<%BANNER%>

Role of particle size and nature of the reactants on the phase evolution, microstructure and densification of MoSi₂-TiB₂...

HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 Abstract
 1. Introduction
 2. Literature review
 3. Experimental procedure
 4. Results and discussion
 5. Summary and conclusions
 References
 Biographical sketch
 
PRIVATE ITEM Digitization of this item is currently in progress.
MISSING IMAGE

Material Information

Title:
Role of particle size and nature of the reactants on the phase evolution, microstructure and densification of MoSi₂-TiB₂ composites produced by in-situ reactions
Physical Description:
viii, 165 leaves : ill. ; 29 cm.
Language:
English
Creator:
Dempere Campillo, Luisa Amelia, 1967-
Publication Date:

Subjects

Subjects / Keywords:
Materials Science and Engineering thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Materials Science and Engineering -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 157-163).
Statement of Responsibility:
by Luisa Amelia Dempere Campillo.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030472463
oclc - 42453771
Classification:
lcc - LD1780 1999 .D389
System ID:
AA00018824:00001

MISSING IMAGE

Material Information

Title:
Role of particle size and nature of the reactants on the phase evolution, microstructure and densification of MoSi₂-TiB₂ composites produced by in-situ reactions
Physical Description:
viii, 165 leaves : ill. ; 29 cm.
Language:
English
Creator:
Dempere Campillo, Luisa Amelia, 1967-
Publication Date:

Subjects

Subjects / Keywords:
Materials Science and Engineering thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Materials Science and Engineering -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 157-163).
Statement of Responsibility:
by Luisa Amelia Dempere Campillo.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030472463
oclc - 42453771
Classification:
lcc - LD1780 1999 .D389
System ID:
AA00018824:00001

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
    Abstract
        Page vii
        Page viii
    1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
    2. Literature review
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    3. Experimental procedure
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
    4. Results and discussion
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
    5. Summary and conclusions
        Page 155
        Page 156
    References
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
    Biographical sketch
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
Full Text










ROLE OF PARTICLE SIZE AND NATURE OF THE REACTANTS ON THE PHASE
EVOLUTION, MICROSTRUCTURE AND DENSIFICATION OF MoSi2-TiB2
COMPOSITES PRODUCED BY IN-SITU REACTIONS











By

LUISA AMELIA DEMPERE CAMPILLO


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


1999























Dedicated to my parents
and my husband Mark













ACKNOWLEDGEMENTS


The pursuit of a doctoral degree abroad is a dream that very few people are able to

fulfill in Venezuela. The critical economic situation limits the number of scholarships

offered and the difference in the level of education and technology available compared to

the United States makes graduate school a significant challenge even for students with an

excellent undergraduate academic performance. I was very lucky having the opportunity

to face this challenge at the University of Florida. I think that this experience has

enriched my life far beyond my academic and personal expectations. I owe this priceless

opportunity to the Universidad de Oriente who granted me a scholarship through the

Consejo Nacional de Investigaciones Cientificas y Technol6gicas. With this scholarship,

I was able to complete my Master of Science degree and I will always be grateful for this

opportunity.

Due to economic problems in Venezuela my scholarship was cancelled yet I was

able to continue with my doctoral studies with a graduate research assistantship obtained

through Prof. Michael Kaufman. I sincerely thank Dr. Kaufman not only for helping to

make it economically possible for me to finish my doctorate, but also for being an

excellent academic advisor and counselor. His patience, support and understanding

attitude were invaluable during the most challenging times in graduate school.








I have to especially thank Dr. Moudgil and the Engineering Research Center

(ERC) for giving me the opportunity to continue with my research while some funding

problems were solved.

I am very grateful to my dissertation committee for their advice and guidance. I

have been especially fortunate to have a very distinguished group of professors on my

committee which through challenging questions and opportune suggestions led me to

understand and therefore explain more clearly the results of my work.

I will never be able to thank my family enough for their absolute support. My

parents not only always believed in me but, without them, the achievement of all my

goals would not have been possible. The unconditional support of my husband has been

determinant during the most difficult and stressful times. I will always be thankful for all

the sacrifices he endured to help me to achieve my dreams.

Finally my gratitude to all those who helped me through training on equipment,

preparing samples, performing analyses, and fixing, lifting and cutting things that

otherwise would have been a tough challenge for me.














TABLE OF CONTENTS



ACKNOWLEDGEMENTS ................................................................. iii

ABSTRACT ................................................................................vii

CHAPTER 1 INTRODUCTION ........................................................1
1.1 Background .................................................................... 1
1.2 A approach ...................................................................... 6

CHAPTER 2 LITERATURE REVIEW ....................................................8
2.1 The Mo-Si System ........................................................... ...8
2.1.1 History of Compositing Efforts of MoSi2 .......................9
2.1.2 Crystal Structure and Mechanical Properties of MoSi2 .........10
2.1.3 Role of Silica ....................................................11
2.2 Reinforcing Approaches ................................................... 13
2.2.1 Artificial Compositing vs. Solid-State Reactions ..............13
2.2.1.1 The Mo-Si-B system ..........................................15
2.2.1.2 The Mo-Si-Ti-B system ...................................16
2.3 Phase Evolution in In-situ Reactions ..........................................19
2.3.1 Mechanisms and Transformation Sequences ................. 20
2.3.2 Role of Oxygen ...............................................24
2.4 Phase Diagrams ..................................................................31

CHAPTER 3 EXPERIMENTAL PROCEDURE .................................... 39
3.1 Selection of Reactant Materials ...................................... ....39
3.1.1 Sampling for Particle Size Analysis ...............................40
3.1.2 Criteria for Raw Materials Selection ...............................40
3.2 Sample Processing ...............................................................43
3.2.1 Hot Pressing of Blended Powders ..................................44
3.2.2 Heat Treatments ................................................ 47
3.2.3 Diffusion Couples ..................................................47
3.3 Analytical Techniques ......................................................48
3.2.1 Metallographic Sample Preparation .................................48
3.3.2 X-Ray Diffraction Sample Preparation ............................48

CHAPTER 4 RESULTS AND DISCUSSION .......................................... 50
4.1 Particle Size Analysis .........................................................50
4.1.1 Elemental Reaction Powders ..........................................50








4.1.2 Displacement Reaction Powders ......................................57
4.2 Evolution of Phases...............................................................61
4.2.1 Displacement Reactions .............................................61
4.2.1.1 Heat treated samples........................................ 61
4.2.1.2 Hot pressed samples....................................... 68
4.2.2 Elemental Powders Reactions...........................................93
4.2.3 Reaction and Transport Mechanisms ...............................113
4.3 Microstructure and Densification .............................................119
4.3.1 Effect of Size and Distribution of sizes of Reactant Powders ...119
4.3.1.1 Displacement reaction using "large" powders ......... 119
4.3.1.2 Displacement reaction "small" vs "large" powders ... 122
4.3.1.3 Elemental reaction "small" vs "large" powders ......130
4.3.2 Effect of the Nature of Reactant Powders ........................134
4.3.3 Effect of Adjusting the Matrix:Reinforcement Ratio ........... 143
4.3.4 Effect of the Metal Particle Size on the Elemental Reaction
Com pacts ............................................................. 145
4.3.4.1 Increasing the size of molybdenum ...................146
4.3.4.2 Increasing the size of titanium ...........................149
4.4 Diffusion Couples..............................................................152

CHAPTER 5 SUMMARY AND CONCLUSIONS ........................................155

REFERENCES......................................................................................157

BIOGRAPHICAL SKETCH...................................................................164














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

ROLE OF PARTICLE SIZE AND NATURE OF THE REACTANTS ON THE PHASE
EVOLUTION, MICROSTRUCTURE AND DENSIFICATION OF MoSi2-TiB2
COMPOSITES PRODUCED BY IN-SITU REACTIONS

By

Luisa Amelia Dempere Campillo

August, 1999


Chairman: Prof. Michael J. Kaufman
Major Department: Materials Science and Engineering

The roles of particle size and nature (elemental vs. compound) of the reactants on

the phase evolution, microstructure and densification of MoSi2-TiB2 composites during

in-situ reactions have been investigated. It is shown that the elemental reaction proceeds

through a larger number of intermediate reaction steps and that the distribution of the

MoSi2 and TiB2 phases in the final microstructures varied considerably depending on

these factors.

In general, the small elemental reactant powders led to greater densification

although the amount of oxides introduced into the system from the surface of the powders

was higher. The dissolution and further segregation of the oxygen introduced into the

system led to the formation of silica pockets, as the most distinctive characteristic of the

resulting microstructures.








It was also found that the shape of the TiB2 grains is strongly influenced by the

size of the initial reactant particles. Smaller powders led to highly compacted pellets and

elongated TiB2 grains due to the simultaneous growth of a large number of fine grains.

Larger powders produced more regular equiaxed TiB2 grains.

The densification obtained through the elemental reaction is significantly higher

compared with the displacement reaction. This is related to the larger driving force for

the initial reactions and earlier completion of the overall reaction allowing longer time for

densification of the final phases.

The restrictions imposed by the stoichiometry of the reactions were surmounted

by adding the appropriate compound powders to the charge prior to consolidation. It was

found that no significant changes in the microstructure or densification were observed by

adjusting the matrix-reinforcement ratio proving that this is a viable processing route for

overcoming the stoichiometric constraints imposed by the reactants selected.

The effect of the initial powder size of the metallic components on the

microstructure is shown specifically, the initial size of the molybdenum powders led to

MoSi2 regions of similar size. Likewise, the size of the initial titanium powder particles

resulted in partially-reacted final microstructures, this is related to the fact that the

formation of TiB2 is the slowest reaction of all the sub-reactions that occur to produce the

final MoSi2-TiB2 composites.













CHAPTER 1
INTRODUCTION


1.1 Background



Intermetallics are playing an important role in the development of new materials

able of sustaining the escalating demands of the aerospace industry. A significant

improvement in weight, operating temperatures or mechanical performance is required

for materials to be considered as replacements in the most demanding applications.

Molybdenum disilicide is one such compound that has potential for high temperature

applications [1]. Its most attractive properties are its high melting point (20200C) [2],

reasonable density (6.24 g/cm3) [3], and excellent high temperature oxidation and

corrosion resistance. However, low ambient fracture toughness and loss of strength at

elevated temperatures have been the most significant limitations to the use of MoSi2 in

structural applications.

The more promising solutions for improving the mechanical properties of brittle

intermetallics such as MoSi2 are based on the incorporation and control of secondary

phases. To date, the artificial introduction of reinforcing phases or their generation via in-

situ reactions have been explored. However, in-situ processing of composites is the route

that provides (1) the best possibility of producing thermodynamically stable

microstructures, (2) the opportunity to eliminate undesirable phases, (3) the potential for








near-net-shape fabrication, and (4) the tendency to be more economical [1]. Nonetheless,

the limitation that the stoichiometries of the reactants and products impose on the volume

fraction of the reinforcing phase that can be generated and the inability to control its

morphology are disadvantages that need to be overcome.

The effect of particle size on the sintering-densification processes, phase

formation sequences, and microstructural scale of both the matrix and reinforcement

phases in composites has been evaluated to some extent in recent years. The proper

selection of particle size and particle size distribution have been found to be determinant

factors in the densification of the composites. In addition, particle size is one of the most

important controlling parameters for the reaction rate in these reaction sintering

processes.

The size and characteristics of the final microstructures obtained through solid-

state reactions depend significantly on the size and distribution of the particulate

reactants. Many studies on molybdenum disilicide composites have demonstrated that

particle size affects significantly the grain size of the MoSi2 that is formed during

processing. Thus, the mechanical properties of the composite are affected since properties

like the fracture toughness of the composite are strongly influenced by the grain size of

the MoSi2 matrix.

The rate of product phase formation and the path of the evolution of phases in

solid-state reactions are also significantly affected by the particle size of the reactants. In

several studies, it has been shown that particle size has a significant effect on the

formation rate of certain phases in ceramic composites, and metal-ceramic composites.

As expected, a faster formation rate is observed when the particle size of the reactants is








smaller, however, the introduction of impurities such as oxides formed at the surface of

the particles is more significant. Thus, particle size is a critical issue in shortening the

times involved in certain processing routes. However, the evolution of phases is also

significantly affected and sometimes leads to the formation of intermediate phases that

substantially affect the microstructural characteristics of the composites.

