the Regression Models for Predicting the Ignition Behavior and
Characteristic of MHH Ignited AkCVTiC Powders___________________________
MHH
Ignition Time
(sec)
Prediction Equations
tlg = 159.83 - 25.83CParticleSi2 e - 20 Al20 3ParticleSize
+ 45 ,%3Wt%excessAl + 4&.33Wt%Al20 3
Ignition
Temperature
(°C)
Tlg = 383.17+61A6Wt%excessAl + 7%.\3Wt%excessAl20 j,
Combustion
Time (sec)
te = 163 - 26.21CParticleSize -19.80 Al20 3ParticleSize
+ 45.63Wt%excessAl + 48.79H^%y4/20 3
Combustion
Temperature
Tc = 1726.17 - 55.&3m%excessA/
(°C)
Powder
Density (g/cc)
- 37.2 ^A l20 3ParticleSizef - 46.9^Vt%excessAl20 3\CParticleSize)
- 5&.67Wt%Al20 3+ 66.96{Al20 3ParticleSizef
-5^.39{Wt%excessAl203XCparticleSize)
- 61.3S{fVt0/oexcessAl2O3){Al2O3ParticleSize)
-125,25(Wt%excessAl20 3)(Wt%excessAI)
p = 3.%0-0.03CparticleSize-0.\7Wt%excessAl
- 0.03Wt%Al203-0.02{Al20 3ParticleSizef
Particle Size
(pm)
PS. = 1.51- 0.1 \CParticleSize - 0. V3W(%excessAl - 0.\AWt%Al20 3
Specific
Surface Area
(m2/g)
SpSA. = 1.05+Q.22CparticleSize+ 0.31Wt%excessAl+0.29Wt°/oAl2O3
Agglomeration
Size (pm)
- 0.14(CparticleSizef - Q.2Q(fVt%excessAl)2-0. XliWtVoexcessALfii f
+0.2 {(CparticleSize^ +0.28(lF/%exces.&4/)2+0.l7(ffrt%excessAl2O3f
+0.22(jVt%excessAlXAl2O3ParticleSize)
+0.19(jVt%excessAiy[fVt%excessAl20 3)
AS. = 54.32 + 5.24CParticleSize + 6.02Al20 3ParticleSize
- 6.7 Wt%ejccessAl20 3 - 2.5(Wt%excessAlf
- 3,64(Wt%excessAl20 3)[Wt%excessAl)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
a)M H
£
110
a>
■
7
L1 '
E
® 48.83
0
1
T
• 1 i T1
— 1— H Lp
•
T
C partide size
(urn)
■a\
i1 ■
, xi
■H
— T■1r 1 — ■— 1—T"f J£
—r
1
A l ^ partide size
(nm)
Excess Al
(wt. %)
1
Excess AljOj
(wt. %)
b) MHH
8
</>
360
|
159.83
I
20
cO)
©
C partide size
(am)
AIjO, partide size
(urn)
o
o
CM
Excess Al
(wt %)
©
o
CM
O
Excess AljOj
(wt %)
Figure 4.21 Prediction Profiler of ignition time of AfeOj-TiC powders (Bar represents the
standard deviation).
R e p ro d u c e d with permission o f the copyright owner. Further reproduction prohibited without permission.
138
a) MH
oP
2^
537 -—
S
3
8.
I
145.67
56
c
-
i
£
o
o
so
c)
O
I1 I '
H
1
1
'
!
•
■
£
D
<D
T
8
se
O
O'
'
'
'
Q
f
(n
i
S
C partide size
Al 0 partide size
(pm)
(pm)
2
i
°
3
iL l
T
H
i
-L
1
•
1
on
■
I
o
Excess Al
(wt%)
T" 1
o
•
.
<
(3N
11
eV
Excess AJ,0,
(wl%)
b) MHH
o
S
603 -
r . 7T -
3
1
383.17 .
E
2
c
O
C
o>
s iT '.
107 H
f-
•"V
- fi 4
F ^'
oC
N
O
w
C partide size
A^Oj partide size
(pm)
(pm)
Excess Al
(wt%)
©
b
f*.
<n
o
o
U
r-)
o
o
9
o
"AT-
M .
Excess ALO,
(w t% )
Figure 4.22 Prediction Profiler of ignition temperature of AhOj-TiC powders (Bar
represents the standard deviation).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
139
a) MH
Io
117-
■E■s
s 5783:
M
Q
E
8‘
O
h*
cn
O
©
CM
CparticleSize
(pm)
Al 0 particle size
(pm)
2
3
O
▼
o
Excess Al
(wt %)
oCM
o
o*•
Excess Al 0
(wt %)
2
3
b) MHH
8
®
E
33
362-
m
' T T
163 '
o
tn
■
13-
H
o
b
rc—
o
o
o
X
1
E
o
TT •
1 ■
■*i
o
o
C particle size
(pm)
©
b
oCM
Al 0 particle size
(pm)
2
3
K ©
CM
Excess Al
(wt %)
Excess Al,0,
(wt %)
Figure 4.23 Prediction Profiler of combustion time of AkCVTiC powders (Bar represents
the standard deviation).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
a) MH
&
8.
=
2236 -
.2
1661 -
V
| ^1929.17-
<A
n3
E
oo
t
,Ski
o
5;
o
CM
o
o
O
O
CM
CM
o
C particle size
AI O particle size
Excess Al
Excess Al,0
(pm)
(]>m)
(wt%)
(wl%)
O
O
Excess Al
(wt%)
Excess AL0
(wt%)
2
3
3
b) M HH
£
a
s»
a.
2090 -
E
1176.17
I
3
A
E
o
O
I-i
■ i-T
«O
o
on
o
s . .
1478 -
m
o
o
o»
o
o
CM
C particle size
AiPa particle size
(urn)
(pm)
CM
CM
Figure 4.24 Prediction Profiler of combustion temperature of AkOa-TiC powders (Bar
represents the standard deviation).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
3
141
From the plots, as the particle size of carbon increased the ignition and combustion
time decreased under MH. However, under MH, the particle size of carbon did not change
the ignition and combustion time significantly.
The temperature profiles of 2 AI2O3 -3TiC -20%excess Al - 20%excess AI2 O3 with a
variation of carbon particle size are shown in Figure 4.25. It was observed that the slowest
ignition time occurred when the smallest carbon particle size was used for both MH and
MHH techniques. When carbon is small, it may be possible for a continuous carbon phase to
act as a barrier to titania-aluminum particle contact, which is believed to be necessary for the
initial reaction to Ti0 2 and Al. When the carbon size is increased, the titania-aluminum
particle contact is greater which may lead to decreased ignition time. However, the sample
with intermediate carbon particle size was ignited faster than one with largest carbon particle
size. This may be because it provided the best reactant particle arrangements and particle
contact between titania-aluminum which resulted in fastest ignition time. In the MH case,
the sample with the largest carbon particle size was observed to have highest combustion
temperature. The reason for this was thought to be due to a decrease surface area of the
larger carbon particle size, decreasing heat loss. When using the MHH technique, a variation
of carbon particle size did not change combustion temperature significantly (as also shown
in the prediction profiler in Figure 4.21). The susceptor in this case provided another heat
source and helped reduced heat loss from the surface of the sample.
Figure 4.26 shows the temperature profile of
20%excess
A I2 O 3
2
AI2 O3 -3TiC -20%excess Al -
with variation of alumina particle size. As the alumina particle size
increased the ignition time and temperature decreased. This was because the rest of the
reactants could act as a continuous phase and thus increased the particle contact between
a lu m in u m
and titania, resulting in faster ignition.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
142
2500
0.075 pm
0.037 urn
0.017 pm
(a) MH
2000
-
ba
1500 -
u
1000
-
500 -
0
50
150
100
200
Time (sec)
2500
0.075 pm
0.037 pm
0.017 pm
(b) MHH
2000
1500
S.
£
£
1000
500
0
100
200
300
400
Time (sec)
Figure 4.25 Temperature profile of 2 AI2 O3 -3TiC -20% excess Al -20% excess A I 2 O 3 with
varying carbon particle size ignited by (a) MH and (b) MHH. The significance of all data is
approximately 99.9%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
2500
63-74 pm I
20-38 pm ]
(a) MH
2000
2
1500
1
I
E
<D
E- 1000
500
0
0
50
100
150
200
Time (sec)
2500
63-74 urn
20-38 urn
0.49 pm
(b)MHH
2000
|§
1500
&
E
£
1000
500
0
100
200
300
400
Time (sec)
Figure 4.26 Temperature profile of 2 AI2 O3 -3TiC -20% excess Al -20% excess A I 2 O 3 with
varying alumina particle size ignited by (a) MH and (b) MHH. The significance of all data is
approximately 99.9%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
Samples with smaller alumina particle size had lower combustion temperatures,
presumably due to the reaction taking place over a longer time where heat losses become
more significant. However, the sample with an intermediate alumina particle size (20-38
microns) was observed to have a lower combustion temperature than the one with the
smallest alumina particle size (0.49 microns). A possible explanation for this was that the
titania-aluminum particle contact network was disrupted when intermediate size alumina
was used. With AI2O3 particle size (20-38microns) being close to Al particle size
(lOmicrons), smaller TiC>2 particles (2-3microns) would be distributed in the intersticies of
packed larger particles of both AI2O3 and Al. Melting occurred locally and only the Al in
direct contacted with TiOi reacted. On the other hand, using the smallest alumina, both Ti02
(2-3microns) and AI2 O3 (0.49 microns) would fill the voids between larger Al particles
(lOmicrons). This structure resulted in longer ignition time. However, Al melting occurred
uniformly throughout the sample and thus the combustion temperature was higher than for
the former case where an intermediate AI2O3 was used. These trends were also in agreement
with the prediction profiles in Figure 4.21-4.24.
The effect of diluent addition on ignition behavior of 2 Ah0 3 -3 TiC-Al are also
shown in Figure 4.21-4.24. The addition of diluents (Al and
A I2 O 3 )
to the reactants
increased ignition temperature, and decreased the combustion temperature. In addition, the
ignition time increased as the amount of the diluents increased. This was due to the
decreased heat generation rate. In addition, it was thought that the high microwave
reflectivity of the Al partially shielded the sample. In other words, sample containing large
amounts of Al had a chance to form an interconnected network of metal particles which
would reflect microwaves and resulted in a shallow microwave penetration depth. Samples
having small amounts of Al particles allowed microwaves to penetrate into the whole
s a m p le
far greater than those having large amounts of Al. Furthermore, as the amounts of Al
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
increased, the amounts of the rest of the reactants (TiCh, C, AI2O3 ) decreased. The decreased
amount of carbon, which are effectively heated with microwave, could result in less efficient
absorption of microwave radiation. The AI2O3 particles also may act as a barrier to reactant
particle contact. Alumina has low electrical conductivity and very low loss tangents thus
allowing microwaves to pass through with very little absorption. These trends were similar
in both MH and MHH techniques.
The temperature profile of
2
Al2 0 3 -3 TiC with various amounts of Al and AI2 O 3 are
shown in Figure 4.27-4.28. It was observed that as the amount of diluents increased, the
ignition time and temperature increased. In addition, the larger the difference between the
initial and ignition temperature was, the lower the combustion temperature became. This
was because more heat energy which generated from the reacted reactant powders was used
in order to increase the temperature of adjacent reactant powders up to the ignition
temperature. It was also observed that, after the reaction was over, the absorption of
microwave energy by the product helped maintain the sample at an elevated temperature
instead of allowing it to cool to room temperature.
The Prediction Profiler Plots of the characteristic of the ignited powders, including
powder density, agglomeration size, particle size, and specific surface area, are displayed in
Figure 4.29-4.32, respectively.
The effect of the addition of Al on the final product density was much more
significant as compared to die other effects (Figure 4.29). The density of product decreased
with increasing amount of Al. As the amount of Al increased, the combustion temperature
decreased. The product would take a shorter time to cool and therefore a liquid phase would
be present in the product for shorter periods. This resulted in the less densified product. It
was noted that the effect of amount of AI2 O 3 on the product density also showed a similar
trend but to a lesser extent.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
146
2500
0% excess Al |
20% excess Al j
40% excess Al j
(a)MH
2000
V
£
3
2
1500
S.
£S
1000
500
0
50
100
150
200
Time (sec)
2500
(b) MHH
£
i
1500
s
£
1000
Um
0% excess Al
20% excess Al
40% excess Al
Si
500
0
0
100
300
200
Time (sec)
400
Figure 4.27 Temperature profile of 2 AI2O3 -3TiC -20% excess AI2 O3 with varying excess
amount of Al ignited by (a) MH and (b) MHH. The significance of all data is approximately
99.9%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
2500
0% excess Al2 0 3
20% excess Al2 0 3
40% excess Al2 0 3|
(a) MH
2000
|
1500
.«
1000
500
0
50
150
100
200
Time (sec)
2500
0% excess Al2 0 3
2 0 % excess AI2 O 3
40% excess AI2 O3
(b) MHH
2000
-
2
£2
&
E
£
1500 -
1000
-
500 -
0
100
200
300
400
Time (sec)
Figure 4.28 Temperature profile of 2 AI2O3 -3TiC -20% excess Al with varying excess
amount of AI2 O 3 ignited by (a) MH and (b) MHH. The significance of all data is
approximately 99.9%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
a) MH
4.09 -
& 3.76 • Y *
* -1
(A
s
Y *
N
3.37 -
* i
■
o
»
3
<=-
■
i
in
£
°
C partide size
(pm)
i
'
a
aT
3
'M
•
i
^
Al 0 partide size
(miti)
2
3
i
O
•
3N
'
Excess Al
(wt %)
1
O
^
1
O
<3
1
1
O
^
<M
Excess Al 0
(wt %)
2
3
b) MHH
4.2 £(A
3.80
T—i
C
s
«
o
o
©
o
p ^.
©
o
C partide size
(pm)
cn
Q
r-
Al 0 partide size
(pm)
2
3
cm
Excess Al
(wt %)
Excess ALO,
(wt%)
Figure 4.29 Prediction Profiler of powder density of AhCh-TiC powders (Bar represents the
standard deviation).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
The effect of various parameters on the agglomeration size of ignited powders are
shown in Figure 4.30. It can be seen that the larger reactant particles produced coarser
reacted powders. The addition of excess AI2O3 to the reactants decreased the agglomeration
size. The reason for this decrease may be that as the amount of AI2O3 increased, the
combustion temperature is lowered below the melting point of the alumina phase and thus
there is less liquid phase present during the reaction. On the other hand, the agglomeration
size did not change significantly with the amount of excess Al.
The effect of the reactants particle size and the addition of diluents on the product
particle size seemed to depend on the corresponding combustion temperature. That is, lower
combustion temperature resulted in smaller particle size of product. For example, as the
combustion temperature decreased with the increase amount of diluents, the particle size of
product decreased (Figure 4.31). Figure 4.32 exhibits the prediction profiler plot of the
specific surface area of the resulting powders, which shows the reverse trends to the plot of
final particle size of product (Figure 4.31).
Some interactions between the parameters also have significant effects on the
investigated responses. Figure 4.33 shows the interaction profile plots of amount of excess
AI2O3
and its particle sizes. The strong interaction between these two parameters was
illustrated. Fastest ignition time was observed when using small AI2O3 particle and small
amounts of AI2O3. With larger additives of AI2O3, larger AI2O3 particles resulted in faster
ignitions. The effect of interaction between the amount of excess AI2O3 and carbon particle
size on ignition temperature are shown in Figure 4.34. The lowest ignition temperature was
obtained with small amounts of AI2O3 and small carbon particles. As amount of AI2O3 was
increased, larger carbon particle sizes would yield lower ignition temperatures.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
150
a) MH
80.08 'J r \
w 52.76 ■ T
1
®
31.5 -
H
' -
’
K
o
o'
H
-
-
■
-
T
■ H
H
o
o
CparticleSize
(pm)
A l ^ particle size
(pm)
Excess Al
(wt %)
Excess A l ^
(wt %)
b) MHH
®
87.9 -
N
to
'J r \
g 54.32
tS
(a 34.58 ®
£o
8
<
v -H i
•
Jr*'
f-
O
cn
o
6
in
r-.
O
T
o
CN
C particle size
A^Oj particle size
(pm)
(pm)
T>
H
O
O
Excess Al
(wt%)
Excess Al,0,
(wt%)
CN
CN
O
Figure 4.30 Prediction Profiler of agglomeration size of AkCVTiC powders (Bar represents
the standard deviation).
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
151
a) MH
J.
2.11 H
<D
£
’ T -p
1-39
h
-
73
1
CL
r
H
055 1
- L- "II -- .'