For a long time, borides have been among the phases contemplated for reinforcing

MoSi2. They have, in general, high hardness, good corrosion resistance, good thermal and

electrical conductivity, and relatively low CTE. Thus, some of the borides that have been

considered are ZrB2, HfB2, MoB, Mo2B5, and TiB2. Among them, the best compromise in

density and improvement in the mechanical properties has been achieved by compositing

MoSi2 with TiB2.

The Mo-Si-Ti-B system has been studied previously with the aim of producing

MoSi2-TiB2 composites through in-situ reactions. Based on an earlier study on

displacement reactions in the Mo-Si-B system, the displacement reaction approach in the

quaternary system was selected with the aim of using the findings on the evolution of

phases in the ternary case. Since the formation of MoSi2-molybdenum boride composites

was studied using molybdenum silicides and boron as reactants, molybdenum and

titanium silicides were chosen as reactant materials in the quaternary displacement

reaction approach. In order to elucidate the dependence of the microstructure on the

nature of the reaction, an elemental powder reaction was also evaluated.

In Costa e Silva's work [1], a very fine microstructure was obtained using the

displacement reaction approach (Figure 1.1). Since the stoichiometry of the selected

reactants led to approximately 60 volume % of TiB2 in the final products, the addition of








pre-reacted MoSi2 was made in order to obtain a 60 volume % MoSi2 matrix. This

addition led to even finer microstructures (Figure 1.2).


Figure 1.1 Microstructure (SEM) of MoSi2-TiB2 composites produced through in-
situ displacement reaction, without any pre-reacted addition.


Figure 1.2 Microstructure (SEM) of MoSi2-TiB2 composites produced through in-
situ displacement reaction, with the addition ofpre-reacted MoSi2.


_PALLL 'I. 'A








In addition, substantial differences in the scale of the microstructures were

obtained using the two different reaction approaches with the displacement reaction

route leading to the finer microstructures (Figures 1.1 and 1.3).



















Figure 1.3 Microstructure (SEM) of MoSi2-TiB2 composites produced though in-
situ elemental reaction, without any pre-reacted addition


The microstructures obtained were evaluated using SEM, TEM and EPMA in an

attempt to elucidate the microstructural evolution. However, phase identification in the

samples was made difficult by the small scale of the microstructures obtained which was

too small for the spatial resolution of the electron-probe microanalysis (EPMA). In the

case of the partially-reacted samples, in most cases, the low strength of the samples

prevented the preparation of samples for transmission electron microscopy (TEM).

In his study, it was impossible to clearly define whether the reactions proceeded

through the formation of intermediate products although it was concluded that the

conversion from three and four reactants to two products almost mandates the existence

of some intermediate steps or phases.








1.2 Approach



This study was undertaken in order to further our understanding of the phase

evolution sequences and their dependence on the nature and size of the reactants in the

production of MoSi2-TiB2 composites. Specifically, it was of interest to understand more

clearly the reasons behind the remarkable differences observed in the microstructures of

the composites produced by Costa e Silva using different reaction approaches. The goal

was to better quantify the reaction paths and to develop improved strategies for

processing such composites using in-situ reactions. In order to accomplish this task,

larger powder sizes were selected for investigation since, as noted above, the structures

that result from using mechanically alloyed powders were too fine for easy

characterization.

Thus, the mechanical alloying step was replaced by low energy tumbling of the

powders not only to avoid the significant comminution of the powders that takes place

during mechanical alloying, but also to reduce the introduction of impurities, such as

iron or alumina from the milling media, to the system.

The reaction approaches used in the previous study were not changed due to the

possibility of using some of the preliminary results and reduce the significant complexity

involved in the analysis of a quaternary reaction system. However, control of the initial

particles sizes was considered to be a necessity in order to reduce the number of variables

affecting the evolution of the composite structures and to compare the different effects

independently.








Thus, the further investigation of this system stems from the need to elucidate the

role of parameters that have been insufficiently explored. The roles of particle size and

nature of the reactant powders on

(1) the optimization of the processing (sintering-densification) of these

composites,

(2) the transformation sequences from different reaction approaches, and

(3) their effect on the final microstructures

need to be addressed with the goal of identifying improved processing strategies for these

composites. While the results may appear to be specific to the Mo-Si-Ti-B system, they

are more generally applicable to other silicides which rely on powder processing.















CHAPTER 2
LITERATURE REVIEW


2.1 The Mo-Si System




Molybdenum disilicide is one of the three intermetallic phases present in the Mo-

Si binary system (Figure 2.1) [2] along with MosSi3 and Mo3Si. Recently it was shown

that the C40(p)-*Cl 1b(a) polymorphic transformation does not occur in binary MoSi2

and was probably the result of contamination [4].


oU
0
2000


I-


0 10 20 30 40 50 60 70 80 90 100
Mo Atomic Percent Silicon Si


Figure 2.1 Mo-Si phase diagram assessed by Gokhale
and Abbaschian [2].








2.1.1 History of Compositing Efforts of MoSi2



In the early 1950s, silicides were already known by their distinctive high melting

points. Among them, molybdenum disilicide showed the appropriate high temperature

oxidation resistance and electrical resistivity to be considered for use as a heating element

by the Kanthal Corporation in Sweden [5]. Thus, a resistance heating element "Kanthal

Super"- constituted almost entirely by MoSi2 was patented.

Simultaneously, Maxwell [6] investigated molybdenum disilicide as a potential

structural material for high temperature applications. His work included the first

compositing efforts for overcoming the limited mechanical properties of this silicide.

Low ambient fracture toughness and loss of strength at elevated temperatures

have been the most significant limitations to the use of molybdenum disilicide in

structural applications and compositing appeared to be the most promising route to

surmount these drawbacks.

The early compositing attempts of Maxwell to reinforce MoSi2 with alumina were

followed by the work of Nowotny and Brukl [7,8]. These efforts led to the development

of a series of ternary phase diagrams of many silicide-based systems.

Alumina and silicon carbide were the reinforcements most widely studied in the

early 1970s by Fitzer revealing some improvement in elevated temperature strength

[9,10]; however, in general no significant enhancement was obtained in MoSi2 properties.

This lack of success led to a period of several years in which MoSi2 received little

attention from the scientific community








Advances in the processing of materials led to a renewed interest in molybdenum

disilicide and its composites in the late 1980s and early 1990s. Most of this work was

focused again on composites containing either silicon carbide or alumina. No significant

improvement in the performance of molybdenum disilicide was achieved compared with

some of the other intermetallics being considered for similar applications such as TisSi3,

TiAl, SiC, and NiAl. Even so, molybdenum disilicide remains as one of the intermetallics

with the greatest commercial success with applications in heating elements, glow plugs in

large diesel engines, and as a major constituent in the insulating tiles on the space shuttle,

among others [9,11].



2.1.2 Crystal Structure and Mechanical Properties of MoSib



Molybdenum disilicide has a tetragonal Cl lb crystal structure (Figure 2.2) and is

characterized by mixed covalent-metallic bonding which limits its ductility and toughness

due to the corresponding high Peierl's stress and the limited number of active slip

systems. This constitutes the main obstacle for MoSi2 in structural applications since

these characteristics preclude the possibility of any significant plastic deformation in this

material.

In a comparative overview of molybdenum silicide properties, Vasudevan and

Petrovic [10] summarized the mechanical properties of MoSi2 pointing out its negligible

ductility and fracture toughness from room temperature to 10000C, a brittle-to-ductile

transition temperature (BDTT) around 10000C, and high temperature strength governed

by plastic flow above 10000 C. The BDTT in polycrystalline MoSi2 is still uncertain; it








has been reported to be between 1000 and 13000C, but recent studies [12] suggest that it

could be even higher.



2.1.3 Role of Silica



Compositing MoSi2 appears to be the most viable route towards its structural

application. The nature of the reinforcements and the processing itself seems to be the

key to achieving microstructures capable of better mechanical performance. However, the

compositing efforts have to consider the limited homogeneity that can be achieved in

MoSi2 using powder processing.


0.7845 nm


Figure 2.2 MoSi2 C lb crystal structure.








The presence of silica in powder-processed MoSi2, due mainly to surface

oxidation, not only has an important detrimental role on the mechanical properties, poor

strength of polycrystalline MoSi2 at higher temperatures has been attributed to the

presence of silica in the grain boundaries [9,10], but consequently leads to the formation

of MosSi3 because of the narrow stability range of MoSi2.

The silica present at the grain boundaries has been found coalesced into globular

shapes rather than wetting the grain boundaries when molybdenum disilicide is

consolidated at high temperatures. This indicates a high surface energy between MoSi2

and SiO2 [13].

Therefore, the processing approaches for MoSi2 composites have to account for

the presence of these two additional phases and their potential interaction with selected

reinforcements. Thus, processing routes leading to the elimination of these phases may

provide an improvement in the high-temperature mechanical properties.

Most of the attempts to eliminate silica from MoSi2 involve the use of chemical

reactions. Thus, the reduction with carbon and aluminum leading to the formation of

silicon carbide and alumina and are among the approaches being used [10,13].

The use of high purity powders and a clean processing and consolidation has also

been explored. However, this route has been the less studied since it implies excessive

processing costs should commercial application be envisaged. Thus, the approaches

involving in-situ reactions seem to be the most plausible routes for producing silica-free

MoSi2 composites.








2.2 Reinforcing Approaches



2.2.1 Artificial Compositing vs. Solid-State Reactions



Two basic approaches have been attempted thus far in the processing of MoSi2

composites, namely, artificial compositing and in-situ reactions. For artificial

compositing, the reinforcement is introduced into the material as a previously-formed

component. In-situ reactions, on the other hand, rely on the formation of the matrix and

the reinforcements during heating and/or consolidation.

The status of artificial composites of MoSi2 has been reviewed recently by

Vasudevan and Petrovic [10]. Their review shows that, of all the composites explored,

MoSi2-SiC composites exhibit the best high temperature properties, though negligible

toughness. A review by Costa e Silva [13] on the use of refractory metals as

reinforcements indicates that attempts to address the low toughness by artificial

compositing using refractory metals appears to have begun with the work of Kehr. In

addition, further work by Fitzer and coworkers showed that interaction between MoSi2

and refractory metals occurs by growth of intermediate layers of silicides.

In efforts to avoid these detrimental reactions, oxide diffusion barriers (A1203,

ZrO2, and Y203) [14-16] were applied to the refractory metals and were observed to

prevent the interactions only if the coating was not damaged during the handling and

consolidation steps; obviously, this is difficult to avoid on a large scale. Thus, the

resulting interfacial reactions due to the lack of thermodynamic equilibria between the

matrix and reinforcement phases, and the difficulties involved in the processing to








prevent these interactions, i.e. coatings, are the most important drawbacks of artificial

compositing, especially in composites for high temperature applications were

interdiffusion is so prevalent.

In order to overcome these obstacles, in-situ reactions have been used to produce

composites that are more thermodynamically stable. To date, two different kinds of in-

situ reactions have been used (1) a direct reaction of the elements involved in the

composite to be produced and (2) displacement reactions involving compound phases.

Displacement reactions are solid-state diffusional phase transformations that

involve at least one compound among the reactants. The reaction is based on the

substitution of one of the elements constituting the compound by another element in the

reactive system to produce new product compounds that are more thermodynamically

stable than the starting reactants. The existence of more than one compound among the

reactants, and the possibility of the existence of intermediate phases, adds more

complexity to the interchange of the elements but the basis of the process remains.

In-situ MoSi2 SiC composites have been processed by Jayashankar and

Kaufman [17-19], Henager, Brimhall and Hirth [20], and Henager and coworkers [21]

among others. All these processing efforts involved the elimination of silica through a

displacement reaction and the simultaneous generation of the reinforcement. In some

cases, the reaction of elemental reactants was explored and the microstructural

differences obtained by both kinds of reactions analyzed. The work of Jayashankar [22]

led to an encouraging improvement in the control of the reinforcement loading, matrix

grain size, and an understanding of microstructural effects on high temperature strength

and deformation in this system.








The compositing efforts using in-situ reactions revealed the main advantages of

this processing approach which include

(1) the possibility of producing thermodynamically stable microstructures,

(2) the opportunity to eliminate undesirable phases such as silica,

(3) the potential for near-net-shape fabrication [23], and

(4) the tendency to be easier and more economical [13].

However, two common disadvantages of the in-situ reactions compared with artificial

compositing are the limitation that the stoichiometries of the reactants and products

impose on the volume fraction of the reinforcing phase that can be generated, and the

inability to control the reinforcing phase morphology [23, 13].



2.2.1.1 The Mo-Si-B system

For a long time, borides have been among the phases contemplated for reinforcing

MoSi2. They have, in general, high hardness, good corrosion resistance, good thermal and

electrical conductivity, and relatively low CTE [24, 25]. Thus, some of the borides that

have been considered are ZrB2 [26, 27], HfB2 [27], MoB, Mo2B5 [12], and TiB2 [12, 23,

26, 27, and 28].

Both, artificial compositing and in-situ reactions have been used to process

MoSi2-molybdenum boride composites. In the case of the in-situ reactions, many

combinations of reactants are possible from elemental powders to displacement reactions

involving one or more compounds. However, in the case of displacement reactions, the

reactions between boron and the molybdenum silicides allow the formation of a large

volume fraction of MoSi2 and, therefore this has been the approach used most frequently.








The displacement reactions that lead to the formation of molybdenum borides

from MosSi3 are accompanied by the formation of molybdenum disilicide due to the

molybdenum loss to form the borides. These reactions were studied by Costa e Silva [13]

as part of an investigation in the processing of in-situ composites. The phase formation

sequence in a MosSi3/B diffusion couple was studied and the order found in the formation

of phases was B-Mo2B5-(MoSi2-Mo2Bs)-MoB-(MoSi2-MoB)-Mo5Si3.

Artificial compositing has also been explored in the processing of these

composites. The artificial compositing ofMoSi2 and MoB5 by Shobu and coworkers [29]

resulted in the presence of a larger number phases after consolidation, namely, MoSi2,

MosB2, MoB, SiO2, and MosSi3. The presence of SiO2 and MosSi3 were attributed to the

oxidation of the MoSi2 starting powders. Also, it was observed that the artificial addition

of Mo2B5 to MoSi2 prevented significantly the grain growth of the MoSi2 phase.



2.2.1.2 The Mo-Si-Ti-B system

Among the boride-MoSi2 composites, the best compromise in density and

improvement in the mechanical properties at high temperatures seems to be achieved by

compositing MoSi2 with TiB2 [27]. Aikin produced MoSi2 composites with 15-45% TiB2,

ZrB2 or HfB2 using the XDTM process [12,27,28]. As mentioned earlier, the best

compromise in density and improvement in the mechanical properties seems to be

achieved by compositing MoSi2 with TiB2 (Figure 2.3, Table 2.1).

In the XDTM process, the reinforcing phase is generated in a melted matrix which

requires temperatures near the melting point of MoSi2 (2020C) during processing








[30,31]. The resulting microstructures appeared to corroborate the general acceptance that

these phases are thermodynamically compatible at the temperatures of interest.








S300



? 200
S30 % TIB2



S 00 oo m 30 % HflB,


O UNREINFORCED
0 . | .
0 200 400 600 800 1000 1200 1400
TEMPERATURE (C)

Figure 2.3 Compressive yield stress as a function of test temperature for MoSi2 reinforced
with 30 vol% of the indicated particles [27].


Table 2.1 Measured Density of MoSi2-Based Alloys [27].
Nominal Composition Density (g/cm')

MoSi2 6.2

MoSi2 + 30vol% SiC 5.3

MoSi2 + 30vol% TiB2 5.7

MoSi2 + 30vol% ZrB2 6.2

MoSi2 + 30vol% HfB2 7.7








The compatibility of MoSi2 and TiB2 has been evaluated by several authors.

Meschter and Schwartz assessed this system using the program SOLGASMIX and the

thermodynamic-properties database from the Facility for the Analysis of Chemical

Thermodynamics (FACT) at McGill University, Montreal. Their results indicated that

MoSi2 is thermodynamically compatible with TiB2 [26] although these results were not

confirmed experimentally.

As mentioned in the statement of the problem, a recent thermodynamic

assessment of the phase equilibria in the quaternary Mo-Si-Ti-B system at 14000C

showed that it is possible to produce MoSi2-TiB2 composites using in-situ displacement

reactions [23] and that an equilibrium between MoSi2 and TiB2 exists at this temperature.

The proposed quaternary isotherm at 14000C shows these two phases exist in equilibrium

next to a MoSi2-TiB2-MoB-Mo2B5 four-phase field (Figure 2.4).

As mentioned earlier, the extremely fine microstructures obtained using different

reactant mixtures hindered the ready evaluation of the phases present in the final

microstructures although X-ray diffraction indicated that the majority phases were indeed

MoSi2 and TiB2.

Thus, two in-situ reactions have been used in the processing of these composites

leading to fully reacted samples [13], an elemental reaction

Mo + 2Si + Ti + 2B = MoSi2 + TiB2

and a displacement reaction.

TisSi3 + 3/7MosSi3 + 10B = 15/7MoSi2 + 5TiB2




















MoB


TIB2


Figure 2.4 Schematic of a part of the calculated 14000C isotherm in the Mo-Si-Ti-
B phase diagram. The four phase field MoSi2-MoB-Mo2B5-TiB2 is indicated, as well as
the B-MosSi3-TisSi3 plane, locus of the compositions produced by the in-situ
displacement reaction study [13].




2.3 Phase Evolution in In-situ Reactions



The mechanisms and transformation sequences that take place during in-situ

reactions must be understood in order to develop reliable strategies in composite design.

The effects of contaminants and impurities on these mechanisms and formation








sequences are expected to be significant and, therefore, these topics will be addressed in

the following two sections.



2.3.1 Mechanisms and Transformation Sequences



The studies on transport-controlled oxidation by Wagner [32] provide the

groundwork for describing the mechanisms of evolution of phases in some in-situ

reactions. Wagner established a criterion for the stability of a flat growth interface based

on the diffusion processes occurring during the oxidation of alloys involving noble

metals. The criterion states that if diffusion in an alloy is rapid compared with diffusion

in the oxide of the less noble metal of the alloy, an oxide layer of uniform thickness is

stable, otherwise a rough oxide-alloy interface will form.

The criterion used by Wagner was adapted by Rapp and coworkers [33,34] in the

case of simple displacement reactions in ternary systems involving metals and their

lowest oxides. They were able to predict morphologies for different reactions and classify

them according to the product morphologies. Thus, two morphological classes were

defined: layered and aggregated and, within the aggregate class, two modifications were

proposed: lamellar and interwoven.

Rapp and coworkers studied the morphological characteristics of interfaces and

modeled the growth kinetics of the product zones [35]. However, their results were

disputed some years later by Backhaus-Ricoult and coworkers [36,37] who developed a

theoretical analysis in a study of morphological stability in ternary systems, based on

diffusion equations under appropriate boundary conditions. Their theoretical approach








was corroborated by their experimental results and it questioned the criterion proposed by

Rapp and Yurek as being not general in application.

Still, Rapp's models have been used by other investigators to rationalize their

results over the years. For example, Henager and coworkers [19, 20] prepared diffusion

couples to study potential reactions for producing composites in the ternary systems

Mo-Si-C and Ni-Al-O. The interpretation of their results is qualitative in nature and

mostly based on Tangchitvittaya, Hirth and Rapp's results on cooperative-precipitation

reactions of iron and nickel oxide [38].

The mechanisms of evolution of the phases of interest in the present investigation

have not been studied in any system combining MoSi2 and TiB2 or analyzed by in-situ

reactions including both phases. Independent studies have been attempted on elucidating

the formation of these compounds in other reaction systems or by using different reaction

techniques such as "self-propagating high-temperature synthesis" (SHS).

The work of Jayashankar and Kaufman [17] is among several recent

investigations aimed at understanding the mechanisms of evolution of phases during in-

situ reactions in MoSi2 composites. They studied the phase evolution of mechanically

alloyed mixtures of Mo and Si, and Mo, Si and C at several temperatures. Their results

showed that some MoSi2 forms during mechanical alloying and that its formation is

significant in the range of 600 to 11000C. Their results also indicate that MosSi3 and the

Nowotny MosSi3C phase are formed as intermediate compounds.

The mechanisms and kinetics of boride formation have also being studied by

several investigators, especially due to the interest in the synthesis of TiB2 as a refractory

material [39, 40, 41, 42]. Holt and coworkers studied the mechanism of reaction of Ti and








B to produce TiB2 by SHS [39]. In these processes, the solid reactants, when ignited,

spontaneously transform into products due to the highly exothermic nature of the

chemical reaction [43].

The results of Holt indicate that a reaction to produce a 50% mixture of TiB and

TiB2 is characterized by a non-planar reaction front leading to a layered-type structure

in the final product. In the case of the reaction to obtain 100% TiB2, the reaction front is

planar with no evidence of layering in the product.

They also analyzed the effect of size distribution of the reactant particles, and the

addition of pre-reacted products as diluents in the compacts. A significant lowering of the

combustion temperature and reaction rate was observed in the case of boron particles

larger than 150 pm, and also upon addition of pre-reacted TiB2 above 10%. It was

proposed that the reaction mechanism should involve only solid-state steps but no

specific steps were proposed and the kinetics of the reactions were not evaluated. These

results were corroborated by Popov and coworkers [40] in a study of the combustion

synthesis of TiB2.

Neronov and coworkers [41] used electron microscopy to study the mechanism of

the B-Ti interaction. The formation of two intermediate products has been proposed:

a Ti3B4 phase forming next to the boron particles, and TiB next to Ti. It is postulated that

the reaction proceeds by the dissolution of Ti through these product layers and its

further reaction with boron. At later stages of the interaction process the titanium

particles become saturated with boron and TiB2 is formed.

The work of Itoh and coworkers [42] established a time-temperature range for the

detection of a 20% of the X-ray diffraction intensity of the peaks obtained after complete








reaction of Ti and B to form TiB2 as 60 minutes at 1100-12000C, indicating the high

activation energy for the formation of TiB2 from its elemental constituents.

As the results of these previous investigations indicate, the kinetics and

mechanisms of the reaction between titanium and boron are not completely understood.

However, some conclusions seem to be clear:

(1) the activation energy for the reaction is probably high,

(2) considerable reaction rates occur only above 10000C,

(3) a solid-state mechanism seems to control the reaction, and

(4) the powder size and morphology probably play a significant role in the

mechanism and overall kinetics of the reaction.

In general, to understand the transformation sequence, the microchemistry in the

reaction zones between the phases and the physical and chemical compatibility of the

product phases are important since they are among the factors that will determine the

high temperature properties of the composites.

One of the factors that govern the transformation sequence is the relative

mobility of the components in the different phases present. In fact, the extent of the

reaction zones could be predicted if the interdiffusion coefficients of the species in the

different phases were known. The interdiffusion process through some of the phases is, in

the majority of the cases, the rate limiting step in the transformation sequences. This links

inevitably the evolution of phases to the kinetic aspects of the process [44].

The kinetic aspects of the diffusional interactions at the interfaces between the

intermediate and final phases in the system during the reaction clearly play a significant

role in the overall phase evolution. The initial phases that form in an interdiffusion








process can influence the subsequent nucleation and interdiffusion processes at that

interface [45]. The formation of phases that can act as diffusion barriers can significantly

reduce the reaction rates and the phase evolution process. Thus, the nature and

characteristics of the final internal interfaces generated in the composites are expected to

determine their performance and their stability at high temperatures.



2.3.2 Role of Oxygen



The analysis of the evolution of phases in any reactive system involves the

evaluation of all the possible combinations of the reactants leading to the formation of the

intermediate and final phases as the process goes along. The presence of impurities and

contaminants added to the system in any of the processing steps can play a crucial role in

the overall kinetics and phase evolution in a chemically active system.

One of the elements expected during powder processing is oxygen. The presence

of oxygen introduced to the system not only as surface oxides in the reactant particles but

also present in solution, as is usually the case of titanium, or in the environment is known

to affect significantly the paths of the phase transformations and the overall kinetics of

the in-situ reactions.

The role of oxygen in the formation of titanium silicides has been extensively

studied [46-49]. The formation and evolution of titanium silicide phases [50-51] and

their oxidation behavior [52-53] has also been analyzed. The interest in these compounds

comes from important technological applications in microelectronics such as ohmic

contacts, interconnections, and diffusion barriers, as well as in very large scale integrated








(VLSI) circuits. The studies on the formation of titanium silicides and the role of oxygen

in the process that are summarized in this section are expected to give some insight in the

effects of oxygen on the in-situ reactions in the Mo-Si-Ti-B system.