. ij :- fI
®
1
•
,SN
t©
o
oS;
o
in
r-»
o
o
C particle size
(pm)
01
o
CM
CM
Excess Al
(wt%)
Al 0 partide size
(pm)
2
O
o
o
3
Excess ALO,
(wt. %)
b) M HH
§
in
1.77 H
1.52
4 -;
o
|
<0
CL
0.45 -
■>— r
T
o
o
cn
o
o
in
p-
§
C partide size
(pm)
1— r
T
0»
°
O
Al 0 partide size
2
F
3
(pm)
w - ;
o
CM
Excess Al
(wt%)
o
1
o
r
o
■«— r
CM
o
Excess AljOj
(wt %)
Figure 4.31 Prediction Profiler of particle size of AhCb-TiC powders (Bar represents the
standard deviation).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
a) MH
|
3.05 -
©
8
~
3
™
S. 1.16 ■
1
0 .7 7 -
<2
K sT
1
V
' — .
-* T
1
"
IT
I1 I1 I Tj
1r ■
V- r - |> i_i— ■
— — —
T"Lr—■
—
• i
1
O
o'
C partide size
(pm)
Al 0 partide size
(pm)
2
Excess Al
(wt %)
3
Excess Al 0
(wt %)
2
3
b) MHH
(9
I
3.76 -
■cJ?
wE. 1.05
& 0.89 H
•o
®
ca.
n
o'o
o
-i— r
co
o
b
p*«-
o
o
C partide size
(pm)
I I :r*
T
a»
o
\
AljOa partide size
(pm)
T
O
T
Excess Al
(wt%)
r
O
■>— r
o
Excess AU0
(wt%)
Figure 4.32 Prediction Profiler of specific surface area of AUOa-TiC powders (Bar
represents the standard deviation).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
110
8
w
<u
E
c
o
■5
co>
7
0
10
20
30
40
Excess Al20 3 (wt%)
Figure 4.33 The effect of the amount of excess AI2 O3 and its particle size on the ignition
time of MH ignited AkOj-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
—
603
0.017 p m c
0.027 pTT! c
0.037 Jim c
O
O
<D
2
2
2L
E
2
c
o
0.042 jam c
0.075 Jim c
O)
107
0
10
20
30
40
Excess Al20 3 (wt%)
Figure 4.34 The effect of the amount of excess AI2 O3 and carbon particle size on the ignition
temperature of MHH ignited AfeOj-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
Figures 4.35-4.37 show the interaction profile plots of C particle size, AI1O3 particle
size, and amount of excess Al with the amount of excess AI2O3, respectively. The sample
with the largest carbon particles and smallest amount of excess AI2O3 would result in the
highest combustion temperature, while small carbon particles would give higher combustion
temperature when large amount of excess AI2O3 was used (Figure 4.35).
The effect of interaction between the amount of excess AI2O3 and its particle size on
combustion temperature was also observed (Figure 4.36). The highest combustion
temperature would occur when using large AI2O3 particles present in small quantities. As the
amount of excess AI2O3 increased, a smaller AI2O3 size would be needed in order to achieve
a high combustion temperature.
The strong interaction between the amount of excess Al and AI2 O3 on combustion
temperature is shown in Figure 4.37. The combination of large amounts of Al and small
amounts of AI2 O3 or vice versa would result in the highest combustion temperature. When
using large amounts of both Al and AI2O3 , the lowest combustion temperature would be
obtained.
Figure 4.38 shows the effect of interaction between the amount of excess Al and
AI2O3
on the agglomeration size of the reacted powders. The smallest agglomeration size
would be obtained when using large amounts of both Al and AI2O3 , presumably due to the
decreased combustion temperature. When using small amounts of AI2O3 , a small amount of
Al would also be required in order to obtain small agglomerates of reacted powder.
The effect of the amount of excess AI2O3 and its particle size on the agglomeration
size of reacted powders is shown in Figure 4.39. The result showed that there was no
interaction between these two parameters. The agglomeration size decreased with increasing
amount of excess AI2O3 regardless of particle sizes used. As AI2O3 particle size was
increased, the resulting agglomeration size increased.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0% excess A1 0
2
3
10% excess Al 0
3
20% excess Al 0
3
O
30% excess Al 0
3
£
4Q% excess A1 0
3
2
2090 .
2
2
2
3
2
&
E
c
.2
to
3
A
E
o
O
1478.
0.017
0.037
0.075
Carbon particle size (pm)
Figure 4.3S The effect of the amount of excess AI2O3 and carbon particle size on the
combustion temperature of MHH ignited AkOa-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
2090.
O
O
a
2
2
8.
E
2
c
o
w
3
A
E
o
O
0% excess Al 0
2
1 0
1478.
3
% excess AI O
2
20% excess Al 0
2
3
3
30% excess AljOj
40% excess Al 0
2
T
0.049
3
T
20
74
Al20 3 particle size (pm)
Figure 4.36 The effect of the amount of excess AI2O3 and its particle size on the combustion
temperature of MHH ignited AhOj-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
2 0 9 0 -----
O
o
3
oC
(0
3
£
E
o
o
1478-----
0
10
20
40
30
E xcess Al (wt%)
0% excess Al 0
2
3
1 0
% excess AljOa
2 0
% excess AljOa
30% excess AI O
2
3
40% excess AljOj
Figure 4.37 The effect of the amount of excess AI2 O3 and Al on the combustion temperature
of MHH ignited AlzOs-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
0% excess Al
10% excess Al
88.0
E
20% excess Al
0)
N
30% excess Al
3
40% excess Al
Cn
2
0)
E
o
o>
O)
<
34.6
10
20
30
40
Excess Al20 3 (wt%)
Figure 4.38 The effect of the amount of excess AI2 O3 and Al on the agglomeration size of
MHH ignited AUC^-TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
0.49 nm aj o
2
8 8 .0 -----
—
1 0
pm ai o
2
3
3
E
3
20-38 pm ai o :
8
0)
44-53 nm ai o :
2
2
63-74 J im Al 0;
c
2
o
2
®
E
o
cn
<
3 4 .6 -----
Excess Al20 3 (wt%)
Figure 4.39 The effect of the amount of excess AI2 O3 and its particle size on the
agglomeration size of MHH ignited AUOj-TiC powders.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
161
It was also found that there was no interaction between any parameters that affect the
final powder density. As seen from Figure 4.40, the density of product decreased as the
amount of excess Al increased with any amount of AI 2O 3 used. It also showed that as the
amount of AI2O 3 increased, the density decreased even though it was not as significant as
the effect of the amount of Al on the final powder density.
4.1.3 Summary
AhCb-TiC powders were produced by a combustion synthesis technique using
microwave heating (MH) or microwave hybrid heating (MHH) as a heating source. A clear
difference in ignition time and temperature was observed for the two heating methods. The
MHH samples required longer time to reach ignition temperature. Thus they ignited at
higher temperature than MH samples and resulted in lower combustion temperatures. The
results showed no significant difference in the characteristics of powders (density, particle
size, surface area, and agglomeration size) ignited by these two heating methods.
The experimental designs were conducted on the synthesis of AliCVTiC powders to
investigate the effects of reaction parameters on ignition behavior and characteristics of the
resulting powders. Firstly, a sixteenth-run Plackett-Burman design was performed to analyze
the significance of reaction parameters and to identify a small number of dominant
parameters for subsequent more investigation. Based on these results, the most significant
factors affecting the investigated responses were amount of excess Al and AI2O 3 , particle
size of C and AI2 O 3 , and also heating method. The thirtieth-run central composite design
(CCD) was then performed to investigate the effect of those parameters on the responses of
interest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
0% excess Al 0
2
4.2
3
10% excess Al 0
3
20% excess Al 0
3
30% excess Al 0
3
40% excess Al 0
3
2
2
2
2
3
S'
CO
c
<D
o
3.4
T
10
~T
20
30
40
Excess Al (wt%)
Figure 4.40 The effect of amount of excess AI2 O 3 and Al on the final density of MHH
ignited AfeCVTiC powders.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
163
The addition of diluents to the reactants increased ignition time and temperature,
while decreasing the combustion temperature. An increased particle size of alumina lowered
the ignition time and temperature. While for the particle size of carbon, the ignition time and
temperature of the ignited sample decreased with an increased particle size of carbon under
MHH only. Under MH, an increased particle size of carbon did not change the ignition time
and temperature significantly. It was also observed that the sample with the largest alumina
or carbon particle size had die highest combustion temperature. It was also found that the
density of product decreased with increasing amounts of Al. The addition of excess AI2 O3 to
the reactants resulted in decreasing the agglomeration size.
The investigated responses also were influenced by the interactions between the
these parameters. In some cases, the interaction effects were significant, but in no case was
the interactive effect as large as for individual parameters. The interaction between the
amount of excess
A I2 O 3
and other parameters was found to have more effect on the
responses than any other combinations of the interaction.
Various empirical models relating these important reaction parameters and their
interactions to the investigated responses were then developed. This provided the
information required for adjusting the setting of the important factors in order to obtain the
desired products.
4.2 Mechanism Studies
The combustion synthesis reaction for mechanism studies was based on equation
(3.1):
3Ti02 + 3C+ (4+ x)Al + yAl20 2 -> 3TiC + (2 + y)Al20 } +xAl
(3.1)
where x and y were 4.74 and 1.25 which corresponded to 20% excess Al and 20% excess
AI2O3 , respectively. Four compositions studied were (1) stoichiometric (x = y =0), (2)
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
164
composition with excess Al (x = 4.74, y = 0), (3) composition with excess AI2O 3 (x = 0,
y=l .25), and (4) composition with excess both Al and AI2 O 3 (x = 4.74, y = 1.25).
The particle size of Ti02, Al, C and AI2O3 were 2-3 pm (99% rutile, Alfa Aesar), 10
pm (99.7%, Alfa Aesar), 0.017 pm (99%, Cabot Corp.) and 0.049 pm (99.8%, Alcoa),
respectively. Based on findings from the first part of this work, samples with the smallest
carbon or alumina particle size took the longest time to ignite. Choosing this particle size for
mechanism studies allowed more control of the reaction and gave more time to observe the
heating behavior of the reactants from the beginning to the moment just before the ignition
occurred.
4.2.1 Microwave heating
The experimental approach was to observe the heating behavior of single
components, two components, three components, and four component interactions during
and after reaction initiation. The combinations of mixtures for each compositions are
summarized previously in Table 3.6. The amounts of reactants used in each mixture are
given in Table 4.24. The samples were ignited under microwave power of 800 watts
operating at 100% duty cycle.
The temperature profile of the individual reactant heated by MH is illustrated in
Figure 4.41. It can be seen that material with moderate electrical conductivity (C) heated
more effectively than either insulating (Ti02 and AI2 O 3 ) or highly conductive (Al) material.
As seen in Figure 4.41, carbon, when exposed to microwaves, reached temperatures of
1000°C in 3 minutes, while no temperature above 500°C was recorded for Al powder. Low
dielectric losses and low thermal conductivity samples,TiCh and AI2 O 3 , were difficult to
heat from room temperature, even though microwave penetration was significant.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
Table 4.24 The amounts of reactants used in each mixtures for four compositions studied.
Composition 4
3TiCh+3C+(4+4.74)AH-l.2SAlA -»
3TiC+3 .25A1A+4.74A1
3T iA
(g)
3C
(g)
(4+4.74)Al
(g)
1.25A1A
(g)
Total
weight (g)
3TiO:+3C+<4+4.74)Ai+l.25AlA
2.62
0.40
2.58
1.40
7.00
3TiO:+3C+(4+4.74)AJ
2.62
0.40
2.58
3TiOi+3C+1.25AIA
2.62
0.40
3Ti0;+(4+4.74)AH-I.23Al;0 ,
2.62
0.40
3C+<4+4.74)AI+l 25A120 3
3TiOj+3C
2.62
3TiOj +(4+4.74)Al
2.62
3T A + 1.25A 1A
2.62
1.40
4.42
2.58
1.40
6.60
2.58
1.40
4.38
0.40
3.02
2.58
5.20
1.40
3C+(4+4.74)AI
0.40
3C+1.25A1A
0.40
2.58
2.58
(4+4.74)Al+l.25A1A
3TiO,
5.60
2.98
1.40
1.80
1.40
3.98
2.62
2.62
0.40
3C
4.02
0.40
2.58
(4+4.74)Al
L.25A1A
2.58
1.40
1.40
Composition 3
3TiO-.+3C+4Al+1.25AlA -»
3TiC+3.25AIA
3TA
(g)
3C
(g)
4A1
(g)
1.25A1A
(g)
Total
weight (g)
3TA+3C+4AH-1.25A1A
2.62
0.40
1.18
1.40
5.60
3TA+3C+4A1
2.62
0.40
1.18
3Ti02+3C+1.25Al20 3
162
0.40
3TA+4A1+1.25A1A
2.62
0.40
3C+4A1+1.25A120 3
3Ti02+3C
2.62
3TiOj +4A1
162
3T A +1.25A 1A
2.62
0.40
1.40
5.20
1.18
1.40
2.98
3.02
3.80
1.18
1.18
4A1+U5A120 j
1.25A1A
1.18
1.40
3C+1.25A1A
4A1
4.42
1.18
0.40
3C
1.40
0.40
3C+4A1
3Ti02
4.20
4.02
1.58
1.40
1.80
1.40
2.58
2.62
2.62
0.40
0.40
1.18
1.18
1.40
1.40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
Table 4.24 (cont.)
Composition 2
3Ti02 +3C+(4+4.74)Al -*
3TiC+2AI2 O3 +4 .74A1
3TiCh
(g)
3C
(g)
(4+4.74)Al
(g)
Total
weight (g)
3Ti0 2+3C+(4+4.74)Al
2.62
0.40
2.58
5.60
3Ti0 2+3C
2.62
0.40
3Ti02 +(4+4.74)Al
2.62
0.40
3C+(4+4.74)Al
3Ti02
3.02
2.58
5.20
2.58
2.98
2.62
2.62
0.40
3C
(4+4.74)Al
0.40
2.58
2.58
3TiCh
(g)
3C
(g)
4A1
(g)
Total
weight (g)
3Ti02 +3C+4Al
2.62
0.40
1.18
4.20
3Ti0 2+3C
2.62
0.40
3Ti0 2 +4Al
2.62
Composition 1
3Ti02+3C+4Al
3TiC+2Al2 0 3
3.02
1.18
3.80
3C+4A1
0.40
1.18
1.58
3C+4AI
0.40
1.18
1.58
3TiQ>
3C
4A1
2.62
2.62
0.40
0.40
1.18
1.18
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
167
1000
>x—x
900 800
—*— excess Al
AUO,
700 i
600 -i
|
500 i
I
^
!
400
-j
300
100 -i
0
20
T
T
T
40
60
80
100
120
T
T
140
160
180
Time (s)
Figure 4.41 Temperature profile of the reactants heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
168
It was observed that the excess Al sample reached lower temperature after 3 minutes
as compared to the stoichiometric Al sample. The stoichiometric Al sample was more
difficult to heat initially, but then the heating rate increased after 40 seconds and higher
temperature was reached. In addition, more arcing was noticeable in the excess Al sample.
This effect was attributed to high reflection from the sample.
There are two main physical loss mechanisms by which microwaves interact with
materials, resulting in internal heating. These are conduction losses in non-insulating
materials and polarization losses in polar dielectric insulators. The former is due to electrical
conduction and the latter is due to dielectric properties.
Electrical conduction plays a key role in the microwave heating of carbon and
aluminum. Heating of carbon and aluminum particles takes place via the joule effect.
Microwave induces electrical current in the particles resulting in an ohmic type of loss
mechanism. The capability of microwaves to dissipate energy into the volume of a material
is dependent on the depth of penetration of the incident radiation. Microwaves penetrate
conducting materials to a depth known as skin depth (§) where the electric field falls to 1/e
of its value at the surface. Based on the material’s conductivity, a skin depth is calculated
according to formula given in equation 2.38.
Conductivity of 103 £2'lcm'1, which is the case with carbon at room temperature,
results in 8 = 32 pm for 2.45 GHz microwave. This depth of penetration is higher order than
the size of carbon powders used (0.075 pm), thus completing heating of carbon particles is
obtained. As shown in Figure 4.41, carbon is heated more effectively than other materials.
Aluminum does not heat as well as carbon because electric field cannot penetrate
much below its surface. Very little microwave energy absorbs in the aluminum particles, and
the remaining energy reflects from its surface. The penetration depth at 2.45 GHz of die
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
aluminum with the electrical conductivity of 34.3 xlO6 O' 1cm' 1 is 1.66 pm. In this case, the
penetration depth is less than the size of aluminum used
(1 0
pm), thus only a skin of the
aluminum powders get heated. The heat is propagated to the rest of the particle by
conventional heat transfer.