The oxygen behavior and its influence on titanium silicide formation was

analyzed through the study of the annealing properties of the systems TiO2/Si and

Ti/TiO2/Si by Berti and coworkers [46].

In this study it was found that titanium oxide, although more stable than any

titanium silicide and Si02, reacts in a silicon environment at temperatures up to 9000C to

form TiSi2 and SiO2 according to reaction (I).



3Si + TiO2 -> TiSi2 + SiO2 (I)



While this reaction is thermodynamically possible, it has some kinetic limitations.

The SiO2 layer that forms at the TiO2/Si interface (Figure 2.5), becomes a diffusion

barrier for silicon and restricts the completion of the chemical reaction.

When unreacted titanium is included in the system (Ti/TiO2/Si), it was observed

that, along with the formation of the SiO2 barrier, the oxygen from TiO2 is dissolved in

the metallic titanium as the temperature is increased before any silicide formation takes

place. The formation of an oxygen solid solution in titanium, when possible, is always

thermodynamically favored. Eventually, the controlled diffusion of silicon by the SiO2

barrier leads to the formation of Ti5Si3.

In general, in the presence of excess silicon, the most stable silicide phase has

been found to be TiSi2. However, in the cases in which some controlling kinetic








mechanism exists, as in this one because of the SiO2 diffusion barrier, TisSi3 is the first

silicide phase to form. As this silicide layer grows, it "pushes" the oxygen into the

unreacted titanium (called the 'snow-plow' effect). At higher temperatures, the

transformation from TisSi3 to TiSi2 occurs. The oxygen concentration in the titanium

increases until it becomes higher than the solubility limit and the precipitation of some

oxides takes place at the surface layer containing some metallic Ti and diffused silicon.




AS DEPOSITED
T,
Si piO, Ti 0

Si'O1
400 +600 C

Si Ti + 0



650 + 750 T
Si TiSi, I,



750 850 C
Si TiSi Tro1
Si


9i 00 SC
Si Ti SiI /


Figure 2.5 Schematic diagram of the Ti/TiO2/Si reactions [46].








One of the analysis techniques used by Berti was XRD in order to identify the

phases present. The analysis of such reaction sequences by x-ray diffraction is very

challenging since the expected oxide, silicide and titanium phases have overlapping

common or very close diffraction lines. In addition, any stresses generated by volume

changes were expected to affect the interplanar spacing causing significant shifts in the

diffraction lines.

A similar effect is caused by the presence of oxygen in titanium. Due to the high

solid solubility limit of oxygen in titanium (TiOo.s), a significant amount of oxygen is

expected to be absorbed by the unreacted titanium as the silicide forms. The presence of

oxygen distorts the hexagonal structure of the titanium leading to shifts in the diffraction

lines.

The study of Berti and coworkers on the effects of oxygen on Ti Si reactions

was completed by also proposing the formation of TiSi2 and Si02 in the case of titanium

oxide in a titanium-rich environment through the reaction (II).



9Si + TiO2 + 3Ti -- 4TiSi2 + SiO2 (II)



Thus, it was proposed that is possible to form titanium silicides by a reaction with

titanium oxides in spite of the greater thermodynamic stability of Ti oxide compared with

any titanium silicide or SiO2.

Some years later, some of Berti's observations were corroborated in a study by

Ottaviani and coworkers [47]. Ottaviani's work on oxygen dissolution in titanium








indicated that the formation of silicides leads to a complete segregation of oxygen into

titanium.

Quessinet and coworkers performed a phase analysis in Ti-Si diffusion couples

[48] and characterized the influence of oxygen on silicide formation [49]. They claim that

the first phase to appear is TisSi3. As significant diffusion takes place and a layer of

titanium silicides is formed, the kinetics of growth of the various phases are determined

by the interdiffusion coefficients within each phase. These diffusion coefficients are

sensitive to oxygen contamination since diffusion barriers might be created.

The silicide layers, as they grow, reject the oxygen into the unreacted titanium.

The eventual accumulation of oxygen in front of the reaction zone builds up a diffusion

barrier between Ti and TisSi3. Thus the binary interdiffusion system Ti/TisSi3 becomes a

ternary diffusion system (Ti/Si/O) affecting not only the diffusion coefficients but the

overall reaction kinetics [48].

The results of Quessinet [49] also indicated that the solubility of oxygen in Ti-rich

silicides is significant when compared with Si-rich titanium silicides like TiSi2 in which

oxygen is almost non-existent. Another significant aspect of this investigation was that no

evidence was found for the formation of a silica layer at the silicide-titanium interface

and that at long times and high temperatures, when significant diffusion takes place, the

diffusion of silicon is stopped when the concentration of oxygen in titanium reaches

certain limits. This occurs when the oxygen concentration in titanium is so high that

titanium is replaced by titanium oxides, mainly Ti203. Finally, a layer that seems to

consist of Si02, TisSi3 and TiO was found at the silicide/oxide interface.








Contrary to the Quessinet results, Heintze and coworkers [50] found a different

formation sequence for the silicides and additional steps in the silicide formation process

at higher temperatures with very high reaction rates. Specifically, in Ti/Si diffusion

couples, they observed the formation of TiSi2 at the beginning of the intermixing process

instead of Ti5Si3.

Similarly, it was found that this layer grows free of contamination, although

oxygen is present in the unreacted metal film. The 'snow plow' effect of the advancing

silicide front was also observed.

The concentration of oxygen present in the unreacted titanium increases as the

advancing silicide front completely repels the dissolved oxygen. At temperatures below

550C when certain concentrations are exceeded, the regular growth of the silicide stops.

This is associated with the formation silicon-deficient, oxygen-rich Ti-Si-O sub-layers.

These barriers limited the diffusion of silicon producing a general reduction in the silicide

reaction rate.

In addition, Heintze found that at high temperatures (above 6500 C), silicide

formation dominates over oxygen diffusion. The silicide front moves so fast that the

oxygen atoms being pushed do not have time to diffuse into the titanium. Thus, the

oxygen concentration increases at the silicide interface and forms Ti-Si-O sub-layers. The

Ti-Si-O sub-layers were found to be non-homogeneous and were expected to be a

mixture of TiSi2 and TiOx. The formation and composition of the Ti-Si-O sub-layers

were corroborated in a similar study by Wee and coworkers [51].

The formation of molybdenum silicides has also been studied by several

investigators [54,55] and the role of oxygen in the formation of the silicides has been








evaluated. However, many molybdenum silicide formation studies were carried out with

experiments that minimize the oxygen content with the purpose of evaluating the

evolution of phases per se [56].

Most of the studies in which molybdenum films have been deposited on silicon

substrates have resulted in the formation of only hexagonal and tetragonal MoSi2 (h-

MoSi2 and t-MoSi2) [55, 57-61]. This is clearly expected being t-MoSi2 the phase with

the most negative energy of formation in the Mo-Si phase diagram [62].

However, the presence of oxygen has been observed to lead to the formation of

other silicide phases. The formation of an oxide layer between a molybdenum film and a

silicon substrate has been found to determine the formation of Mo3Si by limiting the

diffusion of silicon [54]. In addition, excessive oxygen contamination in the molybdenum

film has been found to be the possible cause of the formation of MosSi3 in addition to

t-MoSi2 at temperatures over 8000C.

Although tetragonal MoSi2 is the first phase predicted to form [63,64], h-MoSi2

was observed to form initially. In a study by Doland and Nemanich [56], the formation of

these two phases was analyzed and it was found that h-MoSi2 was the first silicide

formed when the silicon substrate was sufficiently free of contaminants. Annealing of the

samples at temperatures of at least 7880C lead to the transformation from the hexagonal

phase to the tetragonal phase under "clean" conditions. The transformation to t-MoSi2

was accompanied by agglomeration of the silicide film into islands. Thus, Doland and

Nemanich suggested that the formation of h-MoSi2 is due to a lower silicide-silicon

interfacial energy which results in a lower nucleation barrier.








It should be pointed out that all of these Mo-Si and Ti-Si work was performed on

thin films, since the effects of impurities, nucleation and kinetics are critical in the initial

stages of thin film solid reactions, the extrapolation of these thin film results to bulk

samples must be done very carefully.



2.4 Phase Diagrams



In order to design an in-situ process, it is necessary to know which phases will be

in equilibrium under certain conditions. Thus, phase diagrams, if available, provide the

initial insight necessary for evaluating possible alternative routes for in-situ processing.

Unfortunately, the number of higher order systems reasonably well characterized is very

restricted and this imposes a limitation in the processing design with these systems.

Phase diagrams can be calculated if the thermodynamic properties of the phases

involved are known. However, the ability to predict or calculate thermodynamic data is

limited and relies on the existence of well-established binary systems. One method to

estimate the thermodynamic functions of phases is based on the knowledge of the free

energies of formation of the compounds involved. The phase diagram, in turn, can be

constructed via the concept of minimization of free energy or the "taut string" method

[23, 65].

Computer calculations of phase diagrams are continually being refined and utilize

existing experimental thermodynamic data and established phase diagrams. Of the

computer programs devised for phase diagram evaluations, the Thermocalc program

seems to be the most complete in describing higher order systems [66].








In the case of MoSi2-TiB2 composites, the following binary phase diagrams, that

involve the four elements playing a role in the system, are available in the literature:

Mo-Si [2, 67] (Figure 2.1), Mo-B [68,69] (Figure 2.6), Si-B [70] (Figure 2.7), Mo-Ti [71]

(Figure 2.8), Ti-Si [72] (Figure 2.9), and Ti-B [73] (Figure 2.10).


0204050 60 70


Atomic Percent Boron
80 90 95


0 10 20 30 40 50 60 70 80 90 100
Mo Weight Percent Boron B


Figure 2.6 The Mo-B Phase Diagram assessed by Spear and Liao [69].


100
















Atomic Percent Boron


0 1020 30 40 50
2200

2000 L

1800 -
o 14140C /-
S1600
3.2
4 1400- 3 1385
S-1.2 -1270
1200 r 0.4

1000
(Si)
800


60 70 80 90


100

I 2092C


Weight Percent Boron B


Figure 2.7 The Si-B Phase Diagram assessed by Olesinski and Abbaschian [70].



















Atomic Percent Molybdenum
10 20 30 40 50 60 70 8090100
S I I 1 1 1 1 ." O


2600
2400
2200
2000:
1800
1600
1400
1200
1000
800
600
400
0


Figure 2.8 The Mo-Ti Phase Diagram assessed by Murray [71].


10 20 30 40 50 60 70 80 90
Weight Percent Molybdenum


OfTi)
L
.-
L _-.--


16700C (3Ti,Mo)




8620C
~850C


(a~i)


100
Mo


-


1 r.UD


I-



















Atomic Percent Silicon
0 10 20 30 40 50 60 70


2200

2000

1800

1600
0
1400

~1200

1000

800

600
0
Ti


80 90 100


10 20 30 40 50 60 70 80 90 100
Weight Percent Silicon Si

Figure 2.9 The Ti-Si Phase Diagram assessed by Murray [72].



















Atomic Percent Boron


0 203040 50 60 70


90 95 100


3000 L

2500
U /u22000C
k 206020"C 21
g 2000

S/540+l 'C
1500
( i) ( )
STiB-
*? --TiB2
1000 66420C

500. -
0 10 20 30 40 50 60 70 80 90 100
Ti Weight Percent Boron B


Figure 2.10 The Ti-B Phase Diagram assessed by Murray [73].








The Thermocalc program was used by Kaufman [74] to corroborate and calculate

some of these phase diagrams, namely, the Mo-B, Mo-Ti, Ti-Si, Si-B, and Ti-B binary

systems. More recently, the 14000C Mo-Si-B isotherm proposed by Holleck, Benesovsky

and Nowotny [75-77] (Figure 2.11) was reproduced by Costa e Silva, who also calculated

the 14000C isotherm in the quaternary system Mo-Si-Ti-B [23].


Figure 2.11 Proposed 14000C isotherm for the Mo-Si-B system by Holleck,
Benesovsky and Nowotny [75-77].








The binary phase diagrams and the isotherm of the Mo-Si-B system show a

satisfactory agreement with the experimental diagrams. Another ternary system

potentially useful in the study of MoSi2-TiB2 composites is the partial 12500 C isotherm

of the Mo-Ti-Si system assessed by Svechnikov and Kocherzinsky, and also by Alisova

and Budberg [78, 79] (Figure 2.12).


TiAi4 + TiSp + r Moer + Tis M n %

Figure 2.12 Partial 12500C isotherm of the Mo-Ti-Si system [78,79].














CHAPTER 3
EXPERIMENTAL PROCEDURE


The experimental procedure used in this investigation starts with the selection of

the reactant materials and the analysis necessary in this stage. Once the raw materials

were selected, the processing of the samples included hot pressing of powders, heat

treatments and the preparation of diffusion couples. Finally, the analytical techniques

used in the study of the samples are addressed together with any special sample

preparation required for each specific technique.



3.1 Selection of Reactant Materials



A survey of the sizes and purity of the different commercially available elemental

and compound powders was done for a pre-selection of the size ranges to be used in this

study. Information about particle size distribution was requested from suppliers and

manufacturers. After the reactant powders and the different ranges of sizes were selected

and purchased, additional particle size analysis (PSA) of the powders was performed in

order to verify the information provided. To accomplish this, a COULTER LS 230

Particle Size Analyzer with a small volume module was used at the Particle

Characterization Laboratory of the Engineering Research Center (ERC).








The supplier's information including their data regarding average particle size

(APS) and the results of the PSA performed in this study are presented in Table 3.1.



3.1.1 Sampling for Particle Size Analysis



Samples of the powders to be analyzed (about 0.1-0.2 g) were taking directly from

the commercial containers and placed in plastic bottles in de-ionized (DI) water. The

bottles were manually shaken for about 5 minutes just before the analysis to assure

maximum particle suspension. Also, in order to breaking up agglomerates, the wetting

agent Triton X-100 (Lab Chem Inc.) was added to the water and the liquid media used

inside the analyzer. For each sample, the analysis was performed with and without a

wetting agent in order to obtain the most reliable values of the parameters being

measured. It was found that, in some cases, the average particle size decreased using the

wetting agent indicating that more agglomerates were broken. However, in some other

cases an increase in average particle was observed suggesting that the wetting agent was

inducing agglomeration instead.



3.1.2 Criteria for Raw Materials Selection



After the pre-selection of the powders based on the information provided by the

suppliers and an experimental preliminary evaluation of the average particles sizes,

additional factors were considered for the final selection of the powders. A significant

overlap of the distribution of sizes was considered in the final selection of the materials to








be combined. This criterion was used with the purpose of having the best possible

homogeneity of sizes in the reactant mixtures. The results with respect to the distribution

of sizes obtained for all the combinations of reactants chosen are presented in the results

chapter.


Table 3.1 Listing of raw materials
No. Raw Material Purity Indicated Size Supplier APS (plm) APS (gm)
(weight (Supplier) (ERC)
%)
1 Molybdenum 99.9 3-7 lm Johnson Not 9.91
Matthey Available
Electronics
2 Molybdenum 99.9 -150,+325 mesh CERAC Not 69.47
(w 110 44 gim) Specialty Available
Inorganics
3 Silicon 99.999 -325 mesh CERAC 7.18 1.91
(< 44 jm) Specialty
Inorganics
4 Silicon 99.999 -100,+325 mesh CERAC Not 104.3
(149 44 jm) Specialty Available
Inorganics
5 Titanium 99.98 -325 mesh CERAC 9.83 18.99
(< 44 jim) Specialty
Inorganics
6 Titanium 99.5 -150,+325 mesh CERAC Not 80.63
( 110 44 jm) Specialty Available
Inorganics
7 Boron 99 -325 mesh CERAC 5.88 6.51
(< 44 jm) Specialty
Inorganics
8 Boron 99.5 -60 mesh Alfa Not 48.49
(< 250 jm) AESAR Available
9 Molybdenum 99.5 -325 mesh Alfa 10 2.03
Silicide,MoSi (< 44 pm) AESAR
2
10 Molybdenum 99.5 -325 mesh CERAC 5.25 5.57
Silicide,MosS (< 44 jLm) Specialty
i3 Inorganics
11 Molybdenum Not Not CERAC Not 114.9
Silicide, Apply Apply Specialty Apply
Mo5Si3 Inorganics


























Two different size ranges were chosen (1) "small" included powders in the 0.04

and 40 pm range, and (2)"large" was the 40 to 200 glm range. Samples were

identified according to the following nomenclature: EPS to identify Elemental Powder

Samples or DRS in the case of Displacement Reaction Samples, proceeded by the size

range of the powder S for small and L for large, followed by the temperature at which

the sample was prepared and ending by HP, HT (heat treated) or HPT (hot press heat

treated) to indicate the heating process used. The EPSS samples were prepared using the

powder numbers 1, 3, 5 and 7; EPSL using 2, 4, 6 and 8; DRSS with 10, 12 and 7; and

DRSL with 11, 13 and 8 from Table 3.1, unless otherwise indicated.

An additional set of samples was considered for the evaluation of the effect of

metal size on the elemental reaction, and of the adjustment of the matrix-reinforcement

ratio in both types of reactions by addition of the appropriate pre-reacted powders. These

samples are described in Table 3.2.


Table 3.1 continued
No. Raw Material Purity Size Supplier APS (tm) APS (jim)
(weight %) (Supplier) (ERC)

12 Titanium 99.5 -325 mesh CERAC 5.68 12.31
Silicide,TisSi3 < 44 Im Specialty
Inorganics
13 Titanium Not N/A CERAC Not 111.3
Silicide, TisSi3 Apply Specialty Apply
Inorganics
14 Titanium 99.5 -325 mesh Johnson Not 9.24
Boride,TiB2 < 44 pim Matthey Available
Electronics








Table 3.2 Description of selected samples
Sample Description
EPSSMoL EPSS1400-HP powders substituting 1 by 2
EPSSTiL EPSS1400-HP powders substituting 3 by 4
EPSSPRMoSi2 EPSS1400-HP powders adding 9
DRSSPRTiB2 DRSS1400-HP powders adding 14


3.2 Sample Processing



The powders were mixed by tumbling in a glass or plastic vial. The individual

reactant powders were weighed in the appropriate proportions according to the following

two reactions for a total weight of approximately 8 g.



Mo + 2Si + Ti + 2B-+ MoSi2 + TiB2
(elemental reaction)

3/7Mo5Si3 + Ti5Si3 + 10B -* 15/7MoSi2 + 5TiB2
(displacement reaction)


The proportions of MoSi2 (matrix) and TiB2 (reinforcing phase) typically obtained

for these two reactions are 68:32 wt % for the elemental reaction and 48:52 wt % for

the displacement reaction. The silicide compounds used for the displacement reaction

were only available commercially in the sizes indicated in Table 3.1 (Nos. 10 and 12).

Therefore, pellets of those powders (MosSi3 and TisSi3) were hot pressed, ground and

sieved to prepare the displacement reaction samples corresponding to the large range of

sizes (Nos. 11 and 13 in Table 3.1). Before the grinding of the pellets the samples were








analyzed by backscattered electron microscopy and X-ray diffraction and after grinding

and sieving a particle size analysis was performed to verify the range of sizes obtained.



3.2.1 Hot Pressing of Blended Powders


The compaction of the powders by hot uniaxial compression was accomplished

using a Centorr Hot Press (HP) Model 600. The powder charges were loaded into

graphite dies of approximately 16 mm in diameter after the dies, bottom and top punches

were sprayed with boron nitride and the die was lined with graphite paper (Figure 3.1).

The assembly was placed into the chamber of the HP according to the schematic in

Figure 3.2 with no significant ram pressure applied. Temperatures for the HP of the

samples were selected in increments of 1000C starting at 10000C with the purpose of

following the process of formation and evolution of phases taking place.

Temperatures of 10000, 11000, 12000, 13000, 14000 and 15000C were reached for

different samples using both types of reactions. The HP chamber was under a vacuum of

<103 torr during the complete heating cycle. Except in the cases in which the final

temperature was 10000C, samples were heated to 10000C before applying a pressure of

-45 MPa via the hydraulic ram. For the 10000C samples, the pressure was applied at

about 750C with the aim of obtaining some degree of densification in the samples that

will allow their handling for analysis. In all of the cases it was necessary to continuously

adjust the ram pressure to keep a constant load on the sample. However, no sudden drops

of load were observed and no evidence of liquid phase formation were observed in the

dies after the hot pressing of the samples. Typical heating rates were on average





45


350C/min up to 10000C, followed by variable heating rates between 200 and 350C/min

from 1000 to 14000 or 15000C. For the samples used in the phase evolution study the

heating process was stopped at the temperature of interest and the sample was cooled

down and analyzed. In the case of the evaluation of the microstructure and densification

the samples were hot pressed at equal times (1 h) up to 14000 C.








STop Punch









Graphite Paper Discs Ai

| i Graphite Paper Lining


Sample (Powders)


-Graphite Paper Discs

Bottom Punch


Graphite Die


Figure 3.1 Schematic of sample and vial assembly for hot pressing.








The temperature was measured with a Capintec two-color optical pyrometer. This

pyrometer has been shown to be within + 250C accuracy when calibrated using different

previous opportunities with type B (Pt-Rh) [80] and types K, and C thermocouples [13]

before and during this experimental work. The heating was conducted under vacuum

while cooling was carried out initially under vacuum (approximately 1 h) and eventually

in argon at one-half atmospheric pressure.

Samples were removed from the HP chamber after approximately 2 h. They were

removed from the die using a hydraulic press and ground with silicon carbide paper to

remove the graphite lining and any reaction layer formed by reaction with the graphite

during processing.



Applied ram pressure




S- Hot Press Chamber
Conecting Rod

Susceptor
Induction Coil
Loaded Die

Graphite Cyilindric Block
Insulation Bricks

\, Graphite Base Block


Figure 3.2 Schematic of the assembly inside the hot press chamber








3.2.2 Heat Treatments


Two compacts, EPS 1000-HP and DRS 1000-HP, were sectioned in pieces of about

3 mm3 for heat treatment. Five samples in each case were placed in a horizontal tube

furnace under flowing gettered argon and removed at temperature increments of 100C

starting at 10000C up to 14000C. The heating and cooling rates were 5C/min The

samples were removed individually as the temperature reached the selected values and

were cooled in air on an aluminum block. Samples were mounted and individually

polished for analysis.



3.2.3 Diffusion Couples



Diffusion couples of MoSi2 and TiB2 were prepared by hot pressing individual

pellets of the compounds. The powders used were 9 and 14 from Table 3.1. The

compounds were vacuum hot pressed individually, MoSi2 up to 14000C and TiB2 up to

17000C. After the pellets were prepared, a 2 mm thick layer was removed from one of the

circular faces of each of them and the remaining pellet was polished to 5pm. These flat

polished faces were then placed together and hot pressed (45 MPa) at 14000 C and 15500

C for 3 hours.








3.3 Analytical Techniques



3.3.1 Metallographic Sample Preparation



Electric discharge machine (EDM) was used to cut the samples or in some cases a

low speed diamond saw was used. Standard metallographic techniques were used to

prepare the samples for analysis with scanning electron microscopy (SEM), backscattered

electron microscopy (BSEM), energy dispersive spectroscopy (EDS), and wavelength

dispersive spectrometry (WDS).

Samples were mounted in standard epoxy resin and ground on diamond wheels,

due to the exceptional hardness and brittleness of the phases present. They were then

polished with 6 p.m, 1 tm, and 0.45 [tm diamond paste. After polishing, the samples were

ultrasonically cleaned to remove any grinding or polishing contamination.

Secondary electron and backscattered electron images were obtained using a

JEOL 35CF SEM and a JEOL 6400 SEM, the latter equipped with an Oxford Link ISIS

EDS System. WDS microanalysis (EPMA) was performed on a JEOL 733

SUPERPROBE, using Mo, Ti, SiC, B, and Si02 standards.



3.3.2 X-Ray Diffraction Sample Preparation



Samples of the individual powders and the powder mixtures prior to hot pressing

were analyzed with a Philips APD 3720 diffractometer (XRD) without further

comminution. The powders were stuck to a glass slide with Amyl Acetate Colloidal, a





49


volatile organic solution. Hot pressed samples were examined by cutting 0.5 mm thick

cross section slices of the pellets. In some cases, these slices were ground to fine powders

using an alumina mortar and pestle. The slices were then attached to the glass slides using

double sided adhesive tape. Adhesive tape was also used to level the glass slide base with

the sample surface insuring correct diffraction angles.