In electrical conduction, as the charge particles move (e.g. free electron), a current is
induced. The electrical losses due to the ohmic losses occur when charge particles move and
collide with each other. The situation is quite different in dielectric materials (AI2O 3 , TiOi).
Instead of the motion of electrons, electric dipoles play a dominant role in the microwave
heating of the material. The origin of the heating effect lies in the inability of the
polarization to follow the extremely rapid reversals of the electric field. According to Willert
(Wil93b, Wil95), space charge polarization are responsible for microwave loss in AI2 O3 and
Ti02.
For dielectric materials, a penetration depth is calculated from tanS and the dielectric
constant according to a formula given in equation 2.37. Using equation 2.37 and the data
from [Bat95], the depth of penetration for AI2O3 at room temperature is approximately 310
cm and decreases to 160 cm at 600°C. Using data from [Her85, Von54], the depth of
penetration for TiCh is approximately 64.93 cm at room temperature and decreases to 1.43
cm at 600°C.
Even though microwave penetration of TiCh and AI2O3 are significant, they are
difficult to heat from room temperature as seen in Figure 4.41. Because of the low dielectric
loss factor of the materials, only a small amount of microwave energy is absorbed according
to the equation relating dielectric properties to the power absorbed of the material (equation
2.34). In addition, it can be extracted from equation 2.34 that TiC>2 which has higher tan5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
(0.003) and s”eff (0.3) heats more effectively than AI2O3 with a lower tanS (0.002) and e”eff
( 0 .02 ).
When considering the amount of microwave power absorbed from the electrical
conductivity of the material, insulating materials, AI2 O3 (a *
1 0 * 16
fT’cm'1) and T1O2 (o *
10 12 Cy'cm’1) heat less effectively than C with moderate conductivity (c * 103 Q'Icm1) and
highly conductive Al (c * 34.3xl06 Q'lcm''). However, only a skin of Al powders are
heated by microwaves as compared to volumetric heating of C particles.
From these results, therefore, it is likely that when combustible mixture including C,
Al, TiC>2 , and AI2 O3 are heated by microwaves, electrical conduction losses (skin effects: C,
Al) should dominate polarization losses (space charge effects : HO 2 , AI2 O3 ). Several
investigators already reported the beneficial effect of metal dispersion on heating low
dielectric loss materials [Sut92]. A homogeneous dispersion of metal particles offers a
preheating-effect in a mixture [Wil93b].
The heating behavior of individual reactants by MHH also was carried out. One
important issue in this case was whether microwaves would penetrate through the susteptor
and impinge on the sample. If the depth of penetration of the susceptor was too small, then
the susceptor would absorbed all the microwave energy and the sample would be heated
only by conventional heat transfer from the susceptor.
The depths of penetration at 2.4S GHz of the 40%SiC-alumina cement susceptor
used in this study was estimated to be 26.7 cm at room temperature. It dropped to 7 cm at
600°C and 4.8 cm at 1200°C. The results were generated from equation 2.37. An estimate of
the dielectric properties (tanS, e’r) of this susceptor was determined using the dielectric data
for the alumina cement [Cozz96], the dielectric data for 100%SiC [Bat95], and the
Maxwell’s expression for mixture [Mee87b].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
(4.1)
where K'c is the composite dielectric property of interest (tan8, s ’r). Vm and Vd are volume
fraction of the matrix phase (alfrax 66 alumina cement) and of the dispersed phase (SiC),
respectively. K'm and K ’d are matrix and dispersed phase dielectric property of interest,
respectively.
The depth of penetration at 2.4S GHz for this susceptor were between 7 and 4.8 cm
in the 600-1200°C temperature range. The thickness of the susceptor is about 0.6 cm.
Therefore, 40%SiC-alumina cement susceptor attenuated microwaves but did not absorb all
the incident energy. The microwave power that penetrated the susceptor and was available
to heat the sample was determined by
P = l00Po 1—(1 ——X-^r—)
e D„
(4.2)
where P0 is power density at sample surface (incident microwave power of 800w), t3 is the
thickness of the susceptor (0.6 cm), and Dp is the depth of penetration.
The power absorbed by the susceptor used in this study was estimated to be 1.4% of
the incident microwave power at room temperature. It increased to S.4% at 600°C and 8% at
1200°C. This meant that 98.6% of the 800w incident microwave power, which corresponded
to approximately 790 watts, was available for direct microwave interaction with the sample
at room temperature. The microwave power available to interact with the sample decreased
slightly to 94.6% (~760w) at 600°C and 92% (~740w) at 1200°C. With these amounts of
microwave power available to impinge on the sample, it was believed that the powder
samples heated inside this susceptor were not shield from microwave energy during heating.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
Figure 4.42 shows the heating behavior of the reactants heated by MHH. It can be
seen that carbon which was easily heated by microwaves reached a temperature of 760°C as
compared to 1000°C when heating with MH. A similar trend was observed for the
stoichiometric Al samples which are considered moderately heated by microwaves. For
TiOi and AI2 O 3 , a different trend was evident. The samples reached a higher temperature
than when heating under MH. This same behavior was observed for the excess Al sample.
The decrease in maximum temperature with MHH for C and Al can be explained by
the fact that the susceptor used also exhibited strong microwave energy absorption at 2.45
GHz. It partially shielded die sample from the microwave field. The competition between
susceptor and C or Al to absorb microwave radiation thus resulted in less efficient heating of
the sample.
On the other hand, the susceptor helped in the heating of poorly microwave
absorbing materials, Ti02 and AI2 O 3 , and also the excess Al sample which heated very
slowly by microwave energy alone. The susceptor absorbed some of the incident microwave
radiation and converted it to heat for preheating the sample. At the upper temperature where
the sample heated to its critical temperature and absorbed microwave efficiendy, there were
two sources for heating the sample. These included heating from microwave absorption and
from the susceptor. Thus an increase in die maximum temperature when heating these
samples under MHH was observed.
The temperature profile of all combinations for the stoichiometric system heated by
MH is illustrated in Figure 4.43. The only samples that were ignited in these systems were
the mixtures of 3Ti02+3C+4Al and 3C+4A1. X-ray diffraction analysis (Figure 4.44)
showed the presence of TiC and AI2 O 3 for the former mixture. The main compound
produced for the latter was AI4 C3 with some Al remaining.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
173
10 0 0
TiO
* — excess Al
800
AI2O3
* —-Susceptor
700
O
O
8?
3
0
QE
0)
H
600
t
500 i
i
400 J
300
200
100
-
0 T
0
T
20
40
60
80
T
100
120
140
160
180
Time (s)
Figure 4.42 Temperature profile of the reactants heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
174
2000
1800
1600
1400
0
— 1200
2
3
1
1000
E
u
H
800
600
400
200
0
20
40
60
80
100
120
140
160
180
Time (s)
Figure 4.43 Temperature profile of all combination mixtures for the stoichiometric
composition heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
175
Relative Intensity
Jfcd La
3C+4AI
10
20
30
40
50
60
70
80
2 Theta [deg]
1- T iO i 2 - C , 3 - A 1 . 4 - AUCj, 5 - TijOj, 6 - Al3Ti, 7- TiC, 8 - AI2O3
Figure 4.44 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by MH.
R e p ro d u c e d with permission o f the copyright owner. Further reproduction prohibited without permission.
90
176
It was also observed from Figure 4.43 that the addition of Al into TiOz did not seem
to help in raising temperature of the mixture, while the addition of C into Ti0 2 behaved
otherwise. However, XRD analysis for the mixture of 3TiOz+4Al (Figure 4.44) showed the
presence of not only large amount of the reactants Al and TiOz but also small traces ofTuOs
and AI3 TL The AlzC>3 phase was not detected. The presence of Ti2Os indicated that the
deoxidation reaction of TiOz by Al occurred. Also part of the liquid Al reacted with the
reduced Ti to form titanium aluminide, AhTi.
In the case of 3TiOz+3C, only the reactants, C and Ti02 were detected. Even though
the heat provided from C to TiOz helped raise the temperature of the mixture, it did not
result in producing any product.
From these results, it was thought that in order for the 3Ti02+3C+4Al mixture to
ignite heating of both C and Al was required. Firstly, carbon absorbed some of the
microwave radiation and converted it into heat used for wanning the whole sample. Then,
the conductive heating from carbon along with the coupling of microwave radiation with Al
itself heated the surface of Al. Figure 4.45 is a graph of free energy (AG) versus temperature
for some possible reactions which can occur between the reactants of Ti02, C, and Al. It can
be seen that the reaction between TiOz and Al was the most exothermic and was likely to
occur. Thus, once Al was heated efficiently by C, Ti0 2 would react with liquid Al. The
aluminothermic reduction of Ti0 2 then occurred which resulted in the formation of Ti and
Al20 3 . The Ti then reacted with the free carbon to produce TiC. The final TiC and Al20 3
products were evident in XRD analysis of 3Ti02/3C/4Al mixture.
It was thought that there were two main heat sources contributed to the sample
during the aluminothermic reduction of TiOz. These included conductive heating from C and
heating from microwave absorption of TixOzx-i.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
700
□ Ti02 + 2C —
600
A T‘i02+3C —
500 -
O 4/3Al ♦ C---- :
400 ■
i=
• Ti02 ♦ 4/3 Al
300200
-100
■■
-200
-
-300
0
500
1000
1500
2000
2500
Tamparatur* (K)
Figure 4.4S AG vs. temperature for some possible reactions between three reactants. Plotted
using thermodynamic values from [Kub79].
At this stage, heating source from Al is less significant as compared to that from C.
This is because the electrical conductivity of Al decreases with temperature and thus
microwave absorption of Al decreases. Unlike Al, the electrical conductivity of carbon
increases significantly with temperature and thus promotes heating.
Rutile TiC>2 is a microwave transparent material (tan5<0.01) with low electrical
conductivity (~ 10'12 Q‘lcm'1). The reduction of TiCfe to TixC>2x-i introduces defects in the
form of oxygen vacancies in the oxide and increases its electrical conductivity [Bro91]. The
incorporation of oxygen vacancies into crystal structure of TiCh causes a conduction loss
mechanism for microwave absorption [Kat91]. This type of loss mechanism occurs when
vacancies migrate to align themselves with the electric field. In addition, as seen in equation
2.31 and 2.33, as electrical conductivity increases the electric loss factor (s’W) and loss
tangent (tanS) increase. This changes the microwave properties of the oxide. As electrical
conductivity of TiOix-i increases above 1 Q 'cm'1, its loss tangent rises significantly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
[Von54]. Microwave absorption of TiOix-i thus increases and then it becomes a microwave
absorbing material.
Figure 4.46 illustrate the temperature profile of all combinations for the
stoichiometric system heated by MHH. The only samples that were ignited in these systems
were the mixtures of 3TiC>2+3C+4Al and 3C+4A1. However, the ignitions of these mixtures
occurred at longer time (40 seconds) than the ones in the MH case. This was because of
strong microwave absorption at 2.4S GHz by the susceptor.
It was observed that the combustion temperature of 3C+4A1 mixture under MHH
was higher than that observed under MH, while for the 3Ti02+3C+4Al mixture, the
combustion temperature was lowered. As discussed earlier, the heating of C and Al was less
efficient under MHH because of the competition to absorb microwaves between these
materials and susceptor. However, once the susceptor had already been heated by
microwaves, the conductive heat from susceptor could turn into an extra heat source for C
and Al. This resulted in the increase in combustion temperature. When the poorly absorbing
microwave material, TiCh, is added to this mixture of C+Al, heat from susceptor must be
shared to help heating TiC>2 in addition to C and Al. The reaction that occurred in the
TiC>2+C+Al mixture thus took longer time and reached a lower temperature.
The XRD analysis of all combinations for stoichiometric system heated by MHH
(Figure 4.47) showed results similar to the ones heated by MH. Thus it was thought that the
sequential reaction of Ti02 +C +A1 under MHH proceeded in the same fashion as that for
MH.
The temperature profiles of all combinations for die 3Ti02+3C+8.74Al system
heated by MH and MHH are illustrated in Figure 4.48 and Figure 4.49, respectively. When
excess Al was added to the stoichiometric composition, the heat generation rate decreased
which decreased combustion temperature. In addition, as shown previously in Figure 4.41,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2000 ~
1800
1600 J
1400 -
oo
£
3
4AI
Susceptor
1200 -
600 -•
400
200
'
0
20
40
60
80
100
120
140
160
180
Time (s)
Figure 4.46 Temperature profile of all combinations mixtures of the stoichiometric
composition heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
180
>t
55
§
c|
s
3C+4AI
♦
I©
(t
10
20
30
40
50
60
70
80
90
2 Tlieta [deg]
1- TiOj, 2 - C . 3 - M 4 - AUCj, 5 - TijO* 6 - AljTi, 7- TiC, 8 - AljOj
Figure 4.47 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
2000
1800
1600
1400
8.74AI(excess Al)
o
«
3
2
E
£
1200
1000
800
600
400
200
0
20
40
60
80
100
120
140
160
180
Time (s)
Figure 4.48 Temperature profile of all combinations mixtures of the composition with
excess Al heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
182
3Ti02+3C+8.74AI
■3Ti02+8.74AI
•3C+8.74AI
1600 i
•3Ti02
■3C
■8.74AI(excess Al)
■Susceptor
600 1
0 "
0
20
40
60
80
100
120
140
160
180
Time (s)
Figure 4.49 Temperature profile of all combinations mixtures of the composition with
excess Al heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
183
heating excess Al sample under MH was more difficult because of high reflection. More
arcing was also noticeable in the excess Al sample.
In this case, as the amount of Al increases, the Al particles come closer to each other.
The probability of contact between Al particles increases and the area exposed to
microwaves diminishes. The clusters of Al particles are equivalent to bigger particles of Al
particles with lower surface area and thus microwave losses decrease. Bescher et al. [Bes91]
reported that at equal volume fraction of metal in a mixture, die total exposed surface area of
a sample containing a large particles is smaller than that of a sample containing small
particles. The volume of metal affected by skin currents is therefore smaller for the large
metal particles mixture.
The decrease in combustion temperature with increasing amounts of Al to the
3
Ti0 2 +3 C+8 .7 4 Al and 3C+8.74A1 mixtures under MH are shown in Figure 4.48. This
resulted in the increase in ignition time of these mixtures as compared to the stoichiometric
system.
The increased in ignition time and decreased in combustion temperature as the
amount of Al increased was also observed in the 3 Ti0 2 +3 C+8 .7 4 Al system under MHH
(Figure 4.49). In addition, because of strong microwave absorption at 2.4S GHz of the
susceptor, the ignition of these mixtures occurred at longer times than those ignited by MH.
It was also observed that a large amount of Al was easier to heat under MHH thus helping to
raisethe temperature of the 3 Ti0 2 +8 .7 4 Al and C+8.74A1 mixtures.
The temperature profiles of all combinations for the composition with excess AI2 O3
(3 Ti0 2 +3 C+4 Al+1 .2 5 Al2 0 3 ) heated by MH are illustrated in Figure 4.50 and 4.51. It can be
seen that the addition of poorly absorbing microwave material,
20
%Al2O3 , into the
stoichiometric composition increased the ignition time of the 3 Ti0 2 +3 C+4 Al and 3C+4AI
mixtures but did not affect their combustion temperatures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
2000
■a— 3TiOj+3C+4AI+1 .25A1j03
1800 4
a — 3TiO,
1600 i
4AI
1400 -
1200-
— o--3TiO,+1.25AI,0.
O
O
£3
2
8.
E
• — 3C+4A1
1000
-
- -x- - 3C+1 .25A1j0 3
—
800
(.--4AI+1.25AI203
•
-
600 -
3TiO,+3C»4Al
- 3Ti02+3C+1.25AIJ03
— *— 3TiOj+4AI+1,25AI20 3
- 3C+4A1+1 25AIj0 3
400 -
a a t.r.
it
2
200 4
0
20
40
60
80
100
120
140
160 180
200
Time (s)
Figure 4.50 Temperature profile of all combinations mixtures of the composition with
excess AI2 O3 heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185
500
3TiO.
450 -
400-
350 -
O
300
£
250
- - 0 - - 3 T iO2+1.25AIj O3
,
=?
200
- -x- - 3C+1.25ALO,
—
- 4AI+1 ^25AIj 0 3
- 3Ti02+3C+1 25A1j0 3
150-
-3TiO,+4AI+1.25Ai,0.
100
-
50-
0
20
40
60
80
100
120
140
160
180
200
Time ($)
Figure 4.51 Temperature profile of some combinations mixtures of the composition with
excess AI2O3 heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Using small amounts and small sizes of AI2O3 particles, the combustion temperature
of the 3TiC>2+3C+4Al or 3C+4A1 mixture did not change. In addition, it was observed that
the heating rate of single and two component systems decreased with the addition of
20%Al2C>3 under microwave heating. It is seen from Figure 4.51 that the temperature profile
of 3Ti02+3C+1.25Al203 lies above that of 3 C+I.2 5 AI2O3 . This implies that the microwave
absorption of TiC>2 particles along with the heat conduction from C particles helped increase
the heating of
3
C+I.2 5 AI2 O3 . In addition, it can be seen that the heating rate of
3Ti02+3C+1.25Al203 lies below that of 3TiC>2+3C. This was because the heat that occurred
in the system due to the microwave energy absorption of both C and Ti0 2 was used to
preheat the AI2 O3 particles, and thus the heating rate of the 3TiOa+3C decreased. It was also
observed that AI2O3 particles were heated by microwave to some extent. When either TiO?
or Al was mixed with AI2O3 , the heating rate of the mixture decreased. This did not occurr
when C was added into AI2 O3 powders. These results are in agreement with the previous
results where C particles were easily heated by microwave, while Al and low dielectric loss
Ti(> 2 particles were more difficult to heat
The temperature profile of all combinations for the composition with excess AI2 O 3
(3Ti02+3C+4Al+1.25Al203) heated by MHH are illustrated in Figure 4.52 and 4.53. It was
also seen that the addition of AI2 O 3 into the 3Ti02+3C+4Al and 3C+4A1 mixture increased
their ignition times. In addition, the 3Ti02+3C+4Al+1.25Al203 mixtures were ignited at
longer times and thus lower combustion temperatures as compared to those heated by MH.
The mixtures that contained AI2 O 3 also reached higher temperatures after 3 minutes as
compared to those heating by MH. This was due to the higher heating rate of AI2O 3 under
MHH as compared to MH, as discussed earlier. However their temperature profiles still
followed the same trends as the mixtures being heated by MH. The addition of AI2O 3
decreased the heating rate of the mixtures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
2000 n
1800
-a— 3Ti02+3C+4AJ+1.25/^03
1600
-a — 3Ti02
!,
3C
1400
-h— 4AI
Temperature (°C)
—
1.25AljO,
3Ti02+3C
1200
-a— 3Ti02*4AJ
-o- - 3Ti02-*-1.25A1j0 3
1000
- • — 3 0 4 Al
-X--3C+U5AIJO,
-+--4A1+125AJ203
-a — 3Ti02+-3O4AI
- 3Ti02+3O1,25A120 3
- 3Ti02+4AM ^SAIjOj
- 3 0 4 A I+ 1 .2 5 ^ 0 3
- a — Susceptor
0
20
40
60
80
100
120
140
160
180
200
Time (s)
Figure 4.52 Temperature profile of all combinations mixtures of the composition with
excess AI2O3 heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
450
■e—
3TiO-
4 00-
350-
4AI
’+/
— 1.25AI,0O 300-
2 250- - o - - 3 T i 0 2+1.25Al203
200
-
--X--3C+1.25AUO.
— (.--4AI+1.25AUO.
150 i
- 3Ti02+3C+1.25AUO-
100
- 3Ti05+4AI+1.25AUO-
-
-6—Susceptor
50 -
0
20
40
60
80
100
120
140
160
180
200
Time (s)
Figure 4.53 Temperature profile of some combinations mixtures of the composition with
excess AI2O 3 heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
The temperature profile of all combinations for die composition with excess of both
Al and AI2O3 (3Ti02+3C+8.74Al+1.25Al203) heated by MH and MHH are illustrated in
Figure 4.54 -4.55 and Figure 4.56-4.57, respectively.
The addition of both Al and AI2O 3 into the stoichiometric composition resulted in
the longest ignition time of the sample as compared to other cases discussed earlier. When
heating these mixtures using MHH, their ignition times also increased. The presence of
AI2O3 in the mixtures with excess Al (3TiC>2+3C+8.74Al) resulted in an increase in the
heating rates of the mixtures. For example, as discussed earlier, the sample with excess of Al
only (3TiC>2+8.74Al) reached a lower temperature than the stoichiometric sample
(3Ti02+4Al) after heating for 3minutes. The addition of AI2 O3 into this excess Al sample
resulted in an increase of the maximum temperature of the sample. However this increased
temperature still was not as high as die temperature of the stoichiometric sample. That is, the
temperature of 3Ti02+8.74Al (257°C) < 3Ti02+8.74Al+1.25Al20 3 (288°C) < 3Ti02+4Al
(320°C). This was most likely because the distribution of the excess Al was better and more
uniform with the presence of AI2O 3 in sample. The contact between Al particles was reduced
and so was the arcing. Thus the beating efficiency by microwaves of this sample
(3TiC>2+8.74Al+1.25Al203) was increased. These results imply that the addition of AI2 O3
particles, which resulted in a decrease in heating rate and increase in ignition time of the
sample, also had an advantage.
4 .2.2 Conventional Heating
The mechanism governing the combustion reaction of AhCVTiC powders under
conventional heating was investigated in order to compare with those obtained under the
influence of microwave heating.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-a— 3TiOj+3C+
8.74AI+1.25AI20 3
-e—STiOj
3C
— i— 8.74AI
Temperature (°C)
1.25AIj 0 3
—
3TiOs+3C
—
3T102+8.74AI
--O --3T102+1.25A 1203
■ - 3C»8.74AI
- -x- - 3C+1 25AI20 3
— I-. - 8.74AI+1^SAljOj
— • — 3Ti02+3C+8.74AJ
_
_ 3Ti02+3C+1 25AI20 3
— *— 3Ti02+
8.74 AI+1.25AI20
3
AI+
/
--m-- 3C+8.74AI+125A1j03
0
20
40
60
80
100 120 140 160 180 200
Time (s)
Figure 4.54 Temperature profile of all combinations mixtures of the composition with
excess Al and AI2O3 heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191
500 i
-o—3Ti02
450-
3C
400
-h— 8.74AI
. AI2 O3
1 25
350 ■
-•— 3TiOj+3C
H
£
3
15
2.
E
a
H
300
3Ti02+8.74AI
250-
-o- - 3Ti02+1.25AI20 3
/ / /
/
/-+•
-x- - 3C+1 25AI20 3
200
-+ - - S.74AI+1,25AI20 3
150 -
100
- 3Ti02+3C+1.25AI20 3
-31102+
■
8.74AI+1.25A1203
50 J
0 +0
I
20
40
60
' I '
80
100
120
140
160
180
200
Time (s)
Figure 4.55 Temperature profile of some combinations mixtures of the composition with
excess Al and AI2 O 3 heated by MH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2000 i
■a— 3Ti02+3C+
8.74AI+1.25AUO.
■o— 3TiO.
1800 i
1600-
8.74A1
Temperature (°C)
1400■*— 3TiO,+8.74AI
1200
-
1000
-
- - o — 3TiO,+1.25AJ,0.
3C+8.74AI
--X--3C+1.25AUO.
—
- 8.74AI+1.25AI,0
800 ♦ — 3Ti02+3C+
8.74AI
600 -
-3TiO,+3C+1.25AUO.
-3 T i0 2+
8.74AI+1.25AI,0.
400 -
- 3C+8.74AJ+
1.25AI?0 3
200
■a— Susceptor
-
0
20
40
60
80 100 120 140 160 180 200 220 240 260
Time(s)
Figure 4.56 Temperature profile of all combinations mixtures of the composition with
excess Al and AI2 O3 heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
500 -
450 ■!
400 •
350 ■
3TiO-
O
O 300 £
2
2
S.
E
®
I-
3C
8.74AI
250-
200
-
- - o - - 3 T i 0 2+1.25Al203
150 - -x- *
100
,25AI20 3
—►
— 8.74AI+1.25AUO,
-
♦
3Ti02+3C+125AI20 3
-3T i0 +8.74AI+
1.25A120 3
2
50 -
—Susceptor
0
20
40
60
80
100
120
140
160
180
200
Time (s)
Figure 4.57 T e m p e r a tu r e profile of some combinations mixtures of the composition with
excess Al and AI2 O3 heated by MHH.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
The combustion reaction mechanism under conventional heating was conducted by
carrying out DTA experiments. The compositions studied were the same as that investigated
under microwave heating.
Figure 4.58 shows the DTA results for all the stoichiometric mixtures. No endotherm
or exotherm was observed in the mixture of 3Ti+3C. For other mixtures, the first endotherm
encountered during the heating process was about 670°C, which was related to the melting
of aluminum. Although the shape of the DTA curves of 3 Ti0 2 +4 Al and 3 TiC>2 +3 C+4 Al
were very similar,
3
TiC>2+3 C+4 Al exhibited a higher exothermic temperature. The
exothermic peak on heating to higher temperature of the 3 Ti0 2 +4 Al mixture was observed
near 900°C, while the exothermic peak above 1000 °C was observed in the 3 Ti0 2 +3 C+4 Al
mixture. The DTA trace of 4A1+3C had a more pronounced exothermic peak that also
occurred at a higher temperature than the other two. This exotherm was relating to the
formation of AL4C3 . In addition, as seen in Figure 4.45, at the temperature about 900°C, the
reaction between titania and aluminum is the most exothermic and is likely to occur. The
aluminum is present in the liquid phase at this temperature and is likely to be most reactive.
Therefore, on heating a complete mixture of 3 TiC>2 +3 C+4 Al, the reaction between Ti0 2 and
Al is the first to proceed.
An initial endotherm which was observed near 670°C was due to the melting of the
aluminum phase. On heating to higher temperature, the exotherm corresponded to
aluminothermic reduction of Ti0 2 was observed. The reaction products of the 3 TiC>2 +4 Al
reaction consisted of titanium and alumina. The titanium then reacted with the carbon to
form TiC. This was evident from the exothermic peak above 1000°C in the DTA trace of
3
Ti0 2 +3 C+4 Al. The aluminothermic reduction of Ti0 2 occurred at about 900°C which was
higher than the melting point of Al (660.4°C). This was thought to be due to the wettability
of Al melt on Ti0 2 . The wetting angle of liquid Al-solid C and liquid Si-solid TiC>2 systems
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
is reported to be approximately 157° and 107°, respectively [Duf8 8 ]. The surface energy of
liquid Si is similar to that of liquid Al at about 825 dyn cm' 1 under an argon atmosphere
[Lyn74], and thus it is possible that the wetting angle of liquid Al-solid TiO: is near 107°
[Cho95]. It is believed that the poor wettability of Al melt on TiO; should be one of the
reasons why the aluminothermic reduction is initiated at temperature higher than melting of
Al.
Similar thermal effects were observed in the composition with the excess Al
(3 Ti0 2 +3 C+8 .7 4 Al) mixture as shown in Figure 4.59. The only difference observed was that
the endotherms in this case were stronger corresponding to the higher amounts of Al. The
DTA curves of the mixture with excess AI2 O3 (3 Ti0 2 +3 C+4 Al+1 .2 5 Al2 0 3 ) and mixture
with excess of both Al and AI2 O3 (3 Ti0 2 +3 C+8 .7 4 Al+l.2 5 Al2 0 3 ) systems are shown in
Figure 4.60-4.61. Both the endothermic and exothermic peak intensities were decreased in
comparison to those discussed earlier. This was because AI2O 3 acted as a diluent thus
reducing the intensity of the reaction.
XRD analysis for the four stoichiometric mixtures including 3TiC+3C, 3C+4A1,
3
TiC>2 +4 Al, and 3 TiC>2 +3 C+4 Al are shown in Figure 4.62. The main compound produced
from the 3C+4A1 mixture was AI4 C3 with some Al remaining. This result was similar to that
obtained from microwave heating (MH). In the case of the 3 Ti0 2 +3 C mixture, C and Ti0 2
were detected along with some traces of the Ti9 0 n, TigOis and TiO. XRD analysis for the
mixture of 3 Ti0 2 +4 Al showed the presence of not only large amounts of the products AI2O3
but also the reactants Al and also a small trace of TiO. The presence of TiO indicated that
the deoxidation reaction of Ti0 2 by Al had occurred. The liquid Al then reacted with the
oxygen to form AI2 O3 . This result was different from that obtained from MH. The AI2O3
phase was not detected, but only reactants Al and Ti0 2 along with small trace of compound
Ti3 0 s and A^Ti were observed for this mixture under microwave heating.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196
Endothermic
0
200
400
600
800
1000
1200
Temperature (°C)
Figure 4.58 The DTA results for all mixtures of the stoichiometric composition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1400
197
Exothermic
>>
Endothermic
0
200
400
600
800
1000
3TiOj+3C*8.74AI
1200
Temperature (°C)
Figure 4.59 The DTA results for all mixtures of the composition with excess Al.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1400
198
Exothermic
“► 3TiO;
Endothermic
3Ti0,+3C + 1.2SAI,0,
3TiO,*4AI * 1 .2 5 ^ 0 .
3TiO,+3C ♦ 4AI
3TiO,»3C+4AI ♦ 1,2SA^0;
0
200
400
600
800
1000
1200
1400
Temperature (°C)
Figure 4.60 The DTA results for all mixtures of the composition with excess AI2O3 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
199
i
i
Exothermic
Endothermic
3TiO]+3C ♦ 1.25X1,0,
3C+6.74AI +1 2SAljO;
"► 3TiOj+«.74AI ♦ 1.25X1,0,
"► 3TiOj+3C+fl.74AI + 1.25XIA
0
200
400
600
800
1000
1200
1400
Temperature (°C)
Figure 4.61 The DTA results for all mixtures of the composition with excess both Al and
A120 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 00
.2?
55
Ic
3C+4AI
j
L mL
J L A a il
.2
o
cc
88
9 34
10
20
30
50
40
2
60
80
70
90
7Tieta[deg]
1 -T i O i 2 - C , 3 - A l , 4 -T iO , 5 - T isO n, 6 - TitOij,7 - AI4C 3 , 8 -
9
-
Ti C
Figure 4.62 X-ray diffraction patterns for all combination mixtures of the stoichiometric
composition heated by DTA.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
201
XRD of the 3 Ti0 2 +3 C+4 Al mixture showed the presence of the product TiC and
AI2O3 , and also some traces of the reactants and a minor product phase TiO. Heating this
mixture under MH resulted in the products TiC and AI2 O3 . This indicated that the reaction
tended to be more complete under microwave heating than under conventional heating. The
heating of the 3 TiC>2+3 C+4 Al disc samples in the isothermal tube furnace also resulted in
less complete conversion of the reactants when comparing between the two heating
technique, conventional heating and MH [Ato98]. The reactant Ti0 2 and a minor phase TiO
were still observed after heating for 20 minutes at 1000°C in the tube furnace. The factors
that was likely responsible for the more complete conversion of reactants under MH were
the uniform heating and the inverted temperature gradient in microwave heating
These explanations were supported by the phase composition data from XRD
analysis of 3 TiC>2 +3 C+4 Al samples which were quenched after being heated to various
temperatures. The results are presented in Table 4.25.
The results indicated that the deoxidation reaction of Ti0 2 by Al proceeded through
TisOs, Ti 2C>3 , TiO to form Ti and AI 2 O 3 . Choi et al. [Cho95] confirmed the deoxidation
process of Ti0 2 by Al from the reaction couple experiment of Ti0 2 and Al. The reaction
couple of a Ti0 2 disc and an Al disc was heated to 872°C under an Ar atmosphere. From the
concentration profiles of Ti, O, and Al on the cross section of the reaction couple, it was
observed that while the concentration of Ti slightly increased as it approached the interface,
0 slightly decreased. This confirmed the deoxidation process of Ti0 2 by Al.