CHAPTER 4
RESULTS AND DISCUSSION


4.1 Particle Size Analysis



The ranges of sizes for the raw materials used in this study were chosen based on

the following results of the analysis of particle sizes and particle size distributions.



4.1.1 Elemental Reaction Powders



In the case of the elemental powder reactions, the reactants chosen for both ranges

of sizes (small and large) showed a significant overlapping of the particle size

distributions as can be observed in Figures 4.1 and 4.2. Samples prepared for the analysis

of the effect of the metal size showed the distributions in Figures 4.3 and 4.4.

The effects of adding pre-reacted TiB2 to evaluate the consequences of adjusting

the matrix reinforcement ratio was performed using TiB2 powders with the distribution

shown in Figure 4.5 which was overlaid on the EPSS plot (Figure 4.1) to show the

significant overlapping of the distributions. Statistics of the particle size analysis of the

powders used in the elemental powder reactions are shown in Table 4.1. The

corresponding powder numbers in Table 3.1 are indicated.











Table 4.1 Elemental powders and titanium diboride statistics
Material/ Mean Mode Median S.D. 95% Conf. 90 %
Powder (tm) (tim) (gim) Limits (gtm) of
No. ______dist. <
Mo / 1 9.91 14.94 9.44 4.93 0.25-19.60 16.80


Mo/2 69.47 66.44 68.36 10.10 49.70-89.20 82.42


Si/3 1.91 1.92 1.45 1.65 0-5.13 4.44


Si/4 104.30 140.10 100.60 54.70 0-211.00 175.2


Ti/5 18.99 18.00 15.88 13.40 0-45.30 37.58


Ti/6 80.63 80.08 77.85 25.10 31.50-130.00 114.4


B/7 6.51 12.40 5.59 5.15 0-16.60 13.95


B / 8 48.49 72.95 53.06 24.80 0-97.10 77.45


TiB2 / 14 9.24 9.37 8.30 4.94 0-18.90 15.14

















c 4-
E
0 Si
> 3- I Ti

2-




0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 1000
Particle Diameter (pm)

(a)

8 -- Mo
Mo B
7- Si
Ti
6-

5-














(b)
E 4- Si
> 3-

2- /



0
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 1000
Particle Diameter (aim)

(b)

Figure 4.1 Distribution of particle sizes of the "small" elemental powders.
a) Graphic indicating the sub-ranges of sizes in the distribution actually measured
by the COULTER analyzer. b) Curves showing the overlap in the powder distributions.









30- -B
Mo Ti
Si
25- Mo


S20-
O3
E

13



X Si



0
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 1000
Particle Diameter (pm)
(a)
30 B
Mo Ti
SSi
25- -Mo


20-
(0
E
> 15-


10 Ti

Si
5-
B

0-
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 1000
Particle Diameter (pm)
(b)

Figure 4.2 Distribution of particle sizes of the "large" elemental powders.
a) Graphic indicating the sub-ranges of sizes in the distribution actually measured
by the COULTER analyzer. b) Curves showing the overlap in the powder distributions.




























0.1 0.4


Ti


I I I I I I
1 2 4 6 10 20 4
Particle Diameter (pm)
(a)


--Mo

--Si
--Ti


1I I I40
100 200 400


100 200 400 1000


Particle Diameter (pm)


(b)

Figure 4.3 Distribution of particle sizes using "large" molybdenum powders and "small"
Si, Ti and B powders to analyze the effect of metal size in the evolution of phases. a)
Graphic indicating the sub-ranges of sizes in the distribution actually measured by the
COULTER analyzer, b) Curves showing the overlap in the powder distributions.


2-

1-

0.-4
0.04


1000






55


14 Ti
Ti B
12- Mo
-- Si
10-

S8- Mo
Mo
E
4 6-







0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 1000
Particle Diameter (pm)
(a)

14- Ti
Ti B
12- Mo
Si
10-

8 Mo

6-



4-1
Si ..




0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 1000
Particle Diameter (pm)
(b)

Figure 4.4 Distribution of particle sizes using "large" titanium powders and "small" Si,
Mo and B powders to analyze the effect of metal size in the evolution of phases. a)
Graphic indicating the sub-ranges of sizes in the distribution actually measured by the
COULTER analyzer, b) Curves showing the overlap in the powder distributions.






56


8" Mo
-B
7- Mo Si
STi
6- TiB2- TiB2

5

cu 4
E
6- Ti
> 3
Si
2



0
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 10[
Particle Diameter (pm)
(a)

8-
TiB2 Mo M
2 -B




5-
74 SiSi
-\Ti




1 \
E 4- Si
0
>3-

2-/




0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 10(
Particle Diameter (pm)
(b)

Figure 4.5 Distribution of particle sizes for TiB2 overlaid on the distributions for the
elemental powders in the "small" range of sizes, a) Graphic indicating the sub-ranges
of sizes in the distribution actually measured by the COULTER analyzer.
b) Curves showing the overlap in the powder distributions.


30


10








4.1.2 Displacement Reaction Powders


In the case of the displacement reactions, the reactants chosen for the small and

large size ranges also displayed a significant overlapping of the particle size distributions

as can be observed in Figures 4.6 and 4.7. The effect of adding pre-reacted MoSi2 to

evaluate the consequences of adjusting the matrix reinforcement ratio was performed

using MoSi2 powders with the distribution shown in Figure 4.8 overlaid on the DRSS

plot (Figure 4.6) to indicate the similarity of the distributions. The statistics of the particle

size analysis of the powders used in the displacement reactions are shown in Table 4.2.

The corresponding powder numbers in Table 3.1 are indicated.


Table 4.2 Displacement reaction powders and molybdenum disilicide statistics
Material/ Mean Mode Median S.D. 95% Conf. 90 % of
Powder No. (pim) (p~m) (p.m) Limits (itm) dist. <
B /7 6.51 12.40 5.59 5.15 0-16.60 13.95

B /8 48.49 72.95 53.06 24.80 0-97.10 77.45

MoSi2 / 9 2.03 2.79 2.04 1.36 0-4.71 3.87

MosSi3 / 10 5.57 6.45 4.42 4.89 0-15.20 11.81

MosSi3/11 114.9 127.6 114.8 46.26 50.56-172.1 167.8

TisSi3 / 12 12.31 16.40 9.51 11.30 0-34.50 28.25

Ti5Si3/13 111.3 116.3 111.2 31.00 24.20-205.5 151.7


















cE Mo5Si3
E53
13 3-
0 Ti5Si

2-
O I 'I I n' I3



0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 10002000
Particle Diameter (pm)
(a)
-- Ti5Si3

B _- Mo 5Si3
6- B B

5-

4-



0 5 3
23 Mo5Si3/ Ti5Si3

2 -

1 --


0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 10002000
Particle Diameter (pm)
(b)

Figure 4.6 Distribution of particle sizes of the "small" displacement reaction powders.
a) Graphic indicating the sub-ranges of sizes in the distribution actually measured by the
COULTER analyzer. b) Curves showing the overlap in the powder distributions.









7- Ti5Si3
Ti5Si3 Mo5Si3
6- -B

5-


4- B

= 3-


2-


.Mo5 Si 3
0-1
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 10002000
Particle Diameter (pm)
(a)
Ti5Si3
Ti5Si3 Mo5Si3
6- rB

5- B i

4-

E 3-





Mo5Si3

0-_
0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 10002000
Particle Diameter (pm)
(b)


Figure 4.7 Distribution of particle sizes of the "large" displacement reaction powders.
a) Graphic indicating the sub-ranges of sizes in the distribution actually measured by the
COULTER analyzer. b) Curves showing the overlap in the powder distributions.






60


7- M Ti5Siq
MSi2 B Mo SI3
6- B
-- MoSi2



O 4-
55



SSi3
> 3-

2-





0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 1000
Particle Diameter (pm)
(a)
7-
Ti Si
MoSi 5 15Si3
02 B Mo5Si3
6 -B
-- MoSi2
25






2 T\5Si3




0-

0.04 0.1 0.4 1 2 4 6 10 20 40 100 200 400 1000
Particle Diameter (pm)
(b)
Figure 4.8 Distribution of particle sizes for MoSi2 overlaid on the distributions for the
"small" displacement reaction powders. a) Graphic indicating the sub-ranges of sizes in
the distribution actually measured by the COULTER analyzer. b) Curves showing the
overlap in the powder distributions.








4.2 Evolution of Phases



4.2.1 Displacement Reaction



The evolution of phases for the displacement reaction was evaluated initially

through a series of samples heat treated in the range 10000 14000C. The results of this

preliminary study led to the evaluation of an additional set of samples hot pressed at the

same temperatures considered for heat treatments. The results of the SEM evaluation

along with the results of EPMA and XRD analysis for these two sets of samples will be

addressed as separate topics.



4.2.1.1 Heat treated samples

The primary sample used in this evaluation was initially hot pressed at 10000C for

;30 min at 45 MPa using the same general conditions described in section 3.2.1. This

sample was sectioned to obtain the samples to be heat treated as described in section

3.2.2. The very low densification achieved during the hot pressing of the initial sample

(Figure 4.9) hindered significantly the diffusional processes necessary for the

intermediate reactions to take place. Thus, most of the reaction zones that can be

observed are limited (Figures 4.9 4.11) and are only of some significance above 1300 or

14000C (Figures 4.12 and 4.13). In addition, due to the poor densification of the samples,

it was not possible to polish them to the mirror finish desirable for EPMA. Thus, only

limited regions that were polished enough to allow reliable measurements were

examined.





























Figure 4.9 Scanning electron micrograph (SEM) of the sample DRSS1000-HT indicating
the points analyzed by EPMA.


Table 4.3 Electron probe microanalysis (WDS) of the points indicated in Fig. 4.9.
Points Moat. Siat.% Tiat.% Bat. % at.% Possible
_% ____Phase
(1) 59.04 39.93 0.15 0.89 0 MosSi3

(2) 0 1.26 0.08 98.67 0 B

(3) 0 37.86 62.14 0 0 Ti5Si3








As can be observed at 10000 C (Figure 4.9 and Table 4.3), the only phases

apparently present are the initial reactant powders MosSi3 (Point 1), TisSi3 (Point 3) and

B (Point 2) (Table 4.3).

No reaction zones of sufficient dimensions to be measured by EPMA were

observed. However, XRD of this sample showed not only the presence of the reactants

but also the formation of MoSi2 and the molybdenum borides MoB and Mo2B.

Due to the low free energy of formation of MoSi2, its appearance was expected at

the very early stages of the hot pressing process [21,13]. Thus, the formation of MoSi2 is

one of the first reactions to take place.

After heat treatment at 11000 C, the molybdenum borides were sufficiently large

(Figure 4.10) to be detected by EPMA (Table 4.4 Points 1 and 2). They appeared to

have a mixed composition of Mo, Si, and B since they are the products of the reaction

between MosSi3 and B and could not be resolved separately. On the other hand, titanium

silicide did not appear to participate in any reaction at this temperature. At 12000C

(Figure 4.1 land Table 4.5), the TisSi3 phase seems to be enriched in silicon and

approaches the TisSi4 composition (Table 4.5 Points 3 and 4).

The presence of molybdenum borides as a mixture with MosSi3 remains up to

13000C (Figure 4.12) as indicated by points 1 and 2 in Table 4.6. They appear as a

distinct phase at 14000C (Figure 4.13) and it seems that they evolved to MoB2 or that a

mixture of all the possible borides exists as reaction products, which can be assumed due

to the atomic percent values found in some of the points analyzed (e.g. Table 4.4 Point

1, Table 4.6 Points 1 and 2, Table 4.7 Points 1 and 2). The reactions observed in this

set of samples are summarized in Table 4.8.































Figure 4.10 Scanning electron micrograph (SEM) of the sample DRSS 1100-HT
indicating the points analyzed by EPMA


Table 4.4 Electron probe microanalysis (WDS) of the points indicated in Fig. 4.10.
Points Mo at. % Si at. % Ti at. % B at. % Oat.% Possible
Phases
(1) 27.25 18.22 0.12 54.4 0 ?

(2) 32.67 25.08 0.05 42.19 0 ?