Comparison of the results from Table 4.25 and from Figure 4.58 suggests that the
exothermic peak in the DTA trace of the
3
TiC>2 +3 C+4 Al mixture corresponds to a
consecutive reaction of aluminothermic reduction and TiC synthesis. Thus, these results
suggest that the sequential reaction of 3 TiC>2+3 C+4 Al proceeds in three stages: melting of
aluminum, aluminothermic reduction of Ti02, and TiC synthesis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 02
Table 4.25 Chemical Composition Analysis Obtained from XRD for the 3TiO:+3C+4Al
_____
Samples Quenched after Being Heated to Various Temperatures in DTA
Temp (°C)
Ti0 2
Al
C
Ti30 5
Ti20 3
TiO
TiAlj
AI2O3
640
M
M
M
S
-
-
-
-
680
M
M
M
s
-
-
-
-
780
M
M
M
s
m
-
-
-
880
M
M
M
m
S
m
-
S
900
M
M
M
m
S
m
m
S
1000
m
-
m
-
m
m
m
M
M
1200
-
-
-
-
-
m
-
M
M
TiC
M = major phase, S = secondary phase, m=minor phase
37702 + AAl
2Al20 3 + 377
377+3C->377C
(4.3)
(4.4)
Even though the particle size of the reactants used in this study were slightly
different from those used in the other studies [Bow95, Cho95], it was evident from the
results that the mechanistic steps involved in the combustion synthesis of Ah0 3 -TiC by
conventional heating were similar to those cited from the literature as mentioned earlier in
Chapter2. Choi et al. also suggested that the formation of AfeOs-TiC from 3TiC>2+3C+4Al
mixture was governed by TiC formation from reduced Ti and C, which was controlled by
the diffusion of carbon through the solid TiC.
4.2.3 Summary
It can be seen that, under pure microwave heating (MH), materials with moderate
conductivity (C) heated more effectively than either low dielectric loss materials (TiC>2 and
AI2 O 3 )
or highly conductive (Al) material. In addition, it was observed that the excess Al
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
samples were more difficult to heat because of high reflection from the sample which also
led to arcing. When using MHH, the opposite trends were observed. The heating rate of
carbon, which was easily heated by microwaves, and Al powders which were moderately
heated by microwaves were decreased, while the heating rate of TiCh, AI2 O 3 , and high Al
weight samples increased under MHH.
From these heating behaviors, it was thought that in order for 3TiC>2+3C+4Al
mixture to ignite heating of both C and Al was required. Carbon first absorbs some of the
microwave radiation and converts it into heat that is used for wanning the whole sample.
Then, the conductive heating from carbon along with the coupling microwave radiation by
Al itself heats the surface of Al. Once Al is heated efficiently by C, Ti02 reacts with liquid
Al. The aluminothermic reduction of Ti02 then occurrs and results in the formation of Ti
and AI2O 3 . Finally, the Ti then reacts with the free carbon to produce the final product TiC.
In conventional heating, the DTA traces and results from XRD provided evidence
that the combustion reaction is initiated by a reaction between Ti02 and Al. The reaction
proceeded in a three stage process where the aluminum phase melts, then the melting Al
reacts with the titania reducding it to Ti, and finally the titanium reacts with the free carbon
to produce the final products AI2 O 3 and TiC.
These results suggest that the reaction mechanisms during the combustion of TiCh,
C, and Al
to
form A I 2 O 3 -TiC powders using microwave heating and conventional heating
are similar. The difference is the way the reactants receive heat from the heating source.
In conventional heating, material is heated by a heat transfer process. Conventional
heating of materials is a function of the material’s thermal conductivity. Thermal
conductivity is the material constant relating the rate of heat flow to a temperature gradient.
Heat is transfered from the heat source to the surface of material by a combination of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
204
convection (low temperature) and radiation (high temperature) and then it passes to the
cooler center of the material by thermal conduction.
In contrast to conventional heating, heating by microwaves is a function of the
dielectric properties of the material in question. Depending on the microwave absorption of
the materials, heat generation by microwave heating occurs internally throughout the bulk of
the material. This internal heating may enhance reaction rates due to die fact that heating
occurs throughout the reactant mixture rather than by heat conduction from the surface to
the core of the conventionally processed material When microwaves penetrate and
propagate through a dielectric material, the internal electric field is generated within the
affected volume. This electric field induces polarization and motion of charges. The
resistance to these induced motions causes loss and attenuates the electric field. As a result,
volumetric heating occurs.
Thus, in conventional heating, all reactants received heat from the heat source
depending on their thermal conductivities. When the temperature of the sample is raised to
the melting point of Al, the reaction started at the surface of the sample to proceed in three
steps as discussed earlier. In microwave heating, carbon with moderate conductivity first
absorbs some of the microwave radiation and converts it into heat used for wanning the
whole sample. The conductive heating from carbon along with the coupling microwave
radiation by Al itself heats the surface of Al. The skin effect of Al particles under
microwave heating induces a multiple ignition inside die whole sample as shown in Figure
4.63. The heat generated by C and Al then heats Ti02 and thus the aluminothermic reduction
of TiCh by Al is initiated as with the conventional heating.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
205
a) Conventional beating
g
5
# W
W *
w *
» \A /
W *
* W
W
* \A /
*
b) Microwave heating
Microwave
Wavefront
propagation
Figure 4.63 The reaction path within a combustible mixture of Ti02, C, and Al using a)
conventional heating, b) microwave healing.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
SUMMARY AND CONCLUSIONS
The present study was undertaken to synthesize and process AI2 O 3 - TiC powders by
using a combined technology of SHS combustion synthesis and microwave heating. The
feasibility of using microwave energy at 2.45 GHz to synthesize AI2O 3 - TiC powders was
demonstrated.
The experimental designs were conducted on the synthesis of AhCVTiC powders to
investigate the effects of reaction parameters on ignition behavior and characteristics of the
resulting powders. A sixteen-run Plackett-Burman design was first performed to analyze the
significance of reaction parameters and to identify a small number of the dominant
parameters for subsequent investigation. The reaction parameters considered were the
amount of excess Al and AI2O 3 , particle size of each reactant, heating method, and
microwave power. Based on these results, the most significant factors affecting the
investigated responses were identified. These include amount of excess Al and AI2O 3 ,
particle size of C and AI2 O 3 , and also heating method.
The thirty-run central composite design (CCD) was then performed to investigate
the effect of those parameters on die responses of interest. The responses considered
included ignition time and temperature, combustion time and temperature, final density of
the powders, particle size, specific surface area, and agglomeration size.
The combustion synthesis of ALCVTiC powders using microwave heating (MH) or
microwave hybrid heating (MHH) as a heating source was achieved successfully. A clear
difference in ignition time and temperature was observed for the two heating methods. The
MHH samples required longer times to reach ignition temperatures. Thus they ignited at
206
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
207
higher temperature than MH samples and resulted in lower combustion temperature. The
results showed no significant difference in the characteristics of powders ignited by these
two heating methods.
The addition of diluents to the reactants increased ignition time and temperature,
while decreasing the combustion temperature. An increased particle size of alumina lowered
the ignition time and temperature. While for the particle size of carbon, the ignition time and
temperature of the ignited sample decreased with an increased particle size of carbon under
MHH only. Under MH, an increased particle size of carbon did not change the ignition time
and temperature significantly.
It was also observed that the sample with the largest alumina or carbon particle size
had the highest combustion temperature. It was also found that the density of product
decreased with increasing amounts of Al. The addition of excess AI2 O 3 to the reactants
resulted in decreasing the agglomeration size.
The investigated responses also were influenced by the interactions between these
parameters. In some cases, the interaction effects were significant, but in no case were the
interactive effect as large as for individual parameters. The interaction between the amount
o f
excess
A I2 O 3
and other parameters was found to have more effect on the responses than
any other combinations of the interaction.
Various empirical models relating these important reaction parameters and their
interactions to the investigated responses were developed. This provided the infonnation
required for adjusting the setting of the important factors in order to obtain the desired
products.
The key to controlling the desired product configuration is to understand the
combustion mechanism. The particle-particle interactions occurring during a combustion
reaction were studied. To better understand the complex interaction of the reactants on a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
208
particle-particle level as the reaction proceeds, it was necessary to understand how each
component behaved with respect to itself and in combination with the other reactants.
Under microwave heating, material with moderate conductivity (C) heated more
effectively than either low dielectric loss materials (T 1O 2 and AI2O 3 ) or the highly
conductive Al. High A1 weight samples were more difficult to heat because of high
reflections from the sample that also resulted in arcing. The opposite trends were observed
when heating under MHH. The heating rate of carbon, which was easily heated by
microwaves, and Al powders which considered moderately heated were both decreased,
while the heating rate of Ti0 2 , AI2 O 3 , and high Al weight samples increased under MHH.
From these heating behaviors, it is suggested that in order for the 3 TiC>2+3 C+4 Al
mixture to ignite microwave heating of both C and Al is required. Carbon first absorbs some
of the microwave radiation and converts it into heat that is used for wanning the whole
sample. Then, the conductive heating from carbon along with the coupling microwave
radiation by Al itself heats the surface of Al. Once Al is heated efficiently by C, Ti02 reacts
with liquid Al. The aluminothermic reduction of Ti0 2 then occurs which results in product
of Ti and AI2O3 . The Ti then reacts with the free carbon to produce final product TiC.
The mechanism governing the combustion reaction of Al20 3 -TiC powders under
conventional heating was also investigated in order to determine whether or not the
governing reaction mechanism observed under microwave heating was the same as that
observed under conventional heating.
In conventional heating, die DTA traces and results from XRD provided evidence
that the combustion reaction is initiated by a reaction between TiCh and Al. The reaction
proceeded in a three stage process where die aluminum phase melts, and then reacts with the
Ti0 2 to produce Ti and AI2 O 3 , and finally the Ti reacted with the free C to produce TiC.
However, the conversion of reactants into these final products was less complete compared
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
209
to that obtained from microwave heating. The AI2O 3 and TiC products were detected along
with minor traces of some reactants and a TiO phase.
These results lead to the suggestion that the reaction mechanisms during the
combustion of TiCh, C, and Al to form AUOj-TiC powders using microwave heating and
conventional heating were similar. The only difference was the way the reactants received
heat from the heating source.
In conventional heating, material is heated by heat transfer process. Heat is
transferred from heat source to the surface of material by a combination of convection (low
temperature) and radiation (high temperature) and then it passes to the cooler center of the
material by thermal conduction. All reactants (TiCh, C and Al) received heat from the heat
source through heat transfer, depending on their thermal conductivities. When the
temperature of the sample is raised to the melting point of Al, the reaction proceeds in three
steps as mentioned earlier.
In contrast to conventional heating, depending on the microwave absorption of the
materials, heat generation by microwave heating occurs internally throughout the bulk of the
material. Heating by microwaves is a function of the dielectric properties of the material in
question. Carbon with moderate conductivity first absorbs some of the microwave radiation
and converts it into heat used for warming the whole sample. The conductive heating from
carbon along with the coupling microwave radiation by Al itself heats the surface of Al.
Once Al is heated, the heat generated by both C and Al then heat Ti0 2 and thus the
aluminothermic reduction of Ti0 2 by Al begins. The ignition begins from the interior of the
sample due to the heat loss from the its surface. This inverted temperature gradient results in
a combustion wavefront propagated radially outward. Microwave synthesis leads to more
complete conversion of the reactants.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
210
Based on the results of this study, microwave heating appears to be a viable
alternative to conventional heating for synthesizing AI2O3 - TiC powders.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A
HISTORY OF COMBUSTION SYNTHESIS
Year
Activity
Author
1825
Berzelius [Hla91]
ZrOj was made by heating amorphous Zr below
redness and reaction could be ignited even below room
temperature
1892
Moissan [Hla91]
Various nitrides of transition and rare earth were
prepared
1892
Vfllion [Hla91]
Exothermic reaction between Al and different gases
1895
Goldschmidt
[Hla91]
Metal or alloy of the metal were obtained by reacting
various oxide, such as titanium oxide, with aluminum
1896
Moissan [Hla91]
Self-propagating reaction between Al and S or P
1900
Fonzes-Diacon
[Hla91]
Phosphides, arsenides, silicides, borides were prepared
by simultaneously reducing two oxides by thermite
reaction
1900
Matignon [Hla91]
Exothermic reaction between Al and different gases
1902
Muthmann [Hla91]
Cerium nitride & Cerium Hydride were produced
1902
Wedekind [Hla91]
Self-propagating reaction between Al and S or P
1905
Colani[Hla91],
Matignon [Hla91]
Phosphides, arsenides, silicides, borides were prepared
by simultaneously reducing two oxides by thermite
reaction
1907
German
Scientists [Hla91]
“Frank-Caro Process” ; nitridation of calcium carbide
in self-propagation regime
1900s
Anonymous
[HIa91]
Fission of metallic uranium and plutonium and fusion
reaction between hydrogen atoms.
1940s
Alexander[Hla91]
Calcium metal ingot was converted to hydride.
Combustion synthesis of nitrides of alkakaline-earth
metals was proposed
211
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
212
1950s
Walton [Wal59]
Potential of thermite reaction for refractory coating
1953
White [Whi53]
Combustion synthesis of aluminum phosphide
1954
Titterington [Tit54]
Ignition of K:TaF7 and Na in a tubular steel container
1960
Huffadine [Huf60]
Synthesis of molybdenum disilicide from an mixture of
Mo and S powders using a water cooled tubular copper
vessel
1964
Krapf [KraP64]
“Chemical hot press”; the reacting mixture included an
inert component, reactive metals and certain oxides
1967
Stringer [Str67]
“Reaction pressing” which can be applied to both
intermetallic type and simple refractory compounds
1971
Borovinskaya
[Mer71]
“Gasless combustion”
1972
Merzhanov
[Mer72]
“ an acronym SHS”; a wide variety of materials from
abrasive powders of TiC to superconducting oxides of
YbaiCujO?
1976
Merzhanov
[Mer89]
The Institute of Structural Macrokinetics was created
in USSR to serve as the Soviet Center of SHS research
1981
Odawara [Oda81,
Oda84]
SHS reactions coupled with centrifugal motion was
used as a method to deposit corrosion-resistant coatings
on the inner wall of pipes
1982
McCauley [McC82] TiC and TiB2 were studied extensively at the US Army
lab.
1983
Henshaw [Hens83]
TiC and TiB: were studied extensively at die US Army
lab.
1984
DARPA [Gab85]
The US government sponsored combustion synthesis
research through defense advanced research project
agency (DARPA)
1984
Logan [Log84]
processing of TiC, TiB2 , TiB2 -Al2 C>3
1984
Miyamoto [Miy84]
TiC, TiB2 and Sic were synthesized and simultaneously
densified by the high pressure combustion sintering
(HCPS)
1986
Rice [Ric8 6 ]
Simultaneous reaction and densification of ceramics
and composites were studied
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
1986
Munir and Holt
[Holt8 6 ]
The solid-gas (N2) reaction systems were examined and
a theoretical model on the basis of experimental
observations and theoretical calculations was proposed.
SHS of multi-laminated metal foils of Ni and Al was
ignited
1990
Kaieda [Kai90],
Moore [Moor90]
Ni-Ti shape memory alloys were synthesized. Some
basic parameters that play important role with respect
to final product properties were found
1990
Rice [Ric90]
The composites of TiC-l0vol% Ti were produced by
the combination of hot rolling and combustion
synthesis. The product exhibited 7% open porosity and
a fracture strength of 345 GPa
1990
Clark [Dal90,
Cla91]
TiC, SiC, B 4 C , TB 2 , MoSi2 , TisSij, AIN, TiN, and
Al2 0 3 -TiC were ignited using microwave energy
1991
[Pat97]
First international symposia on SHS have been held at
Alma-Ata, Kazakhstan
1992
Moore [Fen92]
The excess of liquid metal in-situ with the combustion
synthesis reaction was used to produce composites in
order to improve the densification of composite
materials synthesized by combustion synthesis.
1992
Merzhanov
[Mer93a]
A quarterly journal “International Journal of SHS”
devoted to SHS has been published with Merzhanov as
the general editor.
1993
[Pat97]
Second international symposia on SHS have been held
at Honolulu, USA
1995
[Pat97]
Third international symposia on SHS have been held at
Wuchan, China
1996
Avakyan,
Nersesyan and
Merzhanov
[Ava96]
Electronic
engineering
materials
such
as
superconductor, ferroelectric and magnetic materials
were synthesized by SHS reaction using metal oxide
and peroxide precursors.
1997
[Pat97]
Fourth international symposia on SHS have been held
at Spain
1999
Fifth international symposia on SHS have been held at
Moscow, Russia on August 1999
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX B
CENTRAL COMPOSITE DESIGN
The central composite design is constructed following the steps given below:
i.
1. Construct a complete or fractional 2 factorial layout, where k is number of factor
investigated
2. Add 2k axial, or star, points along the coordinate axes. Each pair of star points is
denoted, using coded levels, as follow:
0,
0,
.
0
)
(0 ,
±a,
0,
.
0
)
(0 ,
0,
0,
.
(±a,
=
where “a” is a constant
± a)
and F is the number of runs in the factorial
(21evel) portion of die design (F = 2i .).It should be noted that “a” would take
value greater than
minimum
1
, which means that
±1
no longer represents the factor
and maximum.
3. Add “m” repeat observations at the design center:
(0,
0,
0
0)
4. Randomize the assignment of factor-level combinations to the experimental run
sequence
The total number of test runs (n) in a central composite design based on a complete
2k factorial is n = 2k + 2k + m ^
example of a central composite design for 4 factors of
L
interest (k = 4) (prior to randomization) is shown in Table Bl. The design consist of the 2 =
214
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
215
24 = 16, 2k = 2(4) = 8, and in this example six repeat observations (m = 6) is selected. Thus
i
i
the total test runs is 30 and the “a” value is a = F* = (2 4 )4 - 2.
As seen in Table Bl, the CCD is feasible to run 5 levels of each factor. The 5 design
level value in this case is (-a, -1, 0, 1, +a) = (-2, -1, 0, 1, +2). An easy way to resolve the
transformation the design value of (-a, - 1 , 0 , 1 , +a) to the real factor value is as follows:
1. Suppose the minimum and maximum real factor values to be investigated are 0 and 40.
The middle real factor value is (max + min)/2, which is 20. Thus thereal values for -a,
0, and a are 0,20, and 40, respectively.
2. Next, the real factor values for -1 and +1 can be computed as the real value for 0 level ±
A, which in this case is equaled to 20± A.
3. To find the value for A, the “a” value is required (in this example a = 2). Then the
following relation is formed:
maxvalue-middlevalue . . . 40 - 20 , .
-----------------------------= A; therefore, A = ---------- = 10
a
2
implying 20 + A = 30, and 20 - A = 10
4. Thus, the final real factor value for all 5 design levels are as follows:
(-a, - 1 , 0 , 1 , + a) = ( - 2 , - 1 , 0 , 1 , 2 ) = (0,10,20,30,40)
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
216
Table B1 Central Composite Design for 4 factors
Factor
Run
B
A
C
D
1
-1
-1
-1
-1
2
-1
-1
-I
+1
3
4
5
-1
-1
+1
-1
-1
-1
+1
+1
-1
+1
-1
-1
6
-1
+1
-1
+1
7
-1
+1
+1
-1
8
-1
+1
+1
+1
9
+1
-1
-1
-1
10
+1
-1
-1
+1
11
+1
-1
+1
-I
+1
+1
12
+1
-1
13
14
15
16
17
18
19
+1
+1
-I
-1
+1
+1
-1
+1
+1
+1
+1
-I
+1
+1
+1
+1
0
0
0
0
0
0
0
0
0
0
0
0
20
0
0
0
0
21
0
0
0
0
22
0
0
0
0
23
24
25
26
27
28
29
30
-2
0
0
0
+2
0
0
0
0
-2
0
0
0
+2
0
0
0
0
-2
0
0
0
+2
0
0
0
0
-2
0
0
0
+2
Note
2
=2=16
m- 6
2k =2(4) = 8
and
1
1
a = F 4 = (24)4 =2
thus ±a = ± 2
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
APPENDIX C
PARTICLE SIZE DISTRIBUTION OF THE REACTANTS AND PRODUCTS
a) The particle size distribution of all the reactants used in the sixteen-runs PlackettBurman design____________ __________________________________________
Particle size distribution (vol%)
A) Sixteen-Runs
Reactant
Ti0 2
Al
vol% of the
sample has
particle size
less than (pm)
50 vol%ofthe
sample has
particle size
less than (pm)
90 vol% of the
sample has
particle size
less than (pm)
0.9-1.6
0.70
1.60
3.70
1.0-5.0
1.45
4.88
10.29
3.0-4.5
4.44
9.37
18.00
13.61
34.58
87.90
0.49
0.30
0.76
4.05
40-45
14.94
50.23
168.80
0.017*
N/A
N/A
N/A
0.075*
N/A
N/A
N/A
Average
particle size
(pm)
-2 0 0
AI2 O 3
C
mesh
10
* = Data obtained from Cabot Corporation using TEM method
N/A = Data could not be measured using the Coulter Principle
217
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
b) The particle size distribution of all the reactants used in the thirty-runs CCD
Particle size distribution (vol%)
B) Thirty-Run
Reactant
Average
particle size
(pm)
10 vol% of the
sample has
particle size
less than (pm)
50 vol% of the
sample has
particle size
less than (pm)
90 vol% of the
sample has
particle size
less than (pm)
Ti02
2-3
0.78
2.34
5.07
Al
10-14
8.26
15.45
26.4
AI2 O3
0.49
0.30
0.76
4.05
4.70
8.44
12.90
20-38
25.33
32.57
40.79
44-53
26.14
52.40
64.40
63-74
66.44
72.95
96.50
0.017*
N/A
N/A
N/A
0.027*
N/A
N/A
N/A
0.037*
N/A
N/A
N/A
0.042*
N/A
N/A
N/A
0.075*
N/A
N/A
N/A
<10
C
* = Data obtained from Cabot Corporation using TEM method
N/A = Data could not be measured using the Coulter Principle
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Al
A1,0,
Particle size of reactant (pm)
Ti02
C
Al
Heating
method
MW
power
(w)
AhOj
i
p
i l l
* H
90 vol% of the
sample has particle
size less than (pm)
Wt%
excess
10 vol%of the
sample has particle
size less than (pm)
c) The particle size distribution of the sixteenth-runs Plackett-Burman design of the
product AkOj-TiC powders
Parameter
Particle size
distribution
(vol%)
Run
1
0
0
1.0-5.0
0.075
•200mesh
0
MH
1600
3.4
28.7
66.4
2
40
40
0.9-1.6
0.075
3 -4 .5
40-45
MHH
800
23
15
44.7
3
40
0
1.0-5.0
0.075
3 -4 .5
0
MHH
800
8.5
48.2
85.4
4
40
40
I.0-5.0
0.075
3 -4 .5
40-45
MH
1600
N/A
5
0
40
1.0-5.0
0.017
3 -4 .5
40-45
MH
800
4.1
34.6
70.8
6
40
0
0.9-1.6
0.075
3 -4 .5
0
MH
1600
7.8
32.8
64.4
7
40
40
1.0-5.0
0.017
-200mesh
0.49
MHH
1600
5.4
13.6
45.5
8
40
40
0.9-1.6
0.017
-200mesh
0.49
MH
800
N/A
9
0
0
0.9-1.6
0.075
•200mesh
0
MHH
800
4.6
45.8
80.8
10
0
40
1.0-5.0
0.075
-200mesh
0.49
MHH
800
6.5
53.2
86.7
11
0
0
0.9-1.6
0.017
3 -4 .5
0
MH
800
10J
48.2
85.2
12
0
0
1.0-5.0
0.017
3 -4 .5
0
MHH
1600
10.3
53.3
85.9
13
0
40
0.9-1.6
0.075
•200mesh
0.49
MH
1600
3.7
43.7
80.2
14
0
40
0.9-1.6
0.017
3 -4 .5
40-45
MHH
1600
3.4
21.7
57.4
15
40
0
1.0-5.0
0.017
•200mcsh
0
MH
800
14.9
55.2
90.2
16
40
0
0.9-1.6
0.017
-200mesh
0
MHH
1600
18.2
57.5
94.4
N/A
N/A
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
N/A
N/A
220
d) The particle size distribution of the thirtieth-runs Central composite design for MH and
R un
fo r
MH
Factor
Particle size distribution (vol%)
10 v o l% o f
th e sam ple
has particle
size less
than (pm )
5 0 v o l% o f
the sam ple
has particle
size less
than (pm)
90 vol% o f
the sam ple
has particle
size less
than (pm )
C Size
AI2 O 3
Size
Wt%
excess AJ
W t%
excess
A120 3
1
0.027
<10
10
10
4.9
23.7
62.6
2
0.027
<10
10
30
5.0
19.6
52.6
3
0.027
<10
30
10
6.2
31.4
66.2
4
0.027
<10
30
30
4.9
19.5
62.2
5
0.027
44-53
10
10
5.7
38.2
82.2
6
0.027
44-53
10
30
5.5
41.8
74.5
7
0.027
44-53
30
10
6.3
41.8
82.0
8
0.027
44-53
30
30
3.6
31.2
70.1
9
0.042
<10
10
10
4.6
34.4
67.5
10
0.042
<10
10
30
3.7
21.1
63.3
11
0.042
<10
30
10
4.5
33.3
71.7
12
0.042
<10
30
30
3.7
20.4
61.6
13
0.042
44-53
10
10
4.5
38.8
90.4
14
0.042
44-53
10
30
4.4
46.8
96.4
15
0.042
44-53
30
10
4.5
46.7
106.2
16
0.042
44-53
30
30
2.8
27.5
69.2
17
0.037
20-38
20
20
4.0
32.6
64.2
18
0.037
20-38
20
20
4.3
30.4
61.6
19
0.037
20-38
20
20
3.5
34.2
72.6
20
0.037
20-38
20
20
3.6
33.2
67.1
21
0.037
20-38
20
20
4.2
30.0
67.4
22
0.037
20-38
20
20
3.5
32.6
70.3
23
0.017
20-38
20
20
5.1
33.4
81.2
24
0.075
20-38
20
20
6.6
45.0
107.8
25
0.037
0 .4 9
20
20
7.8
60.5
167.4
26
0.037
63-74
20
20
5.4
43.7
103.3
27
0.037
20-38
0
20
7.2
43.6
138.6
28
0.037
20-38
40
20
4.0
30.1
80.3
29
0.037
20-38
20
0
5.1
32.8
72.8
30
0.037
20-38
20
40
3.7
26.3
53.4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
221
R un
fo r
MHH
Factor
C S ize
Particle size distribution (vol%)
AI2 O 3
Size
W t%
excess Al
W t%
excess
AI2 O 3
1 0 vol% o f
th e sam ple
has particle
size less
than (pm )
50 vol% o f
th e sam ple
has particle
size less
than (pm )
90 vol% o f
th e sample
has particle
size less
than (pm)
1
0.027
< 1 0
1 0
1 0
4.2
24.2
67.3
2
0.027
< 1 0
1 0
30
4.6
20.4
55.2
3
0.027
< 1 0
30
1 0
5.6
32.1
68.5
4
0.027
< 1 0
30
30
4.2
18.1
53.4
5
0.027
44-53
1 0
1 0
4.4
29.1
69.7
6
0.027
44-53
1 0
30
5.5
40.0
75.2
7
0.027
44-53
30
1 0
6 .8
44.2
84.4
8
0.027
44-53
30
30
2.9
30.3
67.2
9
0.042
< 1 0
1 0
1 0
4.0
32.2
67.7
1 0
0.042
< 1 0
1 0
30
3.6
21.5
62.8
11
0.042
< 1 0
30
1 0
4.7
39.4
78.9
1 2
0.042
< 1 0
30
30
3.5
17.8
55.6
13
0.042
44-53
1 0
1 0
4.4
41.6
94.2
14
0.042
44-53
1 0
30
4.1
40.9
90.9
15
0.042
44-53
30
1 0
4.5
48.9
110.7
16
0.042
44-53
30
30
2 .6
24.8
66.7
17
0.037
20-38
2 0
2 0
3.4
40.6
68.5
18
0.037
20-38
2 0
2 0
3.7
33.8
60.4
19
0.037
20-38
2 0
2 0
3.3
37.2
70.2
2 0
0.037
20-38
2 0
2 0
3.2
35.7
67.8
21
0.037
20-38
2 0
2 0
3.4
38.6
6 6 .2
2 2
0.037
20-38
2 0
2 0
3.8
32.2
66.4
23
0.017
20-38
2 0
2 0
5.5
48.2
146.9
24
0.075
20-38
2 0
2 0
7.4
45.8
100.5
25
0.037
0 .4 9
2 0
2 0
8 .6
62.3
1S9.2
26
0.037
63-74
2 0
2 0
6.7
60.9
142.4
27
0.037
20-38
0
2 0
9.2
44.3
140.8
28
0.037
20-38
40
2 0
4.2
31.8
90.3
29
0.037
20-38
2 0
0
6.4
44.2
88.5
30
0.037
20-38
2 0
40
3.6
30.2
56.9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF REFERENCES
Ada85
S. Adachi, T. Wada, Y. Miyamoto, M. Koizumi and O. Yamada,
J.Am.Ceram.Soc., 72,805 (1985)
Ada90
S. Adachi, T. Wada, T. Mihara, Y. Miyamoto, and M. Koizumi, “High-Pressure
Self-Combustion Sintering of Alumina-Titanium Carbide Ceramic Composite,”
J.Am.Ceram.Soc., 73[5], 1451-52 (1990)
Ahm88
I. Ahmad, G.T Chandler, and D.E. Clark, “Processing of Superconducting
Ceramics using Microwave Energy,” Mater. Res. Soc. Proc., 124,239-246 (1988)
Ahm9l
I. Ahmad, R. Dalton, and D.E. Clark, “Unique Application of Microwave Energy
to the Processing of Ceramic Materials,” J.Microwave Power, 26, 128-138 (1991)
AntOO
S. Anthonysamy, K. Ananthasivan, V. Chandramouli, I. Kaliappan and P.R.V.
Rao, “Combustion Synthesis of Urania-Thoria Solid Solutions,” J. Nuclear
Materials, 278,346-357 (2000)
Ato98
D. Atong and D.E. Clark, “ Synthesis of TiC-A^C^ Composites Using
Microwavejnduced Self_Propagating High Temperature Synthesis (SHS),”
Ceramic Engineering and Science Proceedings, 19[4], 415-421 (1998)
Ava96
P.B. Avakan, MD. Nersesyan, and A.G. Merzhanov, New materials for electronic
engineering, Am. Ceram. Soc. Bull., 75,50-55 (1996)
Bag88
D.R. Baghurst, A M Chippendale and D.M.P. Mingos, “Microwave Synthesis for
Superconducting Ceramics,” Nature, 332,311 (1988)
Bal88
J.G. Baldoni and S.T. Buljan, “Ceramics for Machining,” Amer.Ceram.Soc.Bull.,
76[2] 381-87 (1988)
Bay94
S.S. Bayya and R.L. Snyder, “Self-Propagating High-Temperature Synthesis
(SHS) and Microwave-Assisted Combustion Synthesis (MACS) of Thallium
Superconducting Phases, Physica, C225,83-90 (1994)
Bat95
J. Batt, J.G. Binner, TE. Cross, N.R. Greenacre, M.G. Hamlyn, R.M. Hutcheon,
W.H. Sutton, and C.M. Weil, “ A Parallel Measurement Programme in High
Temperature Dielectric Property Measurement: An Update, Ceramic Transactions,
59,243-250 (1995)
Bes91
E. Bescher and J.D. Mackenzie, “ Microwave Heating of Cermets,” Microwaves:
Theory and Application in Materials Processing, eds. D.E. Clark, FD. Gac, and
W.H. Sutton, Ceramic Transactions, v.21, American Ceramic Society,
Westerville, OH 557-563 (1991)
222
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
223
Bin92
J.G.P. Binner, I.A. Al-Dawery, C. Aneziris, and T.E. Cross, “Use of the Inverse
Temperature Profile in Microwave Processing of Advanced Ceramics,” Mater.
Res. Soc. Proc., 269,357-362 (1992)
Bin93
J.G.P. Binner and T.E. Cross, “Applications from Temperature Profile Control
During Microwave Sintering,” Engineering Ceramics: Fabrication Science and
Technology, ed. DE. Thompson, 53-61 (1993)
Bor74
IE. Borovinskaya, A.G. Merzhanov, N.P. Novikov and A.K. Filonenko, “Gasless
Combustion of Mixtures of Powdwerd Transition Metals with Boron,”
Comb.Explos.Shock.Wave, 10,2-10 (1974)
Bor86
ME. Borom and M. Lee, “Effect of Heating Rate on Densification of AluminaTitanium Carbide Composites,” Adv.Cera.Mater., I[4] 335-340 (1986)
Bor88
D. Bordui, “Hard-Part Machining with Ceramic Inserts,” Am.Ceram.Soc.Bull.,
67[6], 998-1001 (1988)
Bow91
C.R. Bowen, S. Hulsmann, and B. Derby, “Preliminary Studies on the
Manufactured of Multiphase Ceramics by SHS,” Proceeding of the 2nd European
Ceramic Society Conference, Augsburg, September, 1991, edited by G. Ziegler
and H. Hausner, p. 361
Bow94
C.R. Bowen and B. Derby, “Differential Thermal Analysis of Ignition
Temperatures in a Self-Propagating High-Temperature Synthesis Reaction,” J.