(3) 0 37.06 61.84 1.10 0 TisSi3

(4) 0 37.50 62.5 0 0 Ti5Si3

(5) 0 2.53 0.25 97.22 0 B

(6) 0 2.35 0.03 97.62 0 B





























Figure 4.11 Scanning electron micrograph (SEM) of the sample DRSS 1200-HT
indicating the points analyzed by EPMA


Table 4.5 Electron probe microanalysis (WDS) of the points indicated in Fig. 4.11.
Points Mo at. % Si at. % Ti at. % B at. % O at.% Possible
Phases
(1) 35.72 53.95 0.26 7.34 2.72 MoSi2
(2) 26.3 23.1 0.17 39.17 11.25 ?
(3) 0.95 32.93 54.16 0 11.97 ?
(4) 0.2 41.23 58.57 0 0 TisSi4
(5) 0 2.14 0.05 97.81 0 B
(6) 0 2.64 0 97.36 0 B
(7) 0 2.58 0.01 97.41 0 B
(8) 0 2.59 0.05 97.36 0 B






























Figure 4.12 Scanning electron micrograph (SEM) of the sample DRSS 300-HT
indicating the points analyzed by EPMA.


Table 4.6 Electron probe microanalysis (WDS) of the points indicated in Fig. 4.12.
Points Mo at. % Si at.% Ti at. % B at.% O at. % Possible
Phases
(1) 20.38 12.86 0 66.67 0 ?
(2) 22.96 7.26 0.10 69.68 0 ?
(3) 0 43.25 56.75 0 0 Ti5Si4
(4) 0.01 43.69 46.59 9.70 0 ?
(5) 0 2.80 0.03 97.17 0 B































Figure 4.13 Scanning electron micrograph (SEM) of the sample DRSS1400-HT
indicating the points analyzed by EPMA.


Table 4.7 Electron probe microanalysis (WDS) of the points indicated in Fig. 4.13.
Points Mo at. % Si at. % Ti at. % B at. % O at.% Possible
Phase
(1) 23.78 10.07 0.09 66.06 0 ?
(2) 30.37 0.96 0.10 68.57 0 ?
(3) 2.40 34.79 56.55 0 6.26 ?
(4) 0 39.20 52.45 0 8.35 ?
(5) 0.01 0.15 24.74 48.39 26.71 ?
(6) 0.04 4.76 0.13 94.77 0.30 B
(7) 0.02 5.29 0.17 94.02 0.50 B
(8) 0.07 5.70 0.14 93.36 0.74 B








Table 4.8 Reaction observed in the heat-treated samples.
Reaction

3Mo5Si3 + 9B = 4MoSi2 + 7MoB + 2Mo2B + Si

TisSi3 + Si = TisSi4





The results of this preliminary evaluation were insufficient for determining the

evolution of phases taking place in the displacement reaction. The low densification of

the samples and the relatively short heat treatment times precluded the necessary

interdiffusion required for reaction zones to be observed and analyzed. Therefore, longer

heat treatments were necessary which could preclude the detection of some of the

intermediate steps of the reactions taking place. Thus, a similar set of specimens was

prepared by hot pressing samples at the temperatures used for the tracking of the reaction

through heat treatment.



4.2.1.2 Hot pressed samples

Samples were hot pressed for 1 h and then evaluated by SEM, EPMA and XRD.

In order to put some perspective on the analysis of this reaction system, a list of the

possible contact interfaces between the phases present was made for each temperature.

Thus, the possible reaction interfaces at room temperature in this system are listed in

Schematic 4.1.

The general microstructure obtained for the 10000C sample (Figure 4.14) showed

a relatively higher densification and, in general, the majority of the phases present were

the reactant materials (Table 4.9 -Points 1, 3, 4, and 6). The formation of TisSi4 (Point 2)








was observed in very small proportions close to the regions in which was detected the

formation of molybdenum borides and MoSi2, which should be accompanied by the

release of Si (Point 5). This is expected to occur according to the following reactions.

MosSi3 + 2B = MoSi2 + 2MoB + Mo2B + Si (1)

Ti5Si3 + Si = Ti5Si4 (2)


Figure 4.14 Backscattered electron micrograph of the microstructure of sample
DRSS1000-HP indicating the points analyzed by EPMA


Table 4.9 WDS of the points indicated in Figure 4.14.
Points Mo at. Si at. % Ti at. % B at.% O at.% Possible
% ____Phases
(1) 0.01 38.05 61.94 0 0 TisSi3
(2) 0.02 43.78 56.21 0 0 Ti5Si4
(3) 0 1.95 0.11 97.94 0 B








Table 4.9 continuation
Points Mo at. Si at.% Ti at.% B at.% Oat.% Possible
% ____Phases
(4) 61.23 35.73 1.90 1.15 0 MosSi3
(5) 39.65 25.36 0.51 34.47 0 ?
(6) 0 36.44 61.97 1.59 0 Ti5Si3


Room Temperature
No. of Phases: 3
No. of Possible Interfaces: 3
No. of Reacting Interfaces (*):1

Inter-faces Ti5Si3 B
MosSi3 X X
TisSi3 X


Schematic 4.1 Possible reaction interfaces at room temperature and 10000 C in the
displacement reaction system


The excess Si released by reaction (1) appears to result in the formation of TisSi4

(reaction (2)) only when TisSi3 regions lie adjacent to MosSi3 regions. In general, no

significant reaction zones were observed in the TisSi3 regions in comparison with the

MosSi3 regions (Figure 4.15).

The compositions obtained by EPMA and the specific identity of the intermediate

phases, such as the molybdenum borides, were unclear at this temperature. However, the

reaction zones were larger compared with the equivalent sample heat treated probably

due to the longer hot pressing time.

The reaction zones for the molybdenum boride formation are shown in Figure

4.16 and the corresponding results from the EPMA are presented in Table 4.10.








Unfortunately, the poor densification of this sample precluded proper polishing for

EPMA and some regions could not be examined properly.

The x-ray diffraction analysis of this sample corroborates the formation of

molybdenum borides, namely MoB and Mo2B, and the presence of MoSi2 (Figure 4.17).

It should be pointed out that the sequence of the symbols identifying the peaks in the

XRD plots has no significance in regards to which phase correspond the most. Evidence

of TisSi4 was not detected by x-ray diffraction indicating its limited volume fraction. The

new set of possible reaction interfaces at 10000 C is presented in Schematic 4.2.


Figure 4.15 Backscattered electron micrograph of the microstructure of sample
DRSS1000-HP showing the extent of the reaction found in MosSi3 compared to TisSi3.
















































(c)
Figure 4.16 Backscattered electron micrographs of the reaction zones developed in the
sample DSSS1000-HP indicating the points analyzed by EPMA. (a) Region showing the
formation of MoSi2 (Points 1, 2, and 3). (b) Region showing the formation of a
molybdenum-boron rich reaction zone (Points 1, 2, and 3) surrounding MosSi3. (c) A
significantly larger reaction zone (Points 1 and 2) is observed in this region of the
microstructure.








Table 4.10 Electron probe microanalysis (WDS) of the points in Figure 4.16.
Figures & Mo Si Ti B O Possible
Points at. % at% at. % at. % at% Phases
Figure (a)
(1) 30.22 62.53 0.14 7.11 0 MoSi2
(2) 32.73 67.06 0.20 0 0 MoSi2
(3) 32.51 65.61 0.10 1.79 0 MoSi2
(4) 35.46 19.85 0.20 44.49 0 ?
(5) 39.95 15.23 0.24 44.58 0 ?
(6) 32.56 17.63 0.15 49.66 0 ?
Figure (b)

(1) 38.99 22.86 0.12 32.45 5.59 ?
(2) 37.14 16.45 0.11 45.71 0.60 ?
(3) 32.31 35.76 0.15 31.78 0 ?
Figure (c)

(1) 37.11 20.06 0.10 42.73 0 ?
(2) 39.13 25.47 0.46 32.03 2.91 ?


The XRD results from sample DRSS1100-HP did not showed notable changes

compared to DRSS 1000-HP, except for a slight increase in the extent of the reactions in

the MosSi3 regions in the DRSS1100-HP material. Therefore, it was considered not

necessary to include results almost identical to the ones presented already.








1800
1600

1400
1200
1000
800
600

400
200


Figure 4.17 X-ray diffraction analysis of sample DRSS 1000-HP indicating the
presence of MoB, Mo2B and MoSi2.




Temperature: 10000C
No. of Phases: 6
No. of Possible Interfaces: 15
No. of Reacting Interfaces (*): 2


Inter- i5Si3 B MoB M02B MoSi2
faces
S X X X

_- X X X_ X

0n-- -x x x



=0213 X
-------^r^
Mo2-----



Schematic 4.2 Possible reaction interfaces at 10000 C in the
displacement reaction system.








The results for the sample DRSS1200-HP showed the beginning of the formation

of a TiB2 layer surrounding the former TisSi3 areas, now transformed to TiSi as can be

observed in Figure 4.18 and Table 4.11 Points 4,5,6, and 7. The formation of TiSi takes

place according to the following reaction:

TisSi3 + 4B = 2TiB2 + 3TiSi (3)

The formation of TiSi2 has also been confirmed by XRD and EPMA analysis

(Figures 4.19 and 4.20, Points 2 and 3 in Table 4.12). It may occur by reaction (4) or by

the further reaction of boron with TiSi (5), which seems to be the logical next step in the

process. A mixture of TiSi and TiSi2 is expected according to the Ti-Si phase diagram.

2TisSi3 + 14B = 7TiB2 + 3TiSi2 (4)

2TiSi + 2B = TiB2 + TiSi2 (5)


(7)



(6)
(3)
(4)
(5)




10 UM


Figure 4.18 Scanning electron micrograph (SEM) of the microstructure of sample
DRSS1200-HP indicating the points analyzed by EPMA








Table 4.11 Electron probe microanalysis (WDS) of the points indicated in Figure 4.18.
Points Mo at. % Si at. % Ti at. % Si at. % Oat.% Possible
Phases
(1) 0.1 49.08 48.69 2.14 0 TiSi

(2) 0.07 48.79 48.49 2.65 0 TiSi

(3) 0.16 48.22 48.46 3.17 0 TiSi

(4) 0.17 0.61 30.48 68.44 0 TiB2

(5) 0.26 2.02 28.24 69.48 0 TiB2

(6) 0.14 2.04 29.72 67.39 0.7 TiB2

(7) 0.19 1.91 33.33 64.57 0 TiB2


S2000
L.J!


20 40 60 8o [20] 100


Figure 4.19 X-ray diffraction analysis of sample DRSS1200-HP indicating the presence
of TiSi and TiSi2.






























Figure 4.20 Backscattered electron micrograph of the microstructure of sample
DRSS1200-HPindicating the points analyzed by EPMA


Table 4.12 Electron probe microanalysis (WDS) of the points indicated in Figure 4.20
Points Mo at. % Si at. % Ti at. % B at. % O at.% Possible
Phases
(1) 0.01 47.68 48.00 4.31 0 TiSi
(2) 0.07 63.06 35.48 1.39 0 TiSi2
(3) 0.11 60.70 37.19 2.00 0 TiSi2
(4) 0.06 49.84 50.10 0 0 TiSi
(5) 0.23 0.82 22.70 72.71 3.54 TiB2
(6) 0.37 0.34 30.70 66.70 1.89 TiB2
(7) 0.03 38.77 56.55 0 4.65 ?
(8) 0.02 34.63 47.04 18.22 0.08 ?
(9) 0.08 37.81 53.70 0 8.41 ?
(10) 0.02 37.34 48.98 13.66 0 ?








The final reaction of TiSi2 to form TiB2 can be described by reaction (6):

TiSi2 + 2B = TiB2 + 2Si (6)

The silicon released appears to react with oxygen to form a glassy silica phase at

the center of the former TisSi3 regions in the completely reacted samples at 14000 and

15000C (Figure 4.21). The formation of this silica at the center of these eventual TiB2

regions can be explained by the reduction of the solubility of the oxygen present in the

system from the intermediate phases to the final products.

The oxygen introduced in the system as oxides in the surface of the reactant

particles, is dissolved or rejected in the different phases depending on the solubility limits

characteristic of this element in each phase. Assuming that the solubility of oxygen in the

intermediate titanium silicide phases is lower as they become Si-rich silicides [49], the

oxygen should be segregated towards the center of the particles. Thus, the silicon

released by reactions like (6) in a high oxygen concentration environment is expected to

react with oxygen to form silica since this reaction is thermodynamically favored

compared to the diffusion of Si through TiB2 back to the matrix to react with other

phases. However, after all the oxygen has been consumed, some of the silicon released

through reaction (6) is believed to take part of further reactions in the system.

The formation of TiB2 occurs according to two different reactions. In the case of

reacted-Mo5Si3 TisSi3 adjacent regions, the decomposition of the molybdenum borides

releases part of the boron necessary for TiB2 formation (7) (Figure 4.22). Otherwise the

reaction of TisSi3 to form TiSi (3), TiSi2 (4) and (5), and the eventual reaction with

boron takes place (8).

Mo2B + 7MoB + Ti5Si3 + Si = 4TiB2 + 2MoSi2 + 7Mo + Ti (7)




























Figure 4.21 Backscattered electron image of former TisSi3 region in sample DRSS1400-
HP after complete conversion to TiB2 showing in the center a glassy silica phase.


Figure 4.22 Backscattered electron image of the microstructure of sample DRSS 1200-
HP indicating the dissolution of molybdenum borides at the interface with TisSi3 and the
formation of TiB2.









TiSi (from reaction (3)) + 2B = TiB2 + Si (8)

or TiSi + 2B = TiB2 + TiSi2 (9)

TiSi2 + 2B = TiB2 + 2Si (10)

The MoSi2 MoB agglomerate microstructure observed in Figure 4.22 has been

observed in previous works on this system [13]. In fact, it has been shown that an

orientation relationship exists between these two phases. The formation of new phases at

12000C leads to new types of possible interfaces in the system (Schematic 4.3). However,

some of them clearly do not exist due to their specific locations which limit their contact

with other phases.


The XRD results obtained for sample DRSS1300-HP showed the presence of

both Cl lb (tetragonal) and C40 (hexagonal) MoSi2 along with some of the final

intermediate phases Mo2B and TiSi2 present (Figure 4.21).




DRSS1300-HP o + t-MoSi2
1600 + -MoSi2
S+ + TIB
1400 2IB
o Lo2B
1200 I TISi,
1000
S8oo00 +
+ 0

400 0. + + +1 +
+0 O + + +
O O JI 4/ Ii.


20 40 60 8o [2e] 100 120
Figure 4.23 X-ray diffraction analysis of sample DRSS1300-HP indicating the presence
of C40 MoSi2, Mo2B and TiSi2 in the final stages of the reaction.







Temperature: 1200C
No. of Phases: 9
No. of Possible Interfaces: 36
No. of Reacting Interfaces (*): 5
No. of Constricted Interfaces (C): 8

Inter- TisSi3 B MoB Mo2B MoSi2 TiSi TiSi2 TiB2
faces




B X X X X" X X
-B x- --x -- -x -x

MOB X X X X" XU X
Mo2B X XV- X X x

MoSi2 XU XU X
Tisi X X

TiSi2 X


Schematic 4.3 Possible and constricted reaction interfaces at 12000 C in the displacement
reaction system.


The EPMA results from the DRSS1300-HP are shown in Figure 4.24 and Table
4.13. The presence of a molybdenum boride is apparent in point 2, although TiSi2 was not
observed. Thus, reactions (6) and (7) seem to be the last reactions to take place while
most of the microstructure consists of MoSi2 and TiB2, points 1 and 3, respectively.






























Figure 4.24 Backscattered electron image of the microstructure of sample
DRSS1300-HP indicating the points analyzed by EPMA


Table 4.13 Electron probe microanalysis (WDS) of the points indicated in
Figure 4.24.
Points Mo at. % Si at. Ti at. % B at. % O at. % Possible
% Phases
(1) 33.73 65.35 0.86 0 0 MoSi2
(2) 23.21 48.04 8.29 20.36 0 MoSi2-
TiB2-
Mo-B
mixture?
(3) 1.49 0.19 28.78 69.54 0 TiB2


The phases still present at 13000C form a new set of possible interfaces for the

final reactions in this sample (Schematic 4.4). Similar to the situation encountered at

12000C, some of these interfaces are impossible to occur because of the limitations

imposed during their formation and thus were listed.








Temperature: 1300C
No. of Phases: 4
No. of Possible Interfaces: 6
No. of Constricted Interfaces (C): 2
Inter- MoSi2 TiSi2 TiB2
faces



MoSi2 X-1 X

T1i12 X



Schematic 4.4 Possible and constricted reaction interfaces at 13000 C
in the displacement reaction system.



The EPMA results for the sample DRSS1400-HP indicate the presence of a small

volume fraction of molybdenum boride (Figure 4.25). However, no peaks corresponding

to this phase are observed in the XRD analysis (Figure 4.26) and the extremely fine

pockets of this phase were too small for a reliable analysis using EPMA, Table 4.14,

points 1, 2, and 3. The high atomic percent values obtained for B (> 50%) indicate that

high-B molybdenum borides must be present in what appears to be a mixture of borides

phases in these pockets.

The presence of high-B molybdenum borides is expected when deviations from

the equilibrium MoSi2-TiB2 occur according to the quaternary isotherm presented in

Figure 2.4. A four phase field MoSi2-MoB-Mo2B5-TiB2 adjacent to the equilibrium line

between MoSi2 and TiB2 indicates the possibility of formation of the Mo2B5 phase along

with the formation of MoB.








A general view of the final DRSS1400-HP microstructure is presented in Figure

4.27. There were no major differences found in the case of the DRSS1500-HP. X-ray

diffraction analysis was also found to be identical to the 14000C sample and, therefore,

only the 14000C spectrum is presented (Figure 4.26). The general microstructure was

also comparable to the microstructure of sample DRSS1400-HP (Figure 4.28), and the

presence of very small pockets (less than 1 gm in diameter) of what could be a

molybdenum boride were also found after a very exhaustive search (Figure 4.29 a and b)

as indicated in Table 4.15.


Figure 4.25 Backscatter electron micrograph of sample DRSS 1400-HP indicating the
points analyzed by EPMA









Table 4.14 Electron probe micro-analysis (WDS) of points in Figure 4.25.
Points Mo at. Si at. % Ti at. % B at. % Oat. % Possible Phases

(1) 42.31 3.83 0.12 53.74 0 MoB-
B-rich Mo-B
(Mo2B5 ?)
(2) 43.27 2.7 0.16 53.87 0 MoB
B-rich Mo-B
(MoB5 ?)
(3) 37.55 10.87 0.14 51.43 0 MoB-
B-rich Mo-B
(Mo2B5 ?)
(4) 34.52 65.47 0.01 0 0 MoSi2
(5) 34.54 64.78 0.22 0.46 0 MoSi2
(6) 33.87 63.8 0.04 2.29 0 MoSi2


2500


2000


1500


5 1000


500


20 40 60 80 [28] 100


Figure 4.26 X-ray diffraction analysis of samples DRSS1400-HP and DRSS1500-HP.































Figure 4.27 Scanning electron micrograph (SEM) of sample DRSS1400-HP.


Figure 4.28 Scanning electron micrograph (SEM) of sample DRSS 1500-HP.





























Figure 4.29 Backscatter electron image of sample DRSS1500-HP
indicating the points analyzed by EPMA


Table 4.15 Table 4.14 Electron probe micro-analysis (WDS) of points in Figure 4.29.
Points Mo at. % Si at. % Ti at. % B at. % O at. % Possible
Phases
(1) 27.66 53.53 4.09 14.72 0 MoSi2
(+ Mo-B or
+ TiB2 ?)
(2) 31.67 60.14 1.96 6.23 0 MoSi2
(3) 34.15 63.91 1.02 0.93 0 TiB2
(4) 0.18 0.05 35.87 63.91 0 TiB2


Thus, the general reaction sequence for the displacement reaction can be

described by reactions (1) to (10) according to the indicated reaction balance. Since

reaction (8) can be obtained by adding reactions (9) and (10), the presence of TiSi2 was








considered in the analysis of the evolution of phases but the reactions leading to its

formation eventually were left aside and only reaction (8) was included in the final

equation balance.

The excess Mo and Si liberated in reactions (7) and (8) react to form MoSi2

according to reaction (11).

Mo + 2Si= MoSi2 (1)

It is also expected that the remaining unreacted species complete reaction (12).

Mo2B (from (1)) + 4Si (from (10)) + B (reactant) + Ti (from (7)) = 2MoSi2 + TiB2 (12)

Finally, the displacement reaction can be obtained by balancing the equations

using the proper initial proportions, adding the reactants and products on each side and

canceling the common terms. The last equation represents the most simplified version of

the final reaction.

3MosSi3 + 9B = 4MoSi2 + 7MoB + 2Mo2B + Si (1)

6Ti5Si3 + 24B = 12TiB2 + 18TiSi (3)

18TiSi + 18B = 9TiB2 + 9TiSi2 (5)

Mo2B + 7MoB + TisSi3 + Si = 4TiB2 + 2MoSi2 + 7Mo + Ti (7)

9TiSi2 + 18B = 9TiB2 + 18Si (10)

7Mo + 14Si = 7MoSi2 (11)

Mo2B + 4Si + B + Ti = 2MoSi2 + TiB2 (12)

3Mo5Si3+70B +7 TisSi3+ li + + 15MoSi2+7M~ + 24& + 35TiB2+
7M6S+ 9gieJ2 +;O' t;l'+ +)SSi4 1% l + ?P-91'2 +3 f t~rPJW

3MosSi3 + 7TisSi3 + 70B = 15MoSi2 + 35TiB2

3/7MosSi3 + Ti5Si3 + 10B = 15/7MoSi2 + 5TiB2








It is important to emphasize that the possibility of the formation of some other

intermediate phases cannot be ruled out, since is possible that they could be formed on a

scale not detectable using the analyses performed on these samples. Also, the temperature

intervals used do not allow the inspection of the samples over certain temperature ranges

over which additional phases could have appeared and disappeared and, therefore, not

noticed.

Some important characteristics must be described concerning the final

microstructures of these composites. The presence of pockets of silica (SiO2) found

mostly at the center of previous TisSi3 regions (regions of now TiB2) was a consistent

feature of these microstructures. Often the silica that was found in the center of these

regions had a glassy appearance (Figure 4.30) and was identifiable using x-ray dot maps

in the SEM.



















Figure 4.30 Scanning electron micrograph (SEM) of former TisSi3 region in sample
DRSS1400-HP after complete conversion to TiB2 showing in the center a glassy silica
phase as indicated by silicon and oxygen x-ray maps.








In most cases, however, it seemed that it was pulled out during grinding and

polishing leaving significant "porosity" (Figure 4.31); additional x-ray dot maps (Figure

4.32) and spectra (Figure 4.33) from these regions confirm the presence of silica.

The presence of silica was initially attributed to the final polishing of the samples

using a silica suspension. However, when samples were reground and polished using only

diamond grinding wheels and diamond paste these silica pockets were still observed in

the same regions of the sample indicating that they were an intrinsic feature of the

microstructure.


Figure 4.31 Scanning electron micrograph (SEM) of sample DRSS 1400-HP showing the
pores in the center of TiB2 regions



The presence of silica pockets is expected to be detrimental to the properties of

these composites by the softening of silica at high temperatures due to its relative low

melting point (1710C). A reduction reaction to form the higher melting point phase








silicon carbide, by addition of carbon, is a commonly used way to prevent the presence of

SiO2 in the final microstructures. However, the addition of a new element to these system

may lead to the formation of additional intermediate phases that might affect significantly

the reaction paths proposed for this displacement reaction, and their effect can not be

anticipated.

The reasons for the formation of these silica pockets will be discussed later after

the results concerning the elemental powders reactions are presented.


Figure 4.32 Scanning electron micrograph (SEM) of sample DRSS1400-HP showing a
typical pore in the center of a TiB2 region with some silica particles as indicated by the
silicon and oxygen x-ray maps.





























































0 tSi

0-

0



0-

Ti
0 ". . . ..... ., - i " -- -" .,


4


B
Energy rc.v)


(b)
Figure 4.33 Glassy silica phase partially pulled out from the center of a TiB2 region. (a)
Secondary electron image. (b) Energy dispersive spectrum (EDS) indicating the presence
of mostly Si and O in the region partially pulled out (1), and not pulled out (2).


Counts


Counts


250

200

150

100

50