Therm.Anal., 42,713-719 (1994)
Bow96
C.R. Bowen and B. Derby, “The formation of TiC/A1203 microstructures by
microstructures by SHS reaction,” J.Mater.Sci., 31,3791-3803 (1996)
Bro91
R.J. Brook, Concise Encyclopedia of Advanced Ceramic Material, Pergamon
Press, New York, 486 (1991)
Bru88
R.W. Bruce, “New Frontiers in the use of Microwave Energy: Power and
Metrology,” Microwave Processing of Materials, eds. W.H. Sutton, M.H. Brooks,
and I.J. Chabinsky, Material Research Society Symposium Proceedings, v. 124,
Materials Research Society, Pittsburgh, 3-15 (1988)
Bur98
S.J. Burden, “Comparison of Hot-Isostatically Pressed and Uniaxially HotPressed Alumina-Titanium Carbide Cutting Tools,” Am.Ceram.Soc.Bull., 67[6]
1003-1005 (1998)
Cha98
V. Chandramouli, S. Anthonysamy, PE.V. Rao, R. Divakar and D.
Sundararaman, “Microwave Synthesis of Solid Solutions of Urania and Thoria A Comparative Study,” J. Nuclear Materials, 254,55-64 (1998)
Cho95
Y. Choi and S.W. Rhee, “Reaction of Ti02-Al-C in the combustion synthesis of
TiC-AhOj composite,” JAm.Cera.Soc., 78[4],986-92 (1995)
Cla88
D£. Clark, I. Ahmad and G. Chandler, “Processing of Superconducting Ceramics
Using Microwave Energy, Patent 07/177,774 (1988)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
224
Cla91
D.E. Clark, I. Ahmad and R.C. Dalton, “Microwave Ignition and Combustion
Synthesis of Composites,” Materials Science and Engineering, A144,91-97
(1991)
Cla94
D.E. Clark, D.C. Folz, RX. Schulz, Z.Fathi, AD. Cozzi, A. Boonyapiwat,
P.Korarenko, R.Difiore and C.B.Jones, “Microwave Processing at the University
of Florida, “ Microwave Processing of Materials IV, eds. M.F. Iskander, R.J.
Lauf, and WH. Sutton, MRS Symposium Proceedings, v.347, MRS, Pittsburgh,
489-500 (1994)
Cla96
D.E. Clark and W.H.Sutton, “ Microwave processing of materials,” Annual
Review of Materials Science, eds. M.Tirrell, E.N. Kaufinann, J.A. Giordmaine,
and J.B. Watchtmsn, Jr., V.26, Annual Reviews Inc., Palo Alto, CA, 299-331
(1996)
Coz91
AX). Cozzi, D.K. Jones, Z. Fathi, and D.E. Clark, “Microstructural Evolution of
Yba2Cu307-x Using Microwave Energy,” Microwaves: Theory and Application
in Materials Processing, D.E.Clark, F.D.Gac, and W.H.Sutton,eds., American
Ceramic Society, Westerville, OH, 21,357-64 (1991)
Coz95
AX). Cozzi, D.E. Clark, M X Ferber and V.J. Tennery, “Apparatus for the
Joining of Ceramics Using Microwave Energy,” Microwaves: Theory and
Application in Materials Processing HI, eds. D.E. Clark, D.C. Folz, S.J. Oda and
R. Silberglitt, Ceramic Transactions, v.59, American Ceramic Society,
Westerville, OH, 389-396 (1995)
Coz96
AX). Cozzi, Theory and Application of Microwave Joining, Dissertation,
University of Florida (1996)
Cri82
J.F. Crider, CeramEngng. Sci .P ro c 3,519 (1982)
Cut85
R.A. Cutler, A V. Virka,and J.B. Holt, “Synthesis and Densificadon of OxideCarbide Composites, “CeramXng.Sci.Proc., 6[7-8], 715-28 (1985)
Cut88
RA. Cutler, AC. Hurford, and A.V. Virka, “Pressureless-Sintered A1203-TiC
Composites,” Mater.Sci.Eng., A105/106,183-92 (1988)
Cut92
RA. Cutler, K.M. Rigtrup, and AV. Virkar, J.Am.Ceram.Soc., 75,36 (1992)
Dal90
R.C. Dalton, I. Ahmand and D.E. Clark, CeramXng.Sci.Proc., 11,1729 (1990)
Dan93
D.J. Grellinger and M A Janney, “ Temperature measurement in a 2.45GHz
Microwave Furnace,” Ceramic Transactions, 36 529-538 (1993)
De90
A. De’, “Ultra Rapid Sintering of Alumina with Microwave Energy at 2.45 GHz,”
M.S. Thesis, University of Florida (1990)
De9la
A.S. De’, I. Ahmad, EJ). Whitney and DX. Clark, “Microwave (Hybrid) Heating
of Alumina at 2.45GHz: I. Microstructural Unifoimity and Homogeneity,”
Ceramic transactions, 21,319-328 (1991)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
225
De91b
A.S. De’, I. Ahmad, E.D. Whitney and D.E. Clark, “Microwave (Hybrid) Heating
of Alumina at 2.45GHz: II Effect of Processing Variables, Heating Rates and
Particle Size,” Ceramic transactions, 21,329-340 (1991)
Dec86
R.V. Decareau and R.A. Peterson, Microwave Processing and Energineering, Ellis
Harwood, Ltd., Chichester, UK (1986)
Dim89
P. Dimintrou, H. Halvacek, S.M. Valone, P.G. Behrens, GP. Hansen and J.
Margrave, “Laser- Induced Ignition in Solid State Combustion,” AichE. J., 35,
1085-1096 (1989)
Dol78
S.K. Dolukhanyan, MD. Neresyan, N.A. Martirosyan, and AG. Merzhanov, IZv.
Akad. Nauk. SSSR. Neorg. Mater.,14,1581-84,1978
Duf88
L.C. Dufour, C. Monty, and G.P. Ervas, Surfaces and Interfaces of Ceramic
Materials, Kluwer Academic Publishers, Dordrectht, Netherlands, 173 (1988)
Dun91
S.D. Dunmead and Z.A Munir, J.Mater.Sci., 26,2410 (1991)
Eki87
J.W. Ekin, Adv. Ceram. Mat., 2,586 (1987)
Eni68
J.W. Enig, Proc. Royal. Soc., A305,205 (1968)
Esl89
M. Eslamloo-Grami and Z.A. Munir, Mater.Sci.Report, 3,227-365 (1989)
Esl90
M. Eslamloo-Grami and ZA. Munir, J. Am. Ceram. Soc., 73,2222 (1990)
Fan93
Y. Fang, DJC. Agrawal, DAI. Roy and R. Roy,” Microwave Sintering of Calcium
Strontium Zirconium Phosphate Carmics,” Caramic Transactions, 36,109-114
(1993)
Fen92
H.J. Feng, J.J. Moore, and D.G. Wirth, “Combustion Synthesis of Ceramic-Metal
Composite Materials: The TiC-A1203-Al System,” Met.Trans, 23A, 2373-79
(1992)
Fen95
H.J. Feng and J.J. Moore, “In Situ Combustion Synthesis of Dense Ceramic and
Cermic-Metal Interpenetrating Phase Composites,” Metallurgy and Materials
transactions B, 26:2 [2] 265-273 Apr 1995
Fil7 5
AiC. Filonenko and V.I. Vershinnikov, CombExplos.Shock. Wave, 11,301
(1975)
Fra85
W.L Frankhouser, M.G. Kieszek, K.W. Brendley, And S.T. Sullivan, Gasless
Combustion Synthesis of Refractory Compounds, ParkRidge, New York: Noyes,
152pp, 1985
Gab85
KA. Gabriel, S.G. Wax and J.W. McCauley, Materials Processing by SelfPropagating High-Temperature Synthesis (SHS), DARPA/ ARMY SHS
Symposium Proceedings, Daytona Beach, FL (1985)
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
226
Gre95
G.Darby, D.E. Clark, R. Di Fiore, D.Folz, R. Schulz, A. Boonyapiwat and D.
Roth, “Temperature Measurement During Microwave Processing,” 59,515-522
(1995)
Haa79
P.A. Hass, Am Ceram. Soc. Bull., 58,873 (1979)
Hal88
D.C. Halverson, B. Y. Lum and Z.A. Munir, Proceeding of Symposium on High
Temperature Material Chemistry IV, eds. Z.A. Munir, D. Cubicciotti and H.
Tagawa, 613, The Electrochemical Society, N.J. (1988)
Hen90
L.L. Hench and JiC. West, Principles of Electronic Ceramics, John Wiley & Sons,
New York (1990)
Hens83 W.F. Henshaw, A. Niiler and T. Leete, Cerm.Eng.Sci.Proc., 4,634 (1983)
Her85
J.M Herbert, Ceramic Dielectrics and Capacitors, Gordon and Breach Science,
New York, 226 (1985)
Hip54
A.R. Von Hippel, “Dielectrica and Waves,” John Wieley and Sons, New York, 61
(1954)
Hla91
V. Hlavacek, “ Combustion Synthesis: A Historical Perspective,” Ceram.Bull., 70,
240-243 (1991)
Hol90
C.E. Holcombe and N.L. Dykes, J.Mat.Sci.Lett., 9,425 (1990)
Holt82
J.B. Holt, “Combustion Synthesis of Refractory Materials,” Lawrence Livermore
National Laboratory Report No. UCRL-53258 (1982)
Holt86
J.B. Holt and Z A Munir, “Combustion Synthesis of Titanium Carbide: Theory
and Experiment,” J. Mater. Sci., 21,251-259 (1986)
Holt87
J.B. Holt, “ The Use of Exothermic ractions in the Synthesis and ensification of
Ceramic Materials,” Mat.Res.Soc.Bull., 12,60-64 (1987)
Holt91a J.B. Holt and S.D. Dunmead, “Self-Heating Synthesis of Materials,”
Annu.RevMater.Sci., 21,305-34 (1991)
Holt9 lb J.B. Holt, “Self-Propagating High-Temperature Synthesis,” Engineered materials
Handbook, V4: Ceramics and Glass, 227-231, ASM International (1991)
Huf60
J.B. Huffadine, “The fabication and Properties of Molybdenum Dilicide and
Molybdenum Disilicide-Alumina,” Special Ceramics, p.220, eds. P.Popper,
Academic Press, New York, 1960
Hum80
KD. Humphrey, “ Microwave Sintering of BaTi0 3 ,” MS Thesis. Univ. MissouriRolla, 77 (1980)
Ish89
T. Ishigaki, K. Sato, and Y. Moriyoshi, “Pressureless Sintering of TiC-A1203
Composites,” J.Mater.Sci.Lett, 8,678-80 (1989)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
227
Jan89
M.A. Janney and H.D. Kimery, “Microstructure Evolution in Microwave Sintered
Alumina,”p. 382,Sintering of Advanced Ceramics, eds. C.A. Handwerker, J.E.
Blendall and W.A. Kaysser, American Ceramic Society, Columbus, OH (1989)
Jan91
M.A. Janney, C.L. Calhoun, and H.D. Kimery, “Microwave Sintering of Zirconia8mol% yittria,” Microwaves: Theory and Apphcation in Materials Processing,
D.E.Clark, F.D.Gac, and W.H.Sutton,eds., American Ceramic Society,
Westerville, OH, 21,311-318 (1991)
Jan92
M.A. Janney, HJD. Kimrey, and J.O. Kiggans, Microwave Processing of
Ceramics: Guidelines Used at the Oak Ridge National Laboratory, 173-185, MRS
Symposium Proceedings, v.269, Microwave Processing of Materials m, eds. R.L.
Beatty, W.H. Sutton, and M.F. Iskander, Pittsburgh, PennsylvaniaMRS, 173-185
(1992)
Kai90
Y. Kaieda, M. Otaguchi, and N. Oguro, Combustion and Plasma Synthesis of
High-Temperature Materials, eds. Z.A. munir and J.B. Holt, p. 106, VCH (1990)
Kat85
J.D. Katz, R.D. Blake and C.P. Scherer, “Microwave Sintering of Titanium
Diboride,” Ceram. Eng. Sci. Proc., 10,7-8,857-867 (1985)
Kat91
J.D. Katz, RD. Blake and V.M. Kenkre, “Microwave Enhanced Diffusion?,”
Ceramic Transactions, 21,95-105 (1991)
Kat92
J.D. Katz, “Microwave Sintering of Ceramics,” Annual Review of Materials
Science, eds. R.A. Huggins, J.A. Giordmaine, and J.B. Watchtmsn, Jr., Annual
Reviews Inc., Palo Alto, CA, 153-170 (1992)
Kec90a
L.J. Kecskes, T. Konke, and A. Niiler, J.Amer.Ceram.Soc., 73,1274 (1990)
Kec90b L.J. Kecskes, R J. Benck, and PH. Netherwood Jr., JAmer.Ceram.Soc., 73,383
(1990)
Kha66
B.I. Khaikin and AG. Merzhanov, Combust. Explos. Shock. Waves, 2,22-27
(1966)
Kig91
J.O. Kiggans, CH.. Hubbard, R.R. Steele, H.D. Kimrey, C £. Holcombe and T.N.
Tiegs, Ceramic Transactions, 21,403-408 (1991)
Kin65
A.G. King, “Ceramics for Cutting Metals,” BuUAm.Ceram.Soc., 43[5], 395-401
(1965)
Koi90
M. Koizumi and Y. Miyamoto, Combustion and Plasma Synthesis of HighTemperature Materials, eds. Z.A munir and J.B. Holt, p.54, VCH (1990)
Koz9l
H. Kozuka and JD. Mackenzie, Ceramic Transactions, 21,387-394 (1991)
Kra81
MJC. Krage, Am. Ceram. Soc. Bull., 60, 1234 (1981)
Krap64
S.Krapf, “Ceramic Materials and Methods for their Manufacture,” U.S.Pat.No.
3143423, Aug 4,1964
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
228
Kub79
O. Kubashewski and C.B. Alcock, Metallurgical Thermochemistry, 5th ed.,
Pergamon Press, New York (1979)
Kum91
S.N. Kumar, A. Pant, R.R. Sood, J. Ng-Yelim, R.T. Holt, Ceramic Transactions,
21,395-402,(1991)
Kun96a A.O. Kunrath, T.R. Strohaecker, J.J. Moore, “Combustion Synthesis of MetalMatrix Composites: Part I, The Ti-TiC-A1203 System,” Scripta Materialia, 34[2],
175-181(1996)
Kun96b A.O. Kunrath, T.R. Strohaecker, J.J. Moore, “Combustion Synthesis of MetalMatrix Composites: Part II, The Ti-TixAly-A1203 System,” Scripta Materialia,
34[2], 183-188(1996)
Kun96c A.O. Kunrath, T.R. Strohaecker, J.J. Moore, “Combustion Synthesis of MetalMatrix Composites: Part IE, The Al-TiC-A1203 System,” Scripta Materialia,
34[2J, 189-194(1996)
Leb92
J. Lebrat and A. Varma, “Self-Propagating High-Temperature Synthesis of
NisAl,” Combust. Sci.and Tech., 88,211-221 (1992)
Lee83
M. Lee, M.K. Brun, and T. Tien, “The High-Temperature Fracture Toughness of
SiAlON,” Ceram.Eng.Sci.Proc., 4[9-10], 864-73 (1983)
Lee88
M. Lee and M.P. Borom, “Rapid rate sintering of AJ203-TiC composites for
Cutting Tool Applications,” Adv.Ceram.Mater., 3[1] 38-44 (1988)
Log84
K.V. Logan and J.D. Walton, “T1B2 Formation Using Thermite Ignition,” Ceram.
Eng. Sci. Proc., 5,712-738 (1984)
Log90
K.V. Logan, J.T. Sparrow and W.J.S. Mclemore, “Experimental Modeling of
Particle-Particle Interactions During SHS of TiB2 -Al2 0 3 ,” Combustion and
Plasma Synthesis of High-Temperature Materials, eds. Z.A. Munir and J.B. Holt,
VCH, New York, 219-228 (1990)
Lor91
CP. Lorenson, M.D. Ball, R. Herzig, and H. Shaw, “The Microwave Behavior of
Metallic-Insulator Composite Systems”, MRS. Symp. Proc., v.189,279-282
(1991)
Lyn74
C.T. Lynch, CRC Handbook of Materials Science, CRC Press, Boca Raton, FL, 1,
105 and 109 (1974)
Mac84
JP. MacDowell, Am. Ceram. Soc. Bull., 63,282 (1984)
Mak79
Y.M. Maksimov, A.G. oak, G.V. Lavrenchuk and Y.S. Naiborodenko, “Spin
Combustion of Gasless Systems” CombPxplos.Shock.Wave, 15,415-18 (1979)
Man84
B. Manley, JP . Holt, and Z A Munir, “Sintering of Combustion-Synthesized
Titanium Carbide,” Materials Science Research, 16, Sintering and Heterogeneous
Catalysis, eds. S.C. Kuczynski, AE. Miller and G.A. Sargent, Plenium Press,
New York, 303-316 (1984)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
229
Mar85
N.A. Martirosyan, SJC. Dolukhanyan and M.G. Mershanov, Combust. Explos.
Shock. Waves., 19,569 (1985)
McC82
J.W. McCauley, N.D. Corbin, T. Resetar, and P. Wong, Ceram.Eng.Sci.Proc., 3,
538(1982)
Mee87a T.T. Meek, R.D. Blake, J.J. Petrovic, Ceram. Eng. Sci. Prog., 8,861 (1987)
Mee87b T.T. Meek, “Proposed Model for the Sintering of a Dielectric in a Microwave
Field,” J.Mater.Sci.Let, 6,638-640 (1987)
Mer71
A.G. Merzhanov, VM. Skhiro, P. Borovinskaja, USSR Pat. No. 255 221 (1971)
Mer72
A.G. Merzhanov and I.P. Borovinskaya, Self- Propagating High-Temperature
Synthesis of Refractory Inorganic Compounds,” Dokl. Acad. Sci USSR (Chem.)
(English Transaction), 204[2], 429-432 (1972)
Mer74
A.G. Merzhanov, ArchProcesow Spalania, 5,17-39 (1974)
Mer75
A.G. Merzhanov, Combustion Process in Chemical Technology and Metallurgy,
Eds. A.G. Merzhanov, Chemogolovka, 1975
Mer83
A.G. Merzhanov, “Self-Propagating High-Temperature Synthesis,” physical
Chemistry: Modem Problems, ed. Y.M. K. Khimiya, Moscow, 5 (1983)
Mer89
A.G. Merzhanov, Combustion and Plasma Synthesis of High Temperature
Materials, eds. ZA.Munir, J.B.Holt, 1-53, New York: VCH, 1989
Mer90
A.G. Merzhanov, “Self-Propagating High-Temperature Synthesis: Twenty Years
of Search and Findings,” Combustion and Plasma Synthesis of High-Temperature
Materials, ed. Z.A. Munir and J.B. holt, 1, VCH Inc., New York (1990)
Mer93a A.G. Merzhanov, “Theory and Practice of SHS: Worldwide State of the Art and
the Newest Result,” IntJ. Self-Propagating High-Temperature Synthesis, 2,11358(1993)
Mer93b A.G. Merzhanov, “New Manifications of an Ancient Process,” Chemistry of
Advanced Materials: A Chemistry for the 21st cCentury, eds. C.N.R.Rao,
London: Blackwell, 19-39 (1993)
Met83
A.C. Metaxas and R J. Meredith, Industrial Microwave Heating, Peter Peregrinus,
London, UK, 1983
Miy84
Y. Miyamoto, MJCoizumi, and O.Yamada, “JAm.Ceram.Soc., 67[l 1], C-224
(1984)
Miy90a
Y. Miyamoto, CeramBull., 69,686 (1990)
Miy90b
Y. Miyamoto and M. Koizumi, Combustion and Plasma Synthesis of HighTemperature Materials, eds.Z.A. Munir and J.B. Holt, 163, VCH (1990)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
230
Moo91
E.h. Moore, I. Ahmad and D.E. Clark, “Microwave Thermogravimetric Analyzer,
(MTGA),” Microwaves: Theory and Application in Materials Processing, eds.
D.E.Clark, F.D. Gac and W.H. Sutton, Ceramic Transactions, v.21, American
Ceramic Society, Westerville, OH, 675-681 (1991)
Moor90 J.J. Moore and H.C. Yi, Proceedings of the first US-Japanese Workshop on
Combustion Synthesis, eds. Y. Kaieda and J.B. Holt, p.33, National research
Institute for Metals, Tokyo, Japan (1990)
Moor94 J.J. Moore, D.W. Ready, H.J. Feng, K. Monroe, and B. Mishra, “The Combustion
Synthesis of Advanced Materials,” Journal of the Minerals Metals and Materials
Society,” 46: [11] 72-78 Nov 1994
Moor95a J.J. Moore and H.J. Feng, “Combustion Synthesis of Advanced Materials: Part I.
Reaction Parameters,” Prog. Mater.Sci., 39,243-273 (1995)
Moor95b J.J. Moore and H.J. Feng, “Combustion Synthesis of Advanced Materials: Part II.
Classification, Applications and Modelling,” Prog. Mater.Sci., 39,275-316
(1995)
Mun88
Z.A. Munir, Bull.Am.Ceram.Soc., 67,342-49 (1988)
Mun89
Z.A. Munir and U. Anselmi-Tamburini, “Self-Propagating Exothermic reactions:
The Synthesis of High-Temperature Materials by Combustion,” Mater.Sci.Rep., 3,
277-365 (1989)
New91 R.E. Newnham, S.J. Jang, M. Xu, and F. Jones, “Fundamental Interaction
Mechanisms Between Microwaves and Matter,” Ceramic transactions, 21,51-68
(1991)
Nov75
N.P. Novikov, I.P. Borovinskaya and A.G. Merzhanov, Combustion Synthesis in
Chemical Technology and Metallurgy, eds. A.G. Merzhanov, Chemogolovka
(1975)
Oda81
O. Odawara and J. Ikeuchi, JJpn.InstMetals, 45,316 (1981)
Oda84
O. Odawara and J. Keuchi, J.Am.Ceram.Soc., 64[4], c-86 (1984)
Pam97
R. Pampuch, “Some Fundamental Versus Practical Aspects of Self-Propagating
High-Temperature Synthesis,” Solid State Ionics, 101-103,899-907 (1997)
Par98
H.R.K. Park, Y.S. Han, DiC. Kim, and C.H.R. Kim, “Synthesis of LaCr03
Powders by Microwave Induced Combustion of Metal Nitrate-Urea Mixture
Solution,” JJMater.Sci-Letters, 17,785-787 (1998)
Pat93
K.C. Patil, “Advanced Ceramics: Combustion Synthesis and Properties,”
Bull.Mater.Sci, 16,533-541 (1993)
Pat97
K.C. Patil, S.T. Aruna and S. Ekambaram, “Combustion Synthesis,” Current
Opion in Solid State and Materials Science, 2[2], 158-165, Apr 1997
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
231
Pet92
A. Petrie, P.T.W. Lee, and H.C. Yi, “Combustion Synthesis of TiC-A1203-Al
Composites,” Solid State Phenomena, 25-26,217-224 (1992)
Phi85
K.A. Philpot, “Pre-Combustion Processes in the Combustion Synthesis of NickelAluminum Compounds,” M.S. Thesis, University of California, Davis (1985)
Rab90
B.H. Rabin, G.E. Korth, and RX. Williamson, “Fabrication of Titanium CarbideAlumina Composites by Combustion Synthesis and Subsequent Dynamic
Consolidation,” J.Am.Ceram. Soc., 73[7], 2156-57 (1990)
Ric85a R.W. Rice and W.J. McDonough, J.Am.Ceram.Soc., 68[5], C-122 (1985)
Ric85b R.W. Rice, G.Y. Richardson, J.M. Kunetz, T. Schroeter and W.J. McDonough, “
Effects of Self- Propagating Synthesis Reactant Microstructure,” Material
Processing by Self-Propagating High-Temperature Synthesis (SHS), eds. K.A.
Gabriel, S.G. Wax and J.W. McCauley, DARPA/ARMY SHS Symposium
Proceedings, Oct. 21-23, Daytona Beach, FL, 201-234 (1985)
Ric86
R.W. Rice, W.J. McDonough, G.Y. Richardson, JJM. Kunutz and T. Schroeter,
Ceram.Engng.Sci.Proc., 7,751-760 (1986)
Ric90
R.W. Rice, Ceram.Eng. Sci.Proc., 11,1203 (1990)
Ric91
R.W. Rice, “Review Microstructural aspects of Fabricating Bodies by SelfPropagating Synthesis, J.Mater,Sci., 26,6533-6541 (1991)
Shk71
F.G. Shkadinski, B.I. Khaikin and A.G. Merzhanov, Comb.Explos.Shock.Wave,
7,15(1971)
Shki81
V.M. Shkiro, V.N. Doroshin and IP. Borovinskaya, Combust. Explos. Shock.
Waves USSR, 16,370 (1981)
Spo95
M.S. Spotz, d.j. Skamser, and DX. Johnson, “Thermal Stability of Ceramic
Materials in Microwave Heating,” J.Am.Ceram.Soc., 78[4], 1041-48 (1995)
Str67
R.K. Stringer and L.S. Williams, “Reaction Pressing: A New Fabrication Concept
for IntermetaUic and Metal-Metalloid Compounds”; p.37, Special Ceramics 4,
British Cermic Research Association, Academic Press, New York (1967)
Sub92
J. Subrahmanyam and M. Vijayakumar, “Self-Propagating High Temperature
Synthesis,” J.Mater.Sci, 27,6249-6273 (1992)
Sut88
W.H. Sutton, “Microwave Firing of High Alumina Castables,” Mater. Res. Soc.
Proc., 124,287-295 (1988)
Sut89
W.H. Sutton, “Microwave processing of ceramic Materials,”
Amer.Ceram.Soc .Bull., 68[2], 376-386, (1989)
Sut92
W.H. Sutton, “Microwave Processing of Ceramics - An Overview,” Mater. Res.
Soc. Proc., 269,3-20 (1992)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
232
Sut93
W.H. Sutton, “Key Issues in Microwave Process Technology,” Ceramic
Transactions, 36,3-18 (1993)
Tho92
J.J. Thomas, R.R. Jesse, DX. Johnson and H.M. Jennings, “Nitridation of NonIsothermal Silicon Compacts,” Microwave Processing of Materials m , 269.
Materials Research Society, eds. R.L. Beatty, W.H. Sutton, M.F. Iskander, 277283(1992)
Tin68
W.R. Tinga and W.A.G. Voss, Microwave Power Engineering, Academic Press,
NEW York, 189-199 (1968)
Tin88
W.R Tinga, “Fundamental of Microwave-Material Interactions and Sintering,”
Microwave processing of Materials, eds. W.H. Sutton, M.H. Brooks, and I.J.
Chabinsky, Material Research Society Symposium Proceedings, v. 124, Materials
Research Society, Pittsburgh, 33-43 (1988)
Tit54
R. Titterington and A.G. Simpson, “The Production and Fabrication of Tantalum
Powder,” Symposium on Powder Metallurgy, 11-18 (1954), Special Report No.58
The Iron and Steel Institute, London, U.K. (1956)
Var92
A. Varma and JP. Lebrat, “Combustion Synthesis of Advanced Materials,”
Chemical Engineering Science, 47 [9-11], 2179-2194, Jun-Aug 1992
Von54
A. Von Hippel, Dielectric Materials and Applications, Technology Press of MIT
and John Wiley & Sons, New York (1954)
Wah80
R.P. Wahi and B. Ilscher, “Fracture Behavior of Composites Based on A1203TiC,” J.Mater.Sci, 15,875-85 (1980)
Wal59
J.D. Walton and NX. Poulos, “Cermets for Thermite Reactions,”
J.Am.Ceram.Soc., 42[1], 40 (1959)
Wal89
J.D. Walton, “Early SHS Research at Georgia Tech 1955-60,” presented at
Meeting of Labcom SHS Working Group and Review of BTI/DARPA Advanced
Armor SHS Programs, Aug 8-9,1989, U.S_Army Materials Technology
Laboratory, Watertown, M.A. Unpublished work
Wan87
L. Wang, “An Investigation of the Combustion Synthesis of Copper Aluminides,”
M.S. Thesis, University of California, Davis (1987)
Wan90
L.L. Wang, Z.A. Munir and JJ3. Holt, “ The Combustion Synthesis of Copper
Aluminides,” Metall. Tran., 2 IB, 567-577 (1990)
Wan93
L.L. Wang, Z.A. Munir and Y.M. Maximov, “Review Thermite Reactions” Their
Utilization in the Synthesis and Processing of Materials,” J.Mater.Sci., 28,36933708(1993)
Whi53
W.E. White and A il. Bushley, “Aluminum Phosphide,” Ionic Synthesis, VTV, 23,
McGraw-Hill, New York 1953
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
233
Wil93a
M. Willert-Porada, B. Fisher, and T. Gerdes, “Application of Microwave Heating
to Combustion Synthesis and Sintering of A1203-TiC Ceramics,” Microwave;
Theoty and Application in Materials Processing n, eds. D£. Clark, W.R. Tinga,
and J.R. Laiajr., Ceramic Transactions, V36, American Ceramic Society,
Westerville, OH, 365-375 (1993)
Wil93b
M. Willert-Porada, “Reaction-Rate Controlled Microwave-Processing of Ceramic
Materials,” Ceramic Transactions, 36,277-286 (1993)
Wil95
M. Willert-Porada, “The Role of Space-Charge for Microwave Sintering of
Oxides,” Ceramic Transactions, 59,193-204 (1995)
Yam85
O. Yamada, Y. Miyamoto and M. Koizumi, “ High Pressure Self-Combustion
Sintering of Silicon Carbide,” Am. Ceram. Soc. Bull., 64[2], 319-321 (1985)
Yam86 O. Yamada, Y. Miyamoto, and M. Koizumi, J.Mater.Res.,1,275 (1986)
Yam87 O. Yamada, Y. Miyamoto, and M. Koizumi, JAm.Ceram.Soc., 70[9], C-206
(1987)
Yi89
H.C. Yi and J.J. Moore, J. Mater. Sci., 24,3449,3456 (1989)
Yi90
H.C. Yi and J.J. Moore, “Review Self-Propagating High-Temperature
(Combustion) Synthesis (SHS) of Poeder-Compacted Materials. J.Mater.Sci., 25,
1159-1168(1990)
Yii95
T. Yiin and M. Barmatz, “ Microwave Induced Combustion Synthesis of Ceramic
and Ceramic-Metal Composites,” Microwaves: Theory and Application in
Materials Processing III, eds. D.E.Clark, D.C.Folz, S.J.Oda, and R.Sflberghtt,
Ceramic Transactions, V59, American Ceramic Society, Westerville, OH, 389396 (1995)
Zen80
A.A. Zenin, A.G. Merzhanov, and GA. Nersisyan, Dokl.Acad.Sci.USSR
(Phys.Chem), 250,83-87 (1980)
Zen81
A.A. Zenin, A.G. Merzhanov, and GA. Nersisyan, Combust. Explos. Shock.
Waves, 17,63-71 (1981)
Zha91
S. Zhang and ZA Munir, J.Mater.Sci., 26,368 (1991)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BIOGRAPHICAL SKETCH
Duangduen Atong was born in Bangkok, Thailand in 1971. She received a Bachelor
of Science degree with honors in Material Science from the most prestigious college in
Thailand, Chulalongkorn University in 1992. She worked from 1992 to 1993 with the
National Metal and Material Technology Center (MTEC) as a research and development
project analyst. She then received a scholarship from the Royal Thai Ministry of Science
and Technology to pursue a graduate degree in the United States. She attended Rutgers the
State University of New Jersey, Piscataway, in 1993. Her graduate work was in the area of
tape casting of alumina with emphasis on the effect of binder systems on the packing of
particles of different sizes and shapes. She graduated from Rutgers University with a MS
degree in 1995. She then continued her PhD degree at the University of Florida, Gainesville,
under the guidance of Professor David E Clark. She was involved in a variety of microwave
material processing technology during her tenure in Gainesville campus: surface
modification of glasses using microwave energy, microwave-induced combustion synthesis
of composite powder, to name a few.
She has a strong intention to help establish a microwave material processing center
in Thailand after her graduation.
234
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
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 thesis for the degree of Doctor of Philosophy.
£■
— "
David E. Clark, 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 thesis for the degree of Doctor of Philosophy.
E. Dow Whitney
Professor of Materials Sciel’
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 thesis for the degree of Doctor of Philosophy^
A
cHblsky, Jr/
of Materials Science and
ngineenng
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 thesis for the degree of Doctor of Philosophy.
Hassan E. El-Shall
Associate 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 thesis for the degree of Doctor of Philosophy.
J. Erik Enholm
Professor of Chemistry
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
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 2000
£
M. J. Ohanian
Dean, College of Engineering
Winfred M.
Dean, Gradu
Ilips
{School
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
9 185 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа