Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comparative studies of the oxidation of MoSi₂ based materials Samadzadeh, Seyed Mostafa 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_november_samadzadeh_seyedmostafa.pdf [ 5.97MB ]
Metadata
JSON: 24-1.0166781.json
JSON-LD: 24-1.0166781-ld.json
RDF/XML (Pretty): 24-1.0166781-rdf.xml
RDF/JSON: 24-1.0166781-rdf.json
Turtle: 24-1.0166781-turtle.txt
N-Triples: 24-1.0166781-rdf-ntriples.txt
Original Record: 24-1.0166781-source.json
Full Text
24-1.0166781-fulltext.txt
Citation
24-1.0166781.ris

Full Text

COMPARATIVE STUDIES OF THE OXIDATION  OF MoSi2 BASED MATERIALS  by Seyed Mostafa Samadzadeh  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF APPLIED SCIENCE   in The Faculty of Graduate and Postdoctoral Studies  (Materials Engineering)  The University of British Columbia (Vancouver)  October 2015 © Seyed Mostafa Samadzadeh, 2015   ii   Abstract Molybdenum disilicide (MoSi2) is a promising intermetallic material for high temperature applications (above 1000°C). However, it rapidly oxidizes at temperatures ranging from 400 to 600°C, which given enough time, can lead to its disintegration. Above 1000°C, MoSi2 exhibits better oxidation resistance due to the formation of a continuous SiO2 layer (or alumina layer for the materials doped with aluminum). The experiments in this study were divided into two main categories: low temperature oxidation (300 to 900°C; high oxidation rate expected) and high temperature oxidation (1000 to 1600°C; lower oxidation rate expected due to rapid formation of the protective oxide films). The isothermal exposure time in the low temperature oxidation experiments was from 4 to 240 hrs while it was from 2 to 144 hrs for the high temperature oxidation experiments. Five different, commercially available, MoSi2 based heating elements, i.e. Kanthal Super (labelled by the manufacturer as KS-1700, KS-1800, KS-1900, KS-ER and KS-HT) were used in the experiments. It was found that the oxidation behavior of different materials under investigation depended strongly on their chemical and phase composition, exposure time and temperature.  KS-ER and KS-1800 showed excellent resistance against the low temperature (300 to 900°C) degradation for up to 240 hrs, while KS-HT and KS-1900 underwent significant degradation after 240 hrs of air exposure within the same temperature range. In high temperature oxidation experiments, a dense barrier alumina film (1.5 µm thick at 1000 C to 50 µm thick at 1500C for up to 144 hrs) formed on KS-ER samples.  A dense glassy SiO2 film (3 µm thick at 1000C to 50 µm thick at 1600C for up to 144 hrs) formed on the other types of samples. The glass scale on the surface of KS-1700, KS-1800 and KS-HT was significantly thicker (~3 times) than that on KS-HT over the temperature range of 1200°C to 1600°C after 144 hrs. The rate of alumina formation of KS-ER was relatively higher than the glass film formation of the other types of composite MoSi2 materials. The differences in the oxidation behavior of various MoSi2-based materials were linked to their chemistry and phase compositions.  iii   Preface This research work was conducted as part of a NSERC Collaborative Research Grant awarded to the University of British Columbia, with support from Westport Innovations Inc. of Vancouver. In this study, I was responsible for the materials and samples preparation, running the experiments and samples preparation for characterization, results analysis and presentation, and formulation of the preliminary conclusions. All SEM and EDX analyses presented in this study were performed by Carmen Oprea. I performed the XRD analysis presented in Chapters 4 and 5. This research work was supervised by Prof. Tom Troczynski from the University of British Columbia, and co-supervised by Dr Hamed Karimi Sharif from Westport Innovations Inc. and Adjunct Professor at the University of British Columbia. A version of Chapter 5 has been presented in the 11th International Conference on Ceramic Materials and Components for Energy and Environmental Applications entitled: “Comparative studies of the oxidation of MoSi2 based materials” in two parts. One was “Part 1: Low temperature oxidation (300–900°C)”, and the other was “Part 2: High temperature oxidation (1000–1600°C)”. They both authored by M. Samadzadeh, C. Oprea, H. Karimi Sharif, and T. Troczynski.   iv   Table of Contents  Table of Contents Abstract ................................................................................................................................................ ii Preface ................................................................................................................................................. iii Table of Contents ............................................................................................................................... iv List of Tables ..................................................................................................................................... vii List of Figures ................................................................................................................................... viii List of Symbols .................................................................................................................................. xii Nomenclature ................................................................................................................................... xiii Acknowledgements ........................................................................................................................... xiv 1 Introduction ...................................................................................................................................... 1 1.1 An introduction to the research project ....................................................................................... 1 1.2 Overview of Chapters.................................................................................................................. 3 2 Background information ................................................................................................................. 4 2.1 Introduction ................................................................................................................................. 4 2.2 The development history of MoSi2 based materials .................................................................... 5 2.3 Materials processing for MoSi2 synthesis and forming ............................................................... 6 2.4 Alloying additions to molybdenum disilicide ............................................................................. 8 2.4.1 MoSi2-Si3N4 ......................................................................................................................... 9 2.4.2 MoSi2-SiC ............................................................................................................................ 9 2.4.3 MoSi2-Al2O3 ........................................................................................................................ 10 2.4.4 MoSi2-WSi2 ........................................................................................................................ 10 2.5 Oxidation behavior of molybdenum disilicide .......................................................................... 10 2.6 Summary ................................................................................................................................... 17 3 Scope and objectives ...................................................................................................................... 19 3.1 Scope ......................................................................................................................................... 19 3.2 Objectives .................................................................................................................................. 21 4 Materials and experimental procedures ...................................................................................... 24 v   4.1 Materials and sample preparation ............................................................................................. 24 4.1.1 Kanthal Super 1700 and 1800 ............................................................................................ 25 4.1.2 Kanthal Super 1900 and HT ............................................................................................... 25 4.1.3 Kanthal Super ER ............................................................................................................... 26 4.1.4 Sample preparation ............................................................................................................ 28 4.2 Design of experiments ............................................................................................................... 29 4.2.1 Heat treatment schedule in LTO (300-900°C) ................................................................... 29 4.2.2 Heat treatment schedule in HTO (1000-1600°C) ............................................................... 31 4.2.3 Experimental conditions in low and high temperature oxidation (300-1600°C) ............... 32 4.2.4 Thermal Analysis (TGA / DTA) of MoSi2-based materials............................................... 33 4.3 Uncertainties and errors ............................................................................................................ 34 4.4 Characterization techniques ...................................................................................................... 35 4.5 Test samples nomenclature ....................................................................................................... 36 5 Initial results and experiments ...................................................................................................... 37 5.1 The effects of the pre-oxidized layer of MoSi2-based materials ............................................... 37 5.2 The effect of high temperature pre-treatment of MoSi2-based materials .................................. 40 5.3 Thermogravimetric analysis ...................................................................................................... 42 5.4 Summary and conclusions ......................................................................................................... 44 6 Oxidation behavior of clay-bonded MoSi2 based materials ....................................................... 46 6.1 Low temperature oxidation (LTO) of clay-bonded MoSi2 based materials (300-900°C) ......... 46 6.2 High temperature oxidation (HTO) of MoSi2 based materials (1000-1600°C) ......................... 53 6.3 Overview of oxidation of MoSi2 based materials (300-1600°C) .............................................. 59 7 Oxidation behavior of (Mo, W)Si2 based materials ..................................................................... 61 7.1 Low temperature oxidation (LTO) of (Mo, W)Si2 based materials (300-900°C) ..................... 61 7.2 High temperature oxidation (HTO) of (Mo, W)Si2 based materials (1000-1600°C) ................ 69 7.3 Overview of oxidation of (Mo, W)Si2 based materials (300-1600°C) ...................................... 74 8 Oxidation behavior of Mo(Al, Si)2 based materials ..................................................................... 77 8.1 Low temperature oxidation (LTO) of Mo(Al, Si)2 based materials (300-900°C) ..................... 77 8.2 High temperature oxidation (HTO) of Mo(Al, Si)2 based materials (1000-1600°C) ................ 80 8.3 Overview of oxidation of Mo(Al, Si)2 based materials (300-1600°C) ...................................... 85 9 Comparison of the oxidation of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials ............. 86 vi   9.1 A proposed failure criterion for MoSi2 based materials ............................................................ 86 9.2 Microstructural analysis of MoSi2 based materials after HTO (1000-1600°C) ........................ 87 9.3 Determining the activation energy of diffusion in MoSi2 based materials ................................ 89 9.4 The oxide scale thickness of MoSi2 based materials (1000-1600°C) ........................................ 92 9.5 Comparative oxidation kinetics of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials (300-1600°C) ................................................................................................................................................... 95 10 Summary and conclusions ........................................................................................................... 96 References ........................................................................................................................................ 100     vii   List of Tables Table 4.1 The maximum service temperature of the Super Kanthal heating elements, as recommended by Kanthal [10] ........................................................................................................................................... 25 Table 4.2 Density and elemental compositions of MoSi2-based materials, as measured in the ceramic lab at UBC .................................................................................................................................................. 25 Table 4.3 Test samples nomenclature used in this study .................................................................... 36 Table 6.1 EDX of the surface of KS-1700 (all data in wt%) .............................................................. 49 Table 6.2 Elemental composition of KS-1700 through the cross section after 144 hrs exposed to 1600°C (all data in wt%) ............................................................................................................................. 56 Table 9.1 Composition variation through the cross-sections of samples in temperature range 1000-1600°C ........................................................................................................................................................ 88 Table 9.2 Thermal activation energies calculated by performing Arrhenius Plots ............................. 92 Table 9.3 The oxide scale thickness (all in µm) in high temperature isothermal oxidation ................ 94      viii   List of Figures Figure 2.1 Calculated Gibbs free energy changes ∆G of reactions (1a) and (2a) as a function of temperature [58] .......................................................................................................................................... 12 Figure 2.2 Calculated isothermal sections of Mo-Si-O ternary system at (a) 500°C and (b) at 1600°C. [58] .............................................................................................................................................................. 13 Figure 2.3 Schematic isothermal oxidation curves for MoSi2 at (a) low and (b) high temperatures [58] .................................................................................................................................................................... 17 Figure 4.1 SEM backscattered electron image of the bulk material in an unexposed Kanthal commercial heating element samples viewed in cross-section ................................................................... 27 Figure 4.2 Different types of Kanthal Super bars MoSi2 after the mechanical removal of the protective layer ............................................................................................................................................................ 28 Figure 4.3 Prepared specimens for the experiments after removing the protective layer ................... 29 Figure 4.4 Heat treatment schedule for the experiments at 900°C ...................................................... 30 Figure 4.5 H Heat treatment schedule for the experiments at 1600°C ................................................ 31 Figure 4.6 (a) Sample arrangement during LTO experiments before and after exposure ................... 33 Figure 4.7 PerkinElmer STA-6000 – Simultaneous TGA-DTA used in this investigation ................ 34 Figure 5.1 The appearance and microstructure of (a) (Mo, W)Si2, (b) MoSi2, and (c) Mo(Si, Al)2 based materials after 186 hrs. at 500°C ................................................................................................................. 38 Figure 5.2 The microstructure of the cross section near the surfaces of clay bonded MoSi2 based materials exposed to air for 186 hrs at 500°C ............................................................................................. 39 Figure 5.3 Mass changes versus time at a) 500°C, b) 750°C, c) 1000°C , for different samples pre-oxidized at 1500°C for 66 hrs.  Each point is an average of two samples tested. ....................................... 41 Figure 5.4 TGA and DTA curve of KS-ER powder at temperature (400-995°C) .............................. 43 Figure 5.5 The reactions formed, volatilized, and deposited from KS-ER powder in surrounding of the TA Instrument resulted after experiment .................................................................................................... 44 ix   Figure 5.6 the heat treatment procedure of the experimental work ..................................................... 45 Figure 6.1 Mass variations versus time of KS-1700 (left) and KS-1800 (right) (300-900°C) ............ 47 Figure 6.2 The formation of needle shaped MoO3 on the surface of KS-1700/1800 at 300°C and 400°C after 240 hrs ................................................................................................................................................ 48 Figure 6.3 The surface and section of (a,c) KS-1700 and (b,d) KS-1800 at 500°C after 240 hrs ....... 49 Figure 6.4 XRD analysis of the sample surface of KS-1700 after 240h at 500°C .............................. 50 Figure 6.5 The surface microstructure of KS-1700and 1800 at 500 and 600°C after 240h ................ 51 Figure 6.6 Surface discoloration of the KS-1700 and KS-1800 specimens after 240h (400-900°C) . 52 Figure 6.7 Micrograph of the KS-1700 surface and mass changes after 240 hrs at 800°C ................. 53 Figure 6.8 Mass variations versus time of KS-1700 KS-1800 in the temperature range 1000-1600°C .................................................................................................................................................................... 54 Figure 6.9 Mass variations versus time of KS-1800 in the temperature range 1000-1600°C ............. 55 Figure 6.10 SEM section of the scale development formed on KS-1700 at temperature ranges 1000-1600°C ........................................................................................................................................................ 57 Figure 6.11 SEM section of the scale development formed on KS-1800 at temperature ranges 1000-1600°C ........................................................................................................................................................ 58 Figure 6.12 Mass variations after 72 h isothermal exposure versus temperature of (a) KS-1700 and (b) KS-1800 ...................................................................................................................................................... 59 Figure 7.1 Mass varations versus time of KS-1900 (left) and KS-HT (right) (300-900°C) ............... 63 Figure 7.2 The surface section and microstructure of KS-1900 and KS-HT at 400 and 500°C ......... 64 Figure 7.4a,b SEM backscattered surface of KS-1900 and KS-HT exposed to 700°C for 240 h ....... 65 Figure 7.3a,b SEM backscattered electron image and X-rays diffraction pattern of the surface of KS-HT exposed to 500°C after 240 hrs ............................................................................................................. 66 Figure 7.5 The surface microstructure of KS-1900 exposed to 800-900°C for 240 h ........................ 68 Figure 7.6 The surface discoloration of the specimens after 240h at different temperatures ............. 68 x   Figure 7.7 Mass variations versus time of KS-1900 in the temperature range 1000-1600°C ............. 70 Figure 7.8 Mass variations versus time of KS-HT in the temperature range 1000-1600°C ............... 71 Figure 7.9 SEM section of the scale development formed on KS-1900 at temperature ranges 1000-1600°C ........................................................................................................................................................ 72 Figure 7.10 SEM section of the scale development formed on KS-HT at temperature ranges 1000-1600°C ........................................................................................................................................................ 73 Figure 7.11 The surface appearance of KS 1900/HT bars after 144h at 1300 & 1500°C ................... 74 Figure 7.12 The microstructure section of KS-HT a) 1200°C after 72h b) 1500°C after 144h .......... 74 Figure 7.13 Mass variations after 72 h isothermal exposure versus temperature of (a) KS-1900 and (b) KS-HT ......................................................................................................................................................... 75 Figure 8.1 Mass variations versus time of KS-ER (300-900°C) ......................................................... 78 Figure 8.2 SEM surface of KS-ER exposed to 300-400°C for 240 h (a) The major grey region is Mo(Al,Si)2 (I), the blac regions are Al2O3 (II) and the white regions are Mo5(Al,Si)3 (III). (b) Elongated MoO3 crystals formed on the surface (I) ..................................................................................................... 78 Figure 8.3 (a, b) The SEM surface and section of KS-ER exposed to 500°C and (c, d) the SEM surface of KS-ER exposed to 600, 700°C for 240 h ................................................................................................ 79 Figure 8.4 The surface discoloration of the specimens after 240h at different temperatures ............. 80 Figure 8.5 Mass variations versus time of KS-ER in the temperature range 1000-1600°C ................ 82 Figure 8.6a,b,c,d,e the section and the scale appearance formed on KS-ER at temperature ranges 1000-1500°C ........................................................................................................................................................ 83 Figure 8.7 The microstructure section KS-ER exposed to 1600°C ..................................................... 84 Figure 8.8 Mass variations after 72 h isothermal exposure versus temperature of Mo(Si, Al)2 (KS-ER) .................................................................................................................................................................... 85 xi   Figure 9.1 (a) Low temperature oxidation study-summary, (b) Summation of the absolute values of mass gain and mass loss per unit area, at different temperatures (300˚C-900˚C), for the materials studied in this work ..................................................................................................................................................... 87 Figure 9.2 Plots of specific weight gain versus time of KS-1800 in the temperature range 1000-1600°C .................................................................................................................................................................... 90 Figure 9.3 Plots of specific weight gain square versus time by 2 at five test temperatures during oxidation of KS-1800 .................................................................................................................................. 91 Figure 9.4 Plots of logarithm of parabolic rate constant versus inverse absolute temperature determining activation energy for oxidation of MoSi2-based materials studied here ................................. 91 Figure 9.5 Mass variations after 72h isothermal exposure versus temperature of the materials studied in the temperature range 300-1600°C ......................................................................................................... 95      xii   List of Symbols ΔG Gibbs free energy [kJ.mole-1]  T Temperature [°C] ΔE Activation energy [kJ.mole-1] Kp Parabolic rate constant D Diffusion coefficient [m2s-1] t Time [hour] Q Diffusion activation energy [kJ.mole-1] R Gas constant [kJ.mole-1.K-1)] DL Bulk diffusion coefficient [m2.s-1] DGB Grain boundary diffusion coefficient [m2.s-1]            xiii   Nomenclature LTO Low Temperature Oxidation HTO High Temperature Oxidation LHT Low and High Temperature NG Natural Gas HSIS Hot Surface Ignition Systems HPNG High Pressure Natural Gas HTS Heat Treatment Schedule TGA Thermo-Gravimetric Analysis DTA Differential Thermal Analysis CTE Coefficient of Thermal Expansion SEM Scanning Electron Microscopy EDX Energy-dispersive X-ray spectroscopy XRD X-Ray Diffraction Analysis BSE Back-Scattered Electron KS Kanthal Super        xiv   Acknowledgements I offer my enduring gratitude to the faculty, staff and my fellow students at the University of British Columbia, who have inspired me to continue my work in this field. I owe particular thanks to Prof. Tom Troczynski, for his patient guidance, immeasurable support, and encouragements throughout the course of this work. I also acknowledge Dr. Hamed Karimi Sharif who shared with me his experience, personal insights and provided me with constant support and guidance throughout this project. My sincere thanks go to Carmen Oprea for her assistance in bringing this project to completion. She has been continuously supporting and helping me throughout the years I have been at UBC. Natural Sciences and Engineering Research Council of Canada (NSERC) and Westport Innovations Inc. are greatly acknowledged for financial support of this project.  I want to extend my thanks to my family for their love and encouragement. 1  1   Introduction Chapter 1 Introduction 1.1 An introduction to the research project One of the main challenges in developing high-pressure, direct-injection (HPDI) natural gas (NG) engine technology is the lack of a reliable and durable hot surface ignition systems. The design of natural gas engines is largely based upon the designs of diesel engines, with one important difference providing rationale for the present work. Diesel engines use compression ignition with the help of glow plugs during cold starts.  However, NG-supplied HPDI engines require presence of ignition hot-surface (a "glow plug"), constantly maintained at high temperature of about 1300°C (the exact level of this temperature is still debated, but it also depends on the engine operating characteristics).  In diesel engines, air is compressed first, thus its temperature increases, reaching temperature sufficient for spontaneous ignition of the injected diesel fuel.    It is always necessary to ensure that the fuel and air mixture reach the needed temperature for ignition. Accordingly, glow plugs operating at ~900°C are used to provide the initial source of ignition and warm up the combustion chamber, during cold start operations. The studies have shown that temperatures > 927°C are required for natural gas (NG) ignition within 2 ms from injection, which is higher than for the diesel fuel. It has been shown that the ignition delays even shorter than 2 ms are needed  for optimal engine performance and low emission capabilities [1]. The existing state-of-the-art ceramic glow plugs developed to run intermittently in cold diesel engines, fail prematurely running constantly at ~1300°C in high-performance NG engines, i.e. usually after less than 1,000 hrs of operation, whereas at least 10,000 hrs is looked-for by the industry. As a result, the development of new ignition technology is critical for NG engines. 2 The collaborative research program at the University of British Columbia aims to investigate and address the limitations of the design and materials for Hot Surface Ignition Systems (HSIS) in High-Pressure Natural Gas (HPNG) Direct Injection engines under development by Westport Innovations Inc., of Vancouver, BC. The long-term engineering goal of this program is to provide research support and design assistance to Westport towards development of an acceptable-cost HSIS which will reliably perform for at least 10000 hrs in HPNG engines.  There are only a handful of materials suitable for this application, among which MoSi2 materials seem to be the most promising in terms of their high electrical and thermal conductivity, and high melting point (2030ºC). In addition, molybdenum disilicide based materials are currently used for other high temperature applications such as gas turbines and heating elements. Even though molybdenum disilicide and its composites are recognized as a promising candidates to meet the HSIS criteria for HPNG applications, there are some disadvantages, which need to be addressed before field-application of these materials. The high brittleness and excessive oxidation  at low temperatures (below 1000°C, also referred to as "pesting") and poor creep resistance at high temperatures (above 1200°C) are important disadvantages of MoSi2. One of the principal objectives of this broad research program is to design, fabricate and evaluate MoSi2-based experimental Glow Ring (GR) systems and develop quantitative models of their deterioration process for HPNG Direct Injection engines. The oxidation behavior studies of MoSi2 based materials in air is the main objective of this study and also a part of the objectives of the broad university-industry collaborative research program. The overall aim is to bring a better understanding of MoSi2-based materials deterioration in oxidizing environments. 3 1.2 Overview of Chapters This dissertation is divided into six chapters (with introduction being the first). Chapter 2 provides background information on the project, including historical context, and a review of related literature in this field. It provides context for understanding the purpose and scope of the project. Chapter 3 provides the scope and objectives of this study. Chapter 4 addresses the design of experiments like heat treatment schedule, materials used in this investigation, sample preparations for the experiments, experimental procedures, uncertainties and errors occurred during experiments, and the characterization techniques used in this study. Chapters 5 to 9 present the results and discussions of this study on the comparative oxidation behavior of the MoSi2 based materials. Finally, Chapter 10 gives a summary of the work done, answers the six major research questions stated in the objectives, and gives suggestions for the future work.   4  2   Background information Chapter 2 Background information 2.1 Introduction Molybdenum disilicide (MoSi2) is known since 1907 [2]. MoSi2 is an intermetallic compound with tetragonal C11b structure formed between molybdenum and silicon. It is a grey metallic-looking material with a very high melting point (2030°C), low brittle-ductile transition temperature of approximately 800–1100ºC, and a moderate density (6.24 g.cm-3), and the room temperature fracture toughness of around 4 MPa. 𝑚12. It also exhibits a combination of a low electrical resistivity and a good thermal conductivity at high temperatures [3–8], e.g. 0.4 Ώmm2m-1 and 30 W.m-1K-1 at room temperature, respectively. MoSi2 is a promising intermetallic material for high temperature applications (above 1000°C). MoSi2 materials are widely utilized as heating elements for high-performance furnaces, capable of providing temperatures in excess of 1800°C due to its high electrical and thermal conductivity over a wide range of temperatures. Projected applications of MoSi2-based materials include hot section components in high temperature structural applications such as aircraft engines, gas turbines, glow plugs of diesel engines and industrial gas burners [3, 4, 7, 8]. However, MoSi2 oxidizes relatively fast at temperatures ranging from 400 to 600°C, which given enough time, can lead to its disintegration. Above 1000°C, MoSi2 exhibits better oxidation resistance due to the formation of a continuous SiO2 layer on its surface. Also, it has a brittle-to-ductile transition around 1000°C. Above this temperature, its strength is governed by plastic flow (creep) [3, 7]. Its brittleness and "pesting" oxidation behavior at lower temperatures (defined as oxidation below 1000°C) and a poor creep resistance at higher temperatures (above 1200°C) have hindered its use as in load-bearing parts. In addition, its relatively high coefficient of thermal expansion (8×10−6 K−1 limited use of MoSi2 during thermal cycling. 5 There are indications in the literature that addition of a secondary material into a composite or an alloy may offset the disadvantages of MoSi2 [3, 7]. MoSi2 has high melting point (2030°C ),  high strength to density ratio, low brittle to ductile transition temperature, and high oxidation resistance at  high temperatures, which make it desirable for high temperature structural applications [9, 10]. In spite of the fact that a majority of the silicides of Mo, Nb, W, Ti and Cr have high melting points (2000°C or above), it has been shown that only MoSi2 and Ti5Si3  are the most promising among all the silicides [9]. There are three Mo-Si stoichiometric compounds found: Mo3Si (cubic), Mo5Si3 (tetragonal), and MoSi2 (tetragonal) [7, 9, 11].  The idea to develop MoSi2 and Mo(Si, Al)2 alloy  for heating elements is credited to Kieffer et al. [12]. The first commercial application of MoSi2-based heating elements was developed by Kanthal in 1953 [10]. MoSi2 and its composites have been widely studied since then, and other potential applications have been proposed [9]. MoSi2-based heating elements commonly known as Kanthal Super are in use for several decades for temperatures up to 1700°C. It contains fine grains (5 to 20 µm) of MoSi2, bonded together with aluminosilicate glass phase [9, 13]. More recently, Kanthal Super 1900 and Kanthal Super HT heating elements containing a solid solution alloy of MoSi2 and WSi2 have come into use. Kanthal Super ER heating elements containing MoSi2 doped with aluminum was also introduced in 2010 [10]. Using composites and alloys helped to increase the lifetime service of the elements with a better creep resistance, oxidation resistance and fracture toughness [10]. 2.2 The development history of MoSi2 based materials The first patent of composite materials based on MoSi2 for heat conductor was credited for Kieffer et al. [12], and the first long-time resistance tests in air at temperatures of 1700°C was conducted by Fitzer [14[. Kanthal [10] patented the first commercial heating elements in 1956. Fitzer had observed MoSi2 and Mo5Si3 compounds in his coatings work [14]. Erdoes [15] suggested that the silicides could be considered as coating materials for gas turbine engines. 6 Fitzer et al. [14], investigated the benefits of Al2O3, SiC, and Nb additions to MoSi2 and demonstrated an improvement in mechanical properties. They then suggested in 1978 that MoSi2 is an important matrix material for high temperature structural applications. In 1985, Gac and Petrovic studied addition of SiC whisker to MoSi2 matrix composites and reported improvements in room temperature strength and fracture toughness [8]. In 1986, Bose at Pratt & Whitney began examining SiC-MoSi2 composites for potential aircraft engine applications. In 1988, he studied composite materials based on MoSi2 for aerospace applications [8, 16, 17].  2.3 Materials processing for MoSi2 synthesis and forming The most common techniques to synthesize MoSi2 and the associated composites are powder processing techniques, self-propagating high-temperature synthesis SHS, plasma spray processing, and solid-state displacement reactions. The methods of processing play a substantial role in determination of the impurity content and evolution of microstructure, which leads to the physical, mechanical properties and oxidation resistance [9, 18, 19]. Powder processing involves the consolidation of discrete powders into a bulk form with the help of temperature and/or applying pressure. There are four available pressure-assisted processing techniques in powder processing, namely hot pressing, sinter forging, hot isostatic pressing (HIP), and hot extrusion, which hot pressing is most widely used among them [18–21]. There have been widely and successfully utilized powder processing to process monolithic MoSi2 [22] and a wide variety of MoSi2 based composites, such as those containing Nb, W, C, ZrO2, Al2O3, SiC and TiC reinforced MoSi2, with significant improvements in both flexure strength and fracture toughness [18, 23]. Preparing the materials by hot pressing a mixture of reinforcement and MoSi2 powders is used only for simple shapes, due to poor machinability of MoSi2 materials. Another powder processing technique is mechanical alloying, which 7 involves repeated welding, fracturing and rewelding of powders during high-energy milling under a controlled atmosphere [18, 23]. Formation of MoSi2, using mechanical alloying method, was for the first time reported by Iwatomo & Uesaka in 1990. In this study, Mo and Si powders were used. Following mechanical alloying, the crystalline MoSi2 particles were highly refined, presenting a grain size ranging from 5 to 10 nm. Higher density, lower hot-pressing temperatures for consolidation and better chemical homogeneity are the advantages of using mechanical alloying method [18, 24]. Jayashankar and Kaufman produced silica free MoSi2 matrix composites by mechanical alloying of Mo, Si and C powders, followed by hot pressing [25, 26]. There have been limited studies conducted by self-propagating high-temperature synthesis technique SHS. This technique involves propagation of a high-temperature zone, driven by an exothermic reaction, through a compact of reactants [18]. The content of oxygen in molybdenum disilicide is a critical parameter, because of the possibility of formation of SiO2. It has been shown that in situ reaction sintering or hot pressing of elemental powders is self-purifying leading to a lower oxygen content in the final products than that of the elemental powders [9]. For the application of oxidation resistant coatings, chemical vapor deposition (CVD) technique has been used to deposit MoSi2 on various substrates. Edward K. etc., [27] deposited a uniform layer of MoSi2–SiO2 with excellent adhesion onto molybdenum substrate via chemical vapor deposition (CVD) of SiCl4/H2 and the subsequent metal-organic chemical vapor deposition (MOCVD) of TEOS/N2 at relatively low temperatures. Plasma spray processes present a unique opportunity to synthesis in situ of fine particles and in some cases, near-net shape manufacturing. Castro et al. [28] fabricated monolithic and reinforced MoSi2 materials using the low vacuum plasma deposition technique. This is a kind of plasma spraying techniques, which has been used to produce highly dense bulk composite materials. It was reported that the density of both reinforced and monolithic plasma-sprayed MoSi2 was in the 95-98% range with a highly refined 8 microstructure [28]. Another plasma synthesis approach was recently conducted to manufacture SiC-reinforced MoSi2 composite by Jeng et al. [29, 30]. In this study, they obtained a dense composite with improved toughness and well-bonded interface. Formation of in situ MoSi2-SiC composites by using a solid-state displacement reaction between Mo2C and Si created by Henager et al. [31, 32]:  𝑀𝑜2𝐶 + 5𝑆𝑖 → 2 𝑀𝑜𝑆𝑖2 +  𝑆𝑖𝐶 (1)  This technique involves the reaction of two or three elements to form thermodynamically stable new compounds. It has been shown that the solid-state reactions may be successfully used to synthesize composite materials with impressive combinations of microstructure and properties [31, 32]. Solid-state displacement reactions are most beneficial when utilized in conjunction with another processing methodology, such as plasma processing or hot pressing. However, further improvements in the microstructure and properties of MoSi2 are possible through the judicious combination of several synthesis techniques, such as introduction of solid-state reactions during plasma processing [18]. 2.4 Alloying additions to molybdenum disilicide It has been shown that the pest oxidation phenomenon may be overcome by alloying and processing techniques [33]. The low temperature brittleness, and poor high temperature creep performance of MoSi2 may be offset at least partly by using it together with a second material in a composite or an alloy [3, 6]. MoSi2 can be alloyed with other high melting point silicides such as Mo5Si3, WSi2, NbSi2, CoSi2, and Ti5Si3. It is also thermodynamically stable with a wide variety of potential ceramic reinforcements for composites, including SiC, Si3N4, Al2O3, ZrO2, TiB2, and TiC [34]. Waghmare et al. [35] suggested that alloying additions of Mg, V, Nb, Tc, and Al to the MoSi2 crystal structure increase the ductility of MoSi2, while additions of Ge, P, and Re were expected to have the reverse effect [4]. It is reported that additions 9 of germanium to MoSi2 indicates the low temperature oxidation resistance by formation of a lower viscosity silica glassy phase [6]. Mo5Si3 (of higher melting point of  2180°C) is significantly more creep resistant than MoSi2 [4]. It has been shown that the creep rate of Mo5Si3 is only one-fifth of the creep rate of MoSi2 at 1200°C at the same stress [9]. However, its oxidation resistance is significantly inferior to MoSi2. Activation energies for oxygen diffusion during oxidation of molybdenum silicides are 130 kJ.mol-1 for MoSi2 and 210 kJ.mol-1 for Mo5Si3, It is reported that boron additions to Mo5Si3 significantly improve the oxidation resistance of Mo5Si3. This is due to the formation of a protective borosilicate glass as a result of the boron additions [36].  2.4.1 MoSi2-Si3N4 Composites of MoSi2 and Si3N4 offer enhanced  properties as compared to other MoSi2 based composite systems [37].  It has been shown that Si3N4 additions to MoSi2 improve the low temperature oxidation resistance of MoSi2 [38]. Addition of 30–50 vol.% Si3N4 increases the fracture toughness of MoSi2 matrix-Si3N4 reinforced composites, reaching values as high as 15 MPa 𝑚12⁄  at 1300°C.  [4]. A diesel engine glow plug made of MoSi2–Si3N4 (30:70 vol.%) for automotive applications has been developed [4]. The glow plug lifetime is reported to be as high as 13 years in the diesel fuel combustion environment, at maximum use temperature of ~1000-1200°C, which is very advantageous [4]. Additionally, the higher heating rate of MoSi2–Si3N4 than that of metal glow plugs allows for faster starting of the diesel engine by approximately a factor of two [4]. 2.4.2 MoSi2-SiC Recent works shown that SiC additions to MoSi2 improve the mechanical properties of MoSi2 matrix-SiC reinforced composites. By addition of SiC to MoSi2, the final product (SiC-MoSi2 composite) does not contain residual Si, presence of which is unfavorable in high temperature applications [3, 4, 23].  10 2.4.3 MoSi2-Al2O3 There is a very close thermal expansion coefficient match between MoSi2 (8×10−6 K−1) and Al2O3 (8.1×10−6 K−1). The research carried out by Stefan Köbel [39] showed successful fabrication of MoSi2–Al2O3 laminate composite tubes by plasma spray-forming techniques. SiO2, one of the oxidation products of MoSi2, reacts with Al2O3, to form aluminum silicates including Al2O3·2SiO2, Al2O3·SiO2, and 3Al2O3·2SiO2 [40]. 2.4.4 MoSi2-WSi2 WSi2 is similar to MoSi2 the regarding crystal structure, lattice parameters, melting point, coefficient of thermal expansion, elastic constants and electronic structure [3, 9, 41]. The solid-solution strengthening of WSi2 additions to MoSi2 has proved to be more effective than the dispersion strengthening of the Mo5Si3 particles in the MoSi2-Mo5Si3 alloys [24].Addition of WSi2 improves mechanical properties of MoSi2 e.g., creep resistance and strength [3, 9, 41]. In this regard, Kanthal [10] introduced heating elements containing a solid solution alloy of MoSi2 and WSi2 namely Kanthal Super 1900 and Kanthal Super HT. The high oxidation resistance of MoSi2 is attributed to the formation of a self-healing, glassy silica (SiO2) layer, which prevents the MoSi2 matrix from further oxidation.  However, the low melting-point of glassy phase is detrimental to the high temperature strength.  The formation of weak second phase or liquid phase along grain boundaries enhances grain boundary sliding and mass transport during high temperature exposure, and thus decreases creep resistance. To improve creep resistance of MoSi2, it is imperative to produce composites that are free of SiO2 [18]. Using a second material, e.g. SiC or WSi2, in a composite or an alloy may improve the creep resistance of MoSi2 [3]. 2.5 Oxidation behavior of molybdenum disilicide MoSi2 suffers from accelerated oxidation in the temperature range of 400-600°C that leads to disintegration of the material. This phenomenon was discovered and named “pest oxidation” by Fitzer [14] in 1955. Since then, there have been many attempts to understand the low temperature oxidation behavior 11 of MoSi2 and to determine under what circumstances this phenomenon occurs [42–55]. Westbrook and Wood [56] proposed that the preferential inter-granular diffusion of oxygen coupled with grain boundary embrittlement leads to disintegration or pest oxidation of MoSi2. Berkowitz-Mattuck et al. [7, 57] reported that disintegration or pest oxidation of MoSi2 attributed to enhanced residual stresses introduced via cooling anisotropic materials. Several studies reported that the pest oxidation of MoSi2 is a result of the inter-granular attack, whereby each individual grain was enveloped by the reaction products [43, 50, 58]. It is reported that the pest oxidation takes place because of the improper microstructure and excessive porosity and it is not an intrinsic material property of MoSi2  [14, 59]. Berztiss et al. [27] studied the oxidation behavior of MoSi2 and concluded that the low temperature (pest) oxidation of MoSi2 was a function of the non-optimized fabrication conditions resulting in poor microstructure. Chou & Nieh [49] studied pest oxidation and found reaction oxidation products consisting of SiO2 clusters and MoO3 whiskers. McKamey et al. [46] studied the roles of composition, grain or phase boundaries, and physical defects on the oxidation and fracture of MoSi2. They also found that the pest oxidation occurred through transporting oxygen into the interior of the specimen along pre-existing cracks where it reacted to form simultaneously MoO3 and SiO2. In this regard, Meschter [55] proposed that the pest oxidation did not occur in highly dense MoSi2 (>95% of theoretical density) for exposure times up to 688 hrs at 400-600°C.  There is general agreement that the low temperature oxidation of MoSi2 (also known as pest oxidation in the temperature range 400-600°C) is closely related to the formation of the volatile voluminous MoO3.  According to Liu et al. [58] there are two possible oxidation reactions for MoSi2: 1(a): 2 𝑀𝑜𝑆𝑖2 + 7𝑂2  → 2 𝑀𝑜𝑂3 + 4 𝑆𝑖𝑂2 (2) 2(a): 5 𝑀𝑜𝑆𝑖2 + 7𝑂2  → 2 𝑀𝑜5𝑆𝑖3 + 7 𝑆𝑖𝑂2 (3)  Gibbs free energies of the reactions (1a) and (2a) at different temperatures were calculated by Liu et al. [58] on the basis of the available thermodynamic data, see Figure 2.1. In this study, the thermodynamic 12 data of Mo-Si, Si-O, Mo-O constituent-binary systems and pure elements were taken from Liu et al. [60, 61], Hallstedt [62], Bygden et al. [63] and Dinsdale [64] respectively. They showed that the reaction (2a) is favored over reaction (1a) since the calculated absolute value of ∆𝐺2𝑎 is greater than that of ∆𝐺1𝑎. Figure 2.1 shows the calculated Gibbs free energy changes ∆G of reactions (1a) and (2a) at different temperatures based on 1 mol oxygen [58].  Figure 2.1 Calculated Gibbs free energy changes ∆G of reactions (1a) and (2a) as a function of temperature [58]  The isothermal sections of the Mo-Si-O system at different temperatures were calculated by Liu et al. [58] on the basis of the available thermodynamic data. It shows that SiO2 is more stable than MoO2 and MoO3, and that the terminal molybdenum solid solution (Mo), which has a very low solubility of silicon at 500°C, can be in equilibrium with SiO2. This is concluded through isothermal sections of the Mo-Si-O system at 500°C and 1600°C see Figure 2.2, calculated by Liu et al. [58]. Thus under equilibrium conditions and no kinetic limitations, molybdenum oxides should  not form until all silicon in the alloys is consumed.  It is shown that when oxygen is added to MoSi2, SiO2 and Mo5Si3 are formed, then SiO2 and Mo3Si are formed from Mo5Si3, and finally SiO2 and MoO3 are formed from Mo3Si [58]. Furthermore, even though 13 the thermodynamic conditions are favorable for the formation of silica over molybdenum oxides such as MoO3, it is essential to take into account the kinetics, since the diffusion process plays an important role in oxidation. Sharif [33, 65] reported that the complex oxidation kinetics of MoSi2 results from (a) pesting oxidation, (b) formation of MoO3, (c) formation of different molybdenum silicides, i.e., Mo5Si3 and Mo3Si, during oxidation, (d) transformation of the SiO2 scale into its various phases, and (e) cracking and spallation of the oxide layer. Depending on experimental conditions, one or more of these occurrences may affect the kinetics of oxidation [33].   Figure 2.2 Calculated isothermal sections of Mo-Si-O ternary system at (a) 500°C and (b) at 1600°C. [58]  Liu et al. [58] proposed a linear relationship between the diffusion activation energy Q of silicon in MoSi2, and the melting point Tm using least square fitting of the experimental data for previous experiments for silicides and carbides.  𝑄𝐼𝑅= 22.278𝑇𝑚 − 10321 (4)  𝑄𝐼𝐼𝑅= 10.214𝑇𝑚 − 549.58 (5) 14  where the 𝑄𝐼 and 𝑄𝐼𝐼 values  were calculated as 338.9 kJ.mol-1 and 190.2 kJ.mol-1, respectively for the diffusion of silicon in MoSi2. There are three types of oxidation behavior for MoSi2, as classified by Liu et al. [58], including (i) oxidation below 600°C, (ii) from 600 to 1000°C, and (iii) above 1000°C. Below 600°C, the diffusion rate of ions is too low to permit the formation  of an external continuous silica layer, leading rather to the formation of MoO3. Below 600°C, the low bulk diffusion coefficient of Mo and Si in MoSi2 renders it impossible to sustain the silicon supply required for the formation of a continuous silica layer. This leads to the co-oxidation of silicon and molybdenum, forming a scale of mixed oxides at low temperatures [58].   Since oxidation of MoSi2 is a diffusion-controlled process, it is important to assess the diffusion coefficients of different ions in MoSi2. The bulk diffusion coefficients of oxygen in carbides and silicides near their melting points, DTM, is on the order of 10-11 m2.s-1. In this regard, Liu et al. [58] concluded that the bulk diffusion of  oxygen in  MoSi2 should be about 5.25×10-4 m2.s-1.  The bulk diffusion coefficients of oxygen in MoSi2, DL, can be approximated as [58]:  𝐷𝐿 = 5.25 × 10−4 exp (−338900𝑅𝑇) 𝑚2 𝑠⁄  (6)   The grain boundary diffusion coefficient of oxygen in MoSi2, DGB, is given by [58]:  𝐷𝐺𝐵 = 1.0 × 10−7 exp (−132700𝑅𝑇) 𝑚2 𝑠⁄  (7)  At the  oxidation temperature of MoSi2 (500°C), the DL and DGB values of  oxygen in MoSi2 are about 6.58×10-27 and 1.08×10-16m2s-1, respectively [58].  15 Liu et al. [58] showed that silicon oxide should be in equilibrium with MoSi2 at 500ºC; however, molybdenum oxides are also witnessed after oxidation tests at the same temperature. Therefore, the pest phenomenon could be a result of the low Si self-diffusion coefficient in MoSi2 at these temperatures, which hinders the formation of a silica protective layer, and consequently leads to the co-oxidation of molybdenum and silicon. It has been shown that the pest phenomenon occurs in air in the temperature range from 350 to 550ºC; at the higher temperatures of 600-700ºC, the pest reaction does not take place anymore, but the oxide formed on the surface is not protective [50, 58]. S. Knittel et al. [50] suggested that the pest oxidation occurs at low temperatures and can be controlled by composition, density, microstructure, and water partial pressure. Composition: It has been shown that the Mo-rich molybdenum disilicide is more prone to the “pest phenomenon” than a Si rich molybdenum disilicide. Therefore, the presence of the Mo5Si3 compound is unfavorable to the oxidation resistance [50]. Microstructure and density: the degradation of MoSi2 materials due to pest oxidation reactions are not limited to polycrystalline MoSi2. Crystallographic defects present in a single crystalline MoSi2 provide suitable nucleation sites for oxidation to initiate [51]. In this regard, Westbrook et al. [56] studied a 67% dense sample; the pest oxidation occurred after only 10 hrs of exposure at 400ºC, whereas occurred after 350 hrs for a MoSi2 with 14% porosity. It has been shown that sintered samples with 95% density, does not pest at 500ºC [50, 56]. Water partial pressure: Hansson et al. [59, 66, 67] demonstrated the influence of water partial pressure in the range of temperatures at which a peak of oxidation rate was observed. The authors showed that the peak oxidation rates occurred at temperatures of about 510ºC in O2, but it decreased to 470ºC in O2 with 10% H2O. This was attributed to the formation of molybdenum hydroxides, which volatilize at temperatures lower than the molybdenum oxides. The authors also associated the peak oxidation rate to the evaporation of the molybdenum species from the oxide scale, leaving it with an open structure. This was 16 thought to facilitate the diffusion of oxygen through the oxide scale and consequently to increase the oxidation rate [59, 66, 67]. Kurokawa et al. [52] was studied the effects of structure, concentration and H2O vapor pressure on oxidation of MoSi2 at 500°C. They observed the occurrence of simultaneous oxidation of Mo and Si, resulting in pesting, which was hindered by using fully dense MoSi2 in air and air H2O vapor atmosphere, whereas, porous MoSi2 showed the accelerated oxidation. The temperature range 600-1000°C is a transition range between a non-protective and a protective oxidation behavior of MoSi2. The diffusion coefficient of atoms becomes high enough above 1000°C to give rise to a continuous silica layer at the surface [58]. Above 1000°C, called high temperature here, MoSi2 exhibits oxidation through the formation of a protective (continuous) silica layer. This film is a coherent and adherent silica (SiO2) layer on the surface of MoSi2, thereby hindering oxygen access to the substrate. This is because of partial oxidation of MoSi2 on the layer’s surface, which is resulting in a glassy SiO2-surface. Since the protecting SiO2-film melts at 1723°C, it is suggested that pure MoSi2 can be used up to temperatures of about 1650°C [33, 38, 45, 68–73]. The schematic isothermal oxidation curves for MoSi2 in both low temperatures (below 1000°C) and high temperatures (above 1000°C) are shown in Figure 2.3 [58]. In temperature range 400-600°C, the formation of the volatile voluminous MoO3 in micro cracks increases the kinetics of low temperature oxidation, while the formation of protective thin silica layer above 1000oC inhibits further oxidation and subsequently reduces the kinetics of high temperature oxidation [58].  17  Figure 2.3 Schematic isothermal oxidation curves for MoSi2 at (a) low and (b) high temperatures [58]  Peizhong Feng et al. [71] studied the effects of high temperature pre-oxidation treatment on the temperature ranges of the pest oxidation (400-600°C). They showed that the material with high-temperature pre-oxidation treatment also had a relatively good low-temperature oxidation resistance. 2.6 Summary It has been shown that the oxidation behavior of MoSi2 is a function of temperature, fabrication conditions, microstructure, composition, density, and the oxidizing environment, e.g. containing oxygen and/or water . Furthermore, it has been shown that the pest oxidation phenomenon may be overcome by alloying and other adequate processing techniques; the low temperature brittleness, and the high temperature creep resistance of MoSi2 may be offset at least partly by using it together with a secondary material in a composite or an alloy. Today’s most advanced MoSi2-based heating elements, are very dense (~99 vol%) and alloyed with a secondary material, e.g, B, Al, W, to improve the mechanical properties and oxidation resistance of MoSi2 based components. Despite the substantial attention paid to understanding of the oxidation behavior of MoSi2, there are very few publications on the oxidation behavior of the industrial-grade, engineering MoSi2-based materials doped with Al and W. In order to be able to compare the oxidation behavior of the new industrial-grade, 18 engineering MoSi2-based materials, it is imperative to investigate these materials in same experimental conditions. Also, it is necessary to remove the oxide scale of the materials to make sure that all materials start the oxidation process in the same condition. It is also important to study the oxidation behavior over broader temperature ranges, i.e. 300-1600°C.   There is no published data on the oxidation of various MoSi2 materials over such a broad range of use temperature. Hence, the investigation of the oxidation behavior of MoSi2-based materials containing aluminum or tungsten is an important topic  which needs to be addressed to explain their behavior in different industrially relevant environments. This sets the rationale for the work undertaken in the present project.   19  3   Scope and objectives Chapter 3 Scope and objectives 3.1 Scope The objective of this study is to provide comparative analysis of the isothermal oxidation of five variants of commercially available MoSi2-based materials in the temperature range of 300-1600°C. There is no unified published data on the oxidation of various MoSi2 materials over such a broad range of use temperature. According to the literature, MoSi2 oxidizes at relatively low temperatures, below 1000°C, which can lead to its disintegration [42, 44, 49, 53, 55]. The formation of molybdenum oxide at lower temperatures leads to internal stresses owing to the significant volume change upon oxidation of MoSi2 (~4 times), and as a result of the fracture of the oxidation products which in turn accelerates the rate of oxidation. Above 1000°C, MoSi2 exhibits better oxidation resistance due to the creation of a continuous SiO2 layer (or alumina layer for the materials containing aluminum). However, during service, the protective scale on MoSi2 could be damaged due to erosion, volatilization, and micro-cracks in thermal cycling, or exposure to reducing atmospheres. Therefore, from a practical point of view, it is imperative to investigate the deleterious oxidation behavior of MoSi2 materials when such protective scales are not present. Also, it is necessary to remove the oxide scale of the materials to make sure that all materials start in the same condition.  The experiments in this study were divided into two main categories: low temperature oxidation LTO (300 to 900°C; high oxidation rate expected) and high temperature oxidation HTO (1000 to 1600°C; where low oxidation rate expected). 20 The broad scope of this project is to study the oxidation behavior of different Mo(Si, Al)2, Mo (Si,W)2, and other commercially available MoSi2 based materials over a broad range of temperatures and exposure times. This was done with the aim of providing a database of low and high temperature behavior of the materials studied, in particular for the possible applications in the Hot Surface Ignition Systems for NG engines. An attempt was also made to develop an understanding of the oxidation behavior of various grades of MoSi2-based materials for industrial applications as heating elements.  The scope of this work specifically involved the following: 1. Kanthal Super (KS), commercially available, MoSi2-based materials heating elements studied in this work, included KS-1700 and KS-1800 (clay-bonded MoSi2-based materials), KS-ER is an aluminum doped MoSi2 and KS-1900 and KS-HT are MoSi2 compositions containing tungsten, refer to Table 4.1 and Table 4.2. 2. Experimental specimens were prepared through cutting, grinding and finally removing the pre-oxidized silica (or alumina) layer of the outer surface of the samples. 3. In low temperature oxidation (LTO) behavior studies, the samples in the absence and presence of the pre-oxidized films were exposed to 300-900°C isothermally, for 4 to 240 hrs, and the mass change was determined.  4. In high temperature oxidation (HTO) behavior studies, the samples in the absence of the pre-oxidized films were exposed to 1000-1600°C isothermally. The exposure time was from 2 to 72 hrs. The mass change and the film thickness of the samples was determined. 5. The oxidation activation energy of each material composition at different temperature range was calculated. In addition, the thickness of the oxide scale layer formed on the surface was measured and predicted at different time and temperature exposures. 6. The microstructure and the phase compositions of the as-received and exposed samples were analyzed by Scanning Electron Microscopy (SEM), Energy-dispersive X-ray spectroscopy (EDS), 21 and X-ray diffraction (XRD). The differences in the oxidation behavior of various MoSi2-based materials were linked to their chemistry and phase compositions. 3.2 Objectives The overarching objective of this work is to study the oxidation behavior of Mo(Si, Al)2, Mo (Si, W)2, and MoSi2 based materials in air at 300-1600°C. The aim is to bring a better understanding of the causes of MoSi2-based materials deterioration in oxidizing environment. The objectives of this work attempt to answer the following research questions, each addressing a specific gap in the current literature on this topic: 1. Can the silica (or alumina) pre-oxidized layer formed on the surface of MoSi2 based material bars during manufacturing processes act as a protective layer against the low temperature (below 1000°C) oxidation LTO? Is it needed to heat-treat the samples at high temperatures (1000-1600°C) to form a uniform and protective layer against LTO before usage? Can the heat-treatment at high temperatures limit low temperature oxidation rate? 2. What is the temperature range of the low temperature oxidation for the composites containing aluminum or tungsten as compared to MoSi2 based materials? 3. At which temperature the rate of oxidation (below 1000°C) is maximum for each composition? What differences are there in low temperature oxidation behavior of the MoSi2 materials containing aluminum (or tungsten) from non-alloyed MoSi2 materials? 4. At what temperature the continuous protective silica (or alumina) film starts to form for each material? What are the differences in high temperature oxidation behavior of Mo(Si, Al)2, Mo (Si, W)2 as compared to non-alloyed MoSi2 based materials? What are the rates of high temperature oxide growth of each material?  5. What thickness is the oxide scale of silica (or alumina) formed at different time and temperature exposures?  22 6. What are the characteristics of the oxide layer formed for each isothermal the high temperature oxidation HTO experiment? At what temperature, the oxide layer formed in HTO can act as a protective and uniform layer against the low temperature oxidation LTO? The tasks undertaken to reach the above objectives can be detailed as follows: 1. Evaluate the effects of the pre-oxidized layer formed during the manufacturing processes on the low temperature oxidation behavior through exposing the materials at the temperature of 500°C. According to the literature, 500°C is reported as the peak temperature for the LTO of MoSi2 [43, 44, 49, 53, 55, 74].  To assess the ability of the high temperature heat-treated samples against the LTO, the samples with presence of the pre-oxidized layer are exposed at 1500˚C for 72 hrs to create a protective HTO layer, and then are subjected to 500, 750 and 1000˚C for up to 72 hrs. 2. Compare the low temperature oxidation behavior of MoSi2-based materials aged in air through exposing different types of Kanthal-MoSi2 samples at intermediate temperatures (300-900ºC). To accelerate the rate of degradation, the pre-oxidized layer of the samples was removed uniformly. This step was performed to hasten the onset of oxidation during the laboratory exposures, and to make sure that all materials start the tests in the same condition. The mass changes after exposure at different temperatures were determined to measure how significant the phenomenon can be. 3. The microstructure and the phase compositions of the surfaces and the sections of the exposed samples were analyzed by Scanning Electron Microscopy (SEM), Energy-dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD). The differences in the oxidation behavior of various MoSi2-based materials were linked to their chemistry and phase compositions. 4. Compare the high temperature oxidation behavior of MoSi2 based materials aged in air. This experiment aimed at studying the effect of the HTO through exposing different types of Kanthal-MoSi2 samples at temperature ranges between 1000-1600ºC. The ability of different MoSi2-based materials to form a protective silica (or alumina) film over a broad range of temperatures from 1000°C to 1600°C was investigated. 23 5. The oxide scale growth and thickness of silica (or alumina) was determined through the mass change values and SEM microstructure analyses. The oxidation activation energy was then calculated to quantify the oxidation behavior and the required minimum energy to start a chemical reaction of each composition. 6. The microstructure and the phase compositions of the surfaces/sections of the exposed samples were analyzed by Scanning Electron Microscopy (SEM), Energy-dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD). The ability of the heat-treated oxide scale against the low temperature oxidation behavior for each material composition at different temperatures was evaluated and discussed. In addition, the differences in the oxidation behavior of various MoSi2-based materials were linked to their chemistry and phase compositions.   24  4   Materials and experimental procedures Chapter 4 Materials and experimental procedures 4.1 Materials and sample preparation The experiments were divided into several different series, each with different goal and slightly different procedures. In this chapter, the materials used in this investigation and the experimental procedures will be elaborated in more details. The materials investigated in this study were commercial MoSi2-based compositions manufactured by Sandvik Heating Technology under the product name Kanthal Super, first introduced in 1956 [10]. Kanthal Super (KS) is basically a relatively dense (its porosity is typically < 1 vol%) cermet material consisting of MoSi2 and oxide components, mainly glass (or alumina for the materials containing aluminum). The service temperature of MoSi2-based materials has been raised from 1650°C in 1956 to 1850°C today, through intense research work on materials optimization.  According to the manufacturer, the materials are fabricated by mixing MoSi2 (or alloyed MoSi2) powders, water and for some of them clay used as a bonding phase [10]. The mixture is extruded into 6 mm diameter rods, dried, and then sintered at high temperatures in protective atmosphere (above 1000°C). During the final manufacturing step, the materials are "pre-conditioned", i.e. a silica (or alumina for the materials containing aluminum) scale  forms on the surface of the materials through extended high-temperature treatment in air [10]. After sintering, the materials are more than 99% dense and consist mainly of MoSi2, Mo(Si, W)2, and/or Mo(Si, Al)2 phases [10]. There are different types of Kanthal Super MoSi2-based materials, which are designed for different areas of applications. The maximum service temperature of the heating elements, reported by Kanthal [10], 25 in both oxidizing and reducing environment is presented in Table 4.1, and density and elemental compositions of the materials is shown in Table 4.2. Table 4.1 The maximum service temperature of the Super Kanthal heating elements, as recommended by Kanthal [10] Kanthal Super Type: Maximum Service Temperature (°C) Oxidizing environment Reducing environment KS-1700 1700 1150-1450 KS-1800 1800 1150-1450 KS-1900 1850 1150-1450 KS-HT 1840 1150-1450 KS-ER 1580 1450-1550  Table 4.2 Density and elemental compositions of MoSi2-based materials, as measured in the ceramic lab at UBC Kanthal Super Type: Elemental Compositions wt% Density O Al Si Mo W g. cm-3 KS-1700 15-20 0-2 30-36 40-46 0 5.6 KS-1800 15-20 0-2 30-36 42-48 0 5.6 KS-1900 15-20 0 25-30 22-26 32-37 6.5 KS-ER 15-20 15-20 18-23 45-50 0 5.6 KS-HT 5-10 0 22-26 32-36 25-30 7.1  4.1.1 Kanthal Super 1700 and 1800 KS-1700 and KS-1800 are clay-bonded MoSi2. The maximum service temperature of KS-1700 and KS-1800 are 1700°C and 1800°C, respectively [10]. The typical areas of applications of KS-1700 include industrial furnaces for sintering, glass melting and refining, and heat treatment forging, and in radiant tubes. KS-1800 is mostly used in laboratory furnaces, testing equipment and high temperature sintering production furnaces. 4.1.2 Kanthal Super 1900 and HT KS-1900 is MoSi2 doped with tungsten; the material has the same similar core characteristics as KS-1700 and KS-1800 but is higher in purity (i.e. through reduced Fe content, according to manufacturer's specifications). Maximum operational temperature of KS-1900 is 1850°C. The typical area of applications for Kanthal Super 1900 includes laboratory furnaces, testing equipment, tube furnaces, diffusion furnaces 26 and glass feeders [10]. KS-HT is MoSi2 doped with tungsten; this material is designed for a longer lifetime of small dimension elements at high temperatures (T > 1200°C) when cycling. The hot strength and form stability is also improved [10]. The lower oxidation rate of KS-HT results in a less adhesion to the furnace insulation fibers due to formation of a thinner glaze layer upon oxidation. According to the manufacturer, this also improves the lifetime, because the stresses are reduced between the base material and the surrounding oxide, depending on the difference in thermal expansion coefficients. This is of great importance especially for elements of smaller dimensions used in thermal cyclic conditions, where Kanthal Super 1800 and 1900 elements may be damaged by "banding" [10]. The banding effect is when the element shatters into smaller pieces during thermal cycling. The maximum operating temperature of this material is 1830°C, and the element is suitable for furnace temperatures between 1500-1750°C [10]. 4.1.3 Kanthal Super ER KS-ER is MoSi2 doped with aluminum and the scale that forms at high temperature on the surface is alumina. The major reason for the outstanding features of KS- ER heating elements is that it has a protective surface of alumina. The maximum service temperature of KS-ER is however only 1580°C [10] with the ability to operate directly in a wide range of furnace atmospheres from dry reducing to oxidizing. With Kanthal Super ER heating elements, it is possible to operate firing cycles where the atmosphere can be altered during the cycle between oxidizing, inert, carburizing, nitriding, reducing and rough vacuum. KS-ER can be used in direct contact with high alumina supports without any corrosive reactions. The microstructures of all different unexposed MoSi2-based composite materials are shown in the scanning electron microscopy (SEM) backscattered electron (BSE) image in Figure 4.1. The gray background of KS-ER corresponds to Mo(Si, Al)2, the black regions are alumina, and the white regions are Mo5(Si, Al)3.  27                  Figure 4.1 SEM backscattered electron image of the bulk material in an unexposed Kanthal commercial heating element samples viewed in cross-section  The gray background of KS-1700 and KS-1800 corresponds to MoSi2, the black regions are clay and or silica, and the white regions are Mo5Si3. Likewise, the gray background of KS-1900 and KS-HT corresponds to Mo(Si, W)2, the black regions are silica, and the white regions are Mo5(Si, W)3. The overall KS-1700 KS-1800 KS-ER KS-1900 KS-HT-1 28 composition of the unexposed MoSi2-based materials was measured by energy dispersive X-ray analysis, Table 4.2. 4.1.4 Sample preparation The materials were supplied in the form of ~30 cm long cylindrical rods. The specimens were cut into 21 mm long and 6 mm in diameter using a high-speed diamond saw. The samples were then ground using silicon carbide media from 60 to 320 grit wet sandpapers. Planar mechanical grinding was performed with the intention to create a flat sample surface after sectioning. Planar grinding was considered  complete when the samples have been flattened to a common plane. Water was used as a coolant and lubricant; also, water considerably improves the surface quality of the samples. fperio    Figure 4.2 Different types of Kanthal Super bars MoSi2 after the mechanical removal of the protective layer KS-1700 KS-1800 KS-HT KS-1900 KS-ER 29  Figure 4.3 Prepared specimens for the experiments after removing the protective layer  Following mechanical machining, the next step was cleaning in an ultrasonic bath for 20 minutes with acetone. The samples were then dried in laboratory electric dryer. The dimensions were measured, and afterwards, the weights were measured using a microbalance with accuracy of 0.0001 g. The surface area and volume were calculated. Also, the density of each material was estimated based on their weight and the regular cylindrical geometry (see Table 4.2). Prior to exposure to heat, the crucibles were cleaned by aging at 700°C for 2 hrs to remove any possible contaminations.  4.2 Design of experiments 4.2.1 Heat treatment schedule in LTO (300-900°C) The samples were placed in an air-furnace at designated temperatures and were allowed to oxidize for 4, 12, 36, 72, 84, 120, and 240 hrs. Isothermal low temperature oxidation experiments were performed at different temperatures in air in an electrical-resistance furnace. Furnace heating and cooling rate was about 10°C.min-1, but during cooling and heating of the furnace, the samples were out of the furnace. In all temperatures during experiments, first heated the furnace up to the oxidation temperature and then put the 30 tray of the crucibles with the samples all at once in the furnace so the temperature of the samples jumped from room temperature to the specified temperature for each experiment.  Figure 4.4 shows the heat treatment schedule for the experiments at 900°C and it was followed the same regime for other experiments in the temperature range 300-900°C.  Figure 4.4 Heat treatment schedule for the experiments at 900°C   Each cycle of the LTO tests consisted of: 1. Heat up the furnace from room temperature to maximum temperature at 10°C. min-1 (without samples) 2. Put the samples inside the furnace (The heating rate for the samples estimated 100°C.min-1) 3. Hold for designated incremental time exposure at maximum temperature 4. Take out the samples from the furnace (The cooling rate for the samples estimated 100°C. min-1)  01002003004005006007008009000 40 80 120 160 200 240Temperature °CTime(h)31 4.2.2 Heat treatment schedule in HTO (1000-1600°C) The samples were placed in an air-furnace at designated temperatures and were allowed to oxidize for 2, 12, 36, 72, and 144 hrs. Isothermal high temperature oxidation experiments were performed at different temperatures in air in an electrical-resistance furnace. The test was performed with the heat treatment schedule shown in the Figure 4.5 for high temperature oxidation HTO experiments. Figure 4.5 shows the heat treatment schedule for the experiments at 1600°C and it was followed the same regime for other experiments in the temperature range 1000-1600°C.   Figure 4.5 H Heat treatment schedule for the experiments at 1600°C  Each cycle of the HTO tests consisted of: 1. Heat up the furnace from room temperature to 850°C (based on the literature review in the Chapter 2, the LTO does not happen for MoSi2, above 850°C) at 10°C.min-1 (without samples) 020040060080010001200140016000 20 40 60 80 100 120 140 160 180Temperature °CTime(h)32 2. Put the samples inside the furnace (The heating rate for the samples estimated 100°C.min-1) 3. Heat up the furnace from 850°C to maximum temperature at 10°C.min-1 (with samples) Note: Based on the literature review provided in the Chapter 2, the LTO does not happen for MoSi2 above 850°C. 4. Hold for designated incremental time exposure at maximum temperature (with samples) 5. Cool down the furnace from maximum temperature to 850°C at 10°C.min-1 (with samples) 6. Take out the samples from the furnace (The cooling rate for the samples estimated 100°C. min-1) Oxidative exposures were conducted isothermally in air at temperatures of 1000, 1200, 1300, 1400, 1500, and 1600°C from 2 to 144 hrs (5 days) with two samples of each material for each exposure temperature in low temperature oxidation experiments. The total exposure time in the electrical furnace was 144 hrs for high temperature oxidation experiments, respectively. The weight of the samples were measured and recorded both before and after each exposure time in room temperature. 4.2.3 Experimental conditions in low and high temperature oxidation (300-1600°C) In low and high temperature oxidation experiments, as shown in the Figure 4.6, the samples were placed in a way to maximize the surface area and hence overall mass gain/loss for improved accuracy and precision of the gravimetric method used in this study. After each exposure time, the specimens were reweighed to determine weight changes. Cumulative weight changes of the samples were calculated and reported as a function of oxidation time. 33   Figure 4.6 (a) Sample arrangement during LTO experiments before and after exposure  4.2.4 Thermal Analysis (TGA / DTA) of MoSi2-based materials Thermo-gravimetric analysis (TGA) and differential thermal analysis (DTA) were used to investigate the stability and decomposition of MoSi2 based materials. The experiment was carried out at the University of British Columbia on approximately 25 mg of powdered sample of KS-ER with the particle size 10 to 20 µm, in an alumina crucible under air using a PerkinElmer STA-6000 TGA/DTA, Figure 4.7. TGA data was baseline-corrected using data collected on an empty crucible. The test was performed with the following program for the experiment. 1. Switch gas to air at 20 ml.min-1 2. Hold for 1min at 30°C 3. Heat from 30°C to 400°C at 10°C.min-1 4. Hold for 1min at 30°C 5. Heat from 400°C to 995°C at 2°C.min-1 6. Hold for 5min at 995°C  34  Figure 4.7 PerkinElmer STA-6000 – Simultaneous TGA-DTA used in this investigation  The thermobalance was equipped with a double-symmetrical furnace, where an Al2O3 sample, having the same dimension as the samples, was used as a reference in order to reduce the buoyancy effects. Before and after exposure the samples were weighed in a balance with 0.05 mg readability in order to calibrate the mass gains. At each experiment, the exposures were carried out at least twice in order to establish that the results were reproducible. The sensitivity of the mass gain measurements in the TGA system was 0.05 mg. The gas atmosphere was dry synthetic air (20% oxygen and 80% nitrogen) with a flow rate of 0.8 cm/s and a water concentration of approximately 5 ppm.  4.3 Uncertainties and errors Two samples from each material/condition were tested during all experiments. The weight of the samples were measured three times at room temperature by a high precision electronic weighing balance 35 with a sensitivity of ± 0.0001 g before and after exposure times at room temperature during the oxidation test. The mass gain in milligrams per square centimeter of surface area was calculated for each specimen.  The temperature difference of the temperature controller of the furnace and the temperature of the samples in all experiments was less than 25°C.  4.4 Characterization techniques Following thermogravimetric measurement of oxidation, scale surfaces and/or cross sections of specimens were examined with a scanning electron microscope (SEM) equipped with an energy-dispersive X-ray (EDX) analyzer and an X-ray diffraction (XRD). Upon completion of the experiments, the test samples were sectioned and polished. The oxide film thickness measurements were performed on micrographs of the cross section of the oxide scale obtained by scanning electron microscopy (SEM). The X-ray spectrum provided us information on the chemical composition of the materials. SEM-EDX was used for greater magnification study of the microstructure and the elemental compositions of the samples before and after experiments. To identify the phase compositions of the samples before and after experiments, the X-ray diffraction analysis was performed.   Sample preparation for purposes of examination in the SEM was simple. The samples were mounted in epoxy thermosets - also known as embedding or encapsulation. This has proven to be an advantageous method of preparing ceramic samples for grinding and polishing, because of the easier manipulation and improved edge retention it makes possible. The vacuum impregnation represents a simple and effective method for impregnating a sample with epoxy resin. The samples were placed in embedding molds or other small containers and then transferred to the vacuum vessel. The samples together with the resin were evacuated to a pressure of approximately 1000-1500 Pa (10-15 mbar) to ensure that the epoxy resin was free of bubbles after being mixed. Specimens intended for examination by a scanning electronic microscope were subjected to especially intense cleaning. The specimens were prepared for cross-sectional analysis by grinding on SiC papers from 60 to 1200 grit using water as a lubricant. Silicon carbide (SiC) wet abrasive 36 papers are especially well suited to the grinding of ceramic samples and samples that tend toward pull-outs. The SiC abrasive was harder than the MoSi2 samples. Due to the weak bonding of the abrasive grains, wet abrasive papers work quite gently because the SiC papers wear down within a few minutes and, thus, continuously change to the next finer grain size. 4.5 Test samples nomenclature  The compilation of all types of materials used in this study, their commercial names and phase compositions, as well as nomenclature used in this study, appears in the below Table 4.3. Table 4.3 Test samples nomenclature used in this study Materials Major phases (>95%) Commercial names Materials classification based on major phase composition Nomenclature for the materials studied here Kanthal Super 1700 MoSi2, Mo5Si3, SiO2 KS-1700 Clay bonded MoSi2 based materials MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials Kanthal Super 1800 MoSi2, Mo5Si3, SiO2 KS-1800 Kanthal Super 1900 (Mo, W)Si2, (Mo, W)5Si3, SiO2 KS-1900 (Mo, W)Si2 based materials Kanthal Super HT (Mo, W)Si2, (Mo, W)5Si3, SiO2 KS-HT Kanthal Super ER Mo(Al, Si)2, Mo5(Al ,Si)3, Al2O3 KS-ER Mo(Al, Si)2 based materials     37  5   Initial results and experiments Chapter 5 Results of the preliminary LTO experiments 5.1 The effects of the pre-oxidized layer of MoSi2-based materials  During the manufacturing process, an oxide protective scale forms on surface of MoSi2 materials. This scale layer is called “the pre-oxidized layer” in this study. The ability of the pre-oxidized layer to protect the rest of the materials against low temperature oxidation LTO was studied. The effect of the pre-oxidized layer on the surface was studied at 500°C. It is reported that  the maximum LTO rate of MoSi2 occurs  at this temperature [43, 44, 49, 53, 55, 74]. In this investigation, three samples of KS-HT (MoSi2 doped with tungsten), KS-ER (MoSi2 doped with Al) and a commercial clay bonded MoSi2  were exposed to  500˚C for 186 hrs in air. The microstructure of KS-HT was evaluated under SEM, as seen in Figure 5.1(a): the composition of the cross section in the areas far from the surface did not change during thermal aging at 500˚C. The unexposed sample of KS-HT had an overall elemental composition of 6 wt% O, 22 wt% Si, 40 wt% Mo, 32 wt% W. This material is composed of MoSi2+WSi2; it also contains MoB+WB; however, as boron is a very light element, it is difficult to quantify by EDX, so boron does not appear in the given elemental compositions. The thickness of the surface silica layer was measured at 4-12 µm for the as-received, and 4-8 µm for the exposed sample on the cross section fractured surface. For both samples, the region adjacent to the silica layer consists of (Mo, W)5Si3, which contains less Si than the disilicide. The Mo:W ratio is ~2:1 in the matrix (medium-contrasted) and it is reversed, ~1:2 in the light-contrasted areas. The dark regions are pockets of silica. Therefore, it seems the silica layer (formed during manufacturing processes) on KS-HT was protective against the low temperature oxidation. 38  Figure 5.1 The appearance and microstructure of (a) (Mo, W)Si2, (b) MoSi2, and (c) Mo(Si, Al)2 based materials after 186 hrs. at 500°C  The low temperature oxide film formed at 500˚C in the first few hrs of exposure on the free surfaces of the exposed sample of the clay bonded MoSi2 materials, and grew continuously for the duration of the test, as it can be seen in Figure 5.1(b). The silica layer of the unexposed sample was 2-3 µm thick and the overall composition was 21 wt% O, 1 wt% Al, 35 wt% Si, 43 wt% Mo. The overall composition of the unexposed and exposed sample was similar, except for the region adjacent to the thick silica scale, which had a higher concentration of Mo5Si3 (i.e. it was silica poor). In both cases, the grey area is the MoSi2 matrix, the light-contrasted areas are Mo5Si3, and the dark areas are pockets of silica. The pre-oxidized silica layer on the surface of MoSi2 was not protective against the low temperature oxidation. An oxide layer of thickness of 27-35 µm formed around the circumference in the exposed sample of clay bonded The pre-oxidized layer The LTO oxide film (a) (b) (c) 3.5mm 5.5mm 2.5mm silica Mo5Si3 MoSi2 (Mo, W)Si2 (Mo, W)5Si3 Mo(Si, Al)2 Mo5(Si, Al)3 Al2O3 Silica 39 MoSi2 materials at 500˚C after 186 hrs; this was a mixed layer, consisting mostly of silica and molybdenum oxide, see Figure 5.2.   Figure 5.2 The microstructure of the cross section near the surfaces of clay bonded MoSi2 based materials exposed to air for 186 hrs at 500°C  According to the manufacturer's specifications [10], KS-ER is specially formulated to withstand alternating oxidizing and reducing atmospheres; Al substitutes for about half of the Si in MoSi2, so the general formula of this material is Mo(Si, Al)2. This result in an alumina, instead of silica layer forming on the surface, and alumina is very resistant to reducing conditions [10, 75–78].  The unexposed KS-ER has an overall elemental composition of 17 wt% O, 18 wt% Al, 19 wt% Si, 47 wt% Mo, determined by EDX on 3 areas throughout cross-section. This changes close to the surface into a composition with only 10 wt% Al; the region immediately adjacent to the alumina layer has only 5 wt% Al, as the Al diffuses to the surface to form the oxide. KS-ER was much less affected by oxidation at 500˚C. The composition throughout the cross section was similar to what it was before the thermal aging, see Figure 5.1(c). The oxide layer formed at 500°C Clay bonded MoSi2 based materials 40 The preliminary results have shown that the pre-oxidized layer formed on the surface of the various specimens had different response to oxidation at 500˚C. The silica layer of KS-HT was protective, but the silica layer formed on KS-1700 sample was not protective against oxidation at 500˚C. KS-ER has exhibited excellent thermal oxidation resistance at the low temperature of 500˚C; no considerable changes in compositions and structure were observed in the cross sections of KS-ER. This experiment helped us to design the experiment properly by removing the pre-oxidized layer of the samples before heat treatments to make sure that all materials start in the same condition. 5.2 The effect of high temperature pre-treatment of MoSi2-based materials The ability of MoSi2 based materials to form and maintain a protective silica/alumina film over high temperature heat treatment and the ability of the oxide film formed on the surface of the materials against low temperature oxidation was the aim of this investigation. The samples heat treated (pre-oxidized) at 1500°C for 66 hrs were later exposed to different temperatures (below 1000°C) to test the effect of the high temperature pre-oxidation treatment on thermal oxidation rate against low temperature oxidation. The materials investigated here were KS-1700 and KS-HT, which formed a silica layer, and KS-ER, which formed an alumina layer. All three samples were subjected to oxidation at 500˚C, 750˚C and 1000˚C, for times up to 72 hrs. The mass changes versus exposure time for each material are presented in Figure 5.3. The results showed the pre heat treated samples were resistant against low temperature oxidation since there is no considerable mass changes after exposure at 500˚C, 750˚C and 1000˚C. The glass layer acted as a barrier between the material and the ambient air at different temperatures, as no considerable mass gain was recorded during the experiments. An alumina layer was formed on the surface of KS-ER at high temperature (1500˚C), and the oxidation rate stabilized after 66 hrs. 41    Figure 5.3 Mass changes versus time at a) 500°C, b) 750°C, c) 1000°C , for different samples pre-oxidized at 1500°C for 66 hrs.  Each point is an average of two samples tested.  1000°C 500°C 750°C 42 The heating and cooling rates were relatively low (10˚C.min-1) during the experiments. Therefore, it can be anticipated that the oxidation layers were free of cracks or micro cracks. The results show that the high temperature pre-oxidation treatment significantly improves the low temperature oxidation resistance of MoSi2 based materials (below 1000°C).  5.3 Thermogravimetric analysis Thermo-gravimetric analysis (TGA) and differential thermal analysis (DTA)  were used to investigate the stability and decomposition of MoSi2 based materials. Figure 5.4 shows that a considerable mass gain from initial weight of 31 mg to 40 mg observed during experiment the temperature between 400˚C and 830˚C, but a considerable mass loss from 40 mg to ~27 mg between 830˚C to 1000˚C. According to the literature, MoO3 is volatile above about 800°C and evaporates after its formation, causing a mass loss [79]. It can be speculated from Figure 5.4 that the formation of the oxidation products of KS-ER, e.g, MoO3, Mo5(Si, Al)3, SiO2 started from 400°C and gradually continued, then reached  the  maximum at 810°C. At this temperature MoO3 started to volatilize indicating mass loss and then deposited on the inner surface of the lid, see Figure 5.5(a,b,c). The peak heat flow at 810°C was indicative of a phase change at this temperature and confirmed that the MoO3 evaporated at this temperature, see Figure 5.4. Above the melting point of MoO3, oxidation is accompanied by a total mass loss MoO3 condenses in the cooler parts of the apparatus. The oxide products of the experiment then were collected and identified. The needle-shaped crystalline materials formed on the upper part of crucible, Figure 5.5(a), from volatilized MoO3 during TGA process. Green powder formed on the surface of the lid, as seen in Figure 5.5 (c), was MoO3 was a mixture of MoO3, Mo5(Si, Al)3, SiO2. Alumina was recognized by SEM Alumina in the crucible, the white powder shown in Figure 5.5(c). The whitish powder shown in the Figure 5.5(c) was detected as Al2O3. The equipment was damaged because the reactions formed and plated onto nearby surfaces of the instrument.   43   Figure 5.4 TGA and DTA curve of KS-ER powder at temperature (400-995°C) 44  Figure 5.5 The reactions formed, volatilized, and deposited from KS-ER powder in surrounding of the TA Instrument resulted after experiment  5.4 Summary and conclusions The results of the preliminary LTO experiments helped us to design the experimental work properly. Figure 5.6 shows a schematic heat treatment procedure of the experimental work all material studied over the temperature range 300-1600°C. As it has been described in detail in the Chapter 4, the specimens were cut into 21 mm long and 6 mm in diameter using a high-speed diamond saw. The protective layer of the samples were then shaved and removed from the surface by using a lathe machine. The samples were placed in an air-furnace at designated temperatures and were allowed to oxidize for different time exposures. Samples were removed periodically from the furnace and their weight gain measured (before and after each time exposure) during experiments. (a) (a) (c) MoO3 whiskers (b) (c) Oxidation products of powdered KS-ER 45  Figure 5.6 the heat treatment procedure of the experimental work    46  6   Oxidation behavior of clay-bonded MoSi2 based materials Chapter 6 Oxidation behavior of clay-bonded MoSi2 based materials This part includes an investigation into oxidation behaviors of two types of clay bonded MoSi2-based materials namely Kanthal Super 1700 (KS-1700) and Kanthal Super 1800 (KS-1800). The experiments in this study were divided into two main categories: low temperature oxidation LTO (300 to 900°C; high oxidation rate expected) and high temperature oxidation HTO (1000 to 1600°C; low oxidation rate expected). 6.1 Low temperature oxidation (LTO) of clay-bonded MoSi2 based materials (300-900°C) The isothermal exposure temperature was from 300°C to 900°C for every 100°C and the time exposure was from 4 to 240 hrs (10 days for each temperature) in these low temperature oxidation studies. Figure 6.1 shows the mass variation versus time for KS-1700 and KS-1800. Each point represents average of six measurements of two samples and the bars indicate one standard deviation. The specimens studied here were 21 mm long and 6 mm in diameter. To accelerate the rate of degradation, the pre-oxidized layer of the samples was removed uniformly from the samples before the tests. This step was performed to hasten the onset of oxidation during the laboratory exposures, and to make sure that all materials start the tests in the same condition.  An interesting finding of this investigation was formation MoO3 at 300°C after 240 hrs. As shown in Figure 6.1, at 300°C mass changes started to increase after 120 hrs exposure for both material studied here. After longer periods (>120 h) of oxidation, the sample surfaces begin to show signs of pesting. The reaction resulted in whitish spots on the surface of samples. SEM analysis identified the needle shaped formation of MoO3 on the surface of KS-1700, and KS-1800 samples, see Figure 6.2. It was confirmed the oxidation of 47 MoSi2 occurred even at this low temperature. There is no publication reporting active oxidation of MoSi2 at a temperature as low as 300°C.   Figure 6.1 Mass variations versus time of KS-1700 (left) and KS-1800 (right) (300-900°C)  The overall elemental composition of the surface deposit on the surface of KS-1700, and KS-1800 (as identified by SEM-EDX) corresponds to a mix of MoO3 and silica, with impurities of Na, Mg, Al, Ca, and Fe.  Compared to the composition of the unexposed KS-1700 (35 wt% O, 25 wt% Si, 35 wt% Mo, 2 wt% impurities - Na, Ca, Mg, Al, Fe) this has more impurities (3 wt% overall, but even 8 wt% in specific positions). The "impurities" are defined here as anything beyond Mo, Si, O, e.g. Na, Ca, Mg, Al, Fe.  The surface of the exposed samples contains MoO3 and SiO2 in different proportions, and in some areas a pocket 48 of impure silica is also seen. It was found that the pest disintegration most likely occurred through transporting oxygen into the interior of the specimen of clay bonded MoSi2 based materials, KS-1700 and KS-1800, likely along the exist cracks and/or pores, where it reacted to form MoO3 and SiO2. The internal stresses produced during the formation of MoO3 resulted in disintegration of the sample to powders.     Figure 6.2 The formation of needle shaped MoO3 on the surface of KS-1700/1800 at 300°C and 400°C after 240 hrs  Images of the cross-sections of KS-1700 and KS-1800 after 240 h at 500°C are presented in Figure 6.3; the overall elemental composition did not change between the center and the periphery of the section, indicating that there was no significant internal degradation. The MoSi2 matrix composition (medium-contrasted, white arrow) was also constant; the dark-contrasted regions are silica, with ~1 wt% impurities (Na, Mg, Al, Ca, Fe); the light regions are Mo5Si3. There is no silica layer left, nothing to see and measure on the surface KS-1700 and KS-1800 at 500°C after 240 hrs. It was just observed the isolated particles of MoO3, but did not cover the surface. MoO3 (a) KS-1700-300°C (b) KS-1800-300°C (c) KS-1700-400°C (d) KS-1800-400°C 49     Figure 6.3 The surface and section of (a,c) KS-1700 and (b,d) KS-1800 at 500°C after 240 hrs  At 500°C, there was no considerable mass change for KS-1700. Additionally, The SEM/EDX of the surface of KS-1700 at this temperature was indicating that there was no significant oxidation on the surface or internal degradation, Table 6.1. XRD analysis of KS-1700 after 240 h at 500°C is shown in Figure 6.4 the surface of the sample shown only MoSi2, not seen oxides of Mo and Si. This might be attributed to the high density of KS-1700 and KS-1800 (˃95 vol%). It has been shown that fully dense (above 95%), crack free MoSi2 does not pest [6, 52, 66]. Table 6.1 EDX of the surface of KS-1700 (all data in wt%)  O Na Al Si Fe Mo As-received 34.66  1.19 28.33 0.72 35.11 240 hrs at 500°C 35.29 1.72 0.78 28.55  33.65    The EDX analyses of the exposed samples of KS-1700  and KS-1800 after 240 hrs at 500°C performed on the cross-section indicate that the overall composition were constant throughout the section, and similar to that of the unexposed samples, the matrix, silica inclusions and the Mo5Si3 particles were also similar.  MoO3 (c) KS-1700-500°C (a) KS-1700-500°C (d) KS-1800-500°C (b) KS-1800-500°C 50    Figure 6.4 XRD analysis of the sample surface of KS-1700 after 240h at 500°C  Intensity 51 After 240 hrs at 700°C, the silica deposit on the surface of KS-1700 was not uniform and it was extensively cracked with pockets of silica and isolated particles of MoO3, Figure 6.5, hence it was not protective anymore. The dark areas in the Figure 6.5 showed the low concentrations of Mo and high levels of Si and O, meaning that the silica deposit was very thick, while the high levels of O and Mo indicate that they were mostly MoO3. The silica layer also contained small amounts of impurities originating mostly from the alumino-silicate (clay) used as a binder during manufacturing [10]. The overall EDX elemental composition (cross-section) of the as-received KS-1800 after 240 h at 700°C is 16 wt% O, 36 wt% Si, 47 wt% Mo, with 1 wt% impurities - Na, Ca, Mg, Al, Fe. The overall composition of the surface of the exposed sample is 52 wt% O, 28 wt% Si, 19 wt% Mo, with 1 wt% impurities - Na, Ca, Mg, Al, Fe.     Figure 6.5 The surface microstructure of KS-1700and 1800 at 500 and 600°C after 240h  Similar to KS-1700, the dark-contrasted areas of the surface of KS-1800 after 240 hrs exposure at 600°C, in Figure 6.5, are silica; there are numerous and some very large, up to 40 µm, silica inclusions; the light-contrasted areas are Mo5Si3, and the medium grey is the MoSi2 matrix; the white line adjacent to the silica film is Mo5Si3. The amount of impurities seems to be lower than in KS-1700, but they are still up to 3 wt% in the silica inclusions with ~5 wt% impurities.  The surface discoloration of the samples heat treated  for 240 h at 400-900°C is presented in Figure 6.6. The formation of different low temperature oxidation products can be seen even with the naked eye. This KS-1800-600°C-X KS-1700-700°C 52 kind of oxidation can lead in time to the disintegration of the material. The dark and dull color of the specimens indicates the presence of MoSi2 on the surface (i.e. for KS-1700 at 500°C). The shiny dark (i.e. for KS-1700 at 800°C, dark areas) indicates amorphous silica. The whitish color (i.e. for KS-1700 at 700°C and 800°C and for KS-1800 at 400°C) is for a mix of MoO3 and SiO2.  Figure 6.6 Surface discoloration of the KS-1700 and KS-1800 specimens after 240h (400-900°C)  According to the literature, MoO3 is volatile about 800°C and evaporates after its formation, causing a mass loss [79]. Figure 6.7 shows mass variations versus time and micrograph of the KS-1700 surface after 240 hrs at 800°C. In the first 24 hrs exposure of KS-1700 sample at 800°C, the oxidation proceeds by the initial formation of MoO3 crystals and amorphous SiO2 on the surface and the MoO3  then evaporates. 53  Figure 6.7 Micrograph of the KS-1700 surface and mass changes after 240 hrs at 800°C  As it is shown in Figure 6.7, the formation of silica and molybdenum oxide occurred simultaneously and the mass gain was continuously increasing; after 24 hrs, mass gain stopped and MoO3 started to evaporate;  the loss of mass stopped when all MoO3 depleted from the surface, at around 100 hrs of the exposure time. The micrograph in Figure 6.7 shows the region where MoO3 evaporated from the surface of KS-1700. Kanthal Super 1800 had a better resistance against the low temperature oxidation than KS-1700. 6.2 High temperature oxidation (HTO) of MoSi2 based materials (1000-1600°C) At temperatures higher than 1000°C, compact scales of alumina and silica provide the best resistance to oxygen diffusion because of their high thermodynamic stability and low diffusivities for both cations  and anions [6, 9]. For high temperature applications, it is desirable to have a continuous and coherent scale of SiO2 or Al2O3, the formation of which often comparable with the oxides of other alloying elements or composites [9]. At higher temperatures up to 1300°C, alumina is  an excellent barrier to oxygen, while above 1300°C oxygen permeation  through silica occurs at a lower rate [6]. High temperature oxidation product of four types of materials studied here (KS-1700, KS-1800, KS-1900, and KS-HT) is silica and the scale forms at high temperature on the surface of KS-ER is alumina. The mass variation after 72 hrs of exposure from 1000°C to 1600°C, every 100°C, is presented in Figure 6.8 and Figure 6.9 for KS-1700 and KS-1800, respectively. The plots of specific weight gain versus KS-1700-800°C Depletion of MoO3 54 time for isothermal oxidation of both KS-1700 and KS-1800 at 1000, 1200, 1300, 1400, 1500, and 1600°C, shown in Figure 6.8 and Figure 6.9, started with a rapid weight gain followed by a slower rate of oxidation. During the initial part of high temperature oxidation, when a protective oxide layer had not yet formed, oxygen was readily available to react with silicon, and a rapid oxidation rate at 1400, 1500, and 1600°C was observed. Clearly, there is a change in the rate of oxidation with increasing temperature. The rate of oxidation at 1500 and 1600°C was extensively higher than that of 1300 and 1400°C, while it was observed an insignificant mass gain in the temperature range 1000-1200°C. In the first few hours, it is assumed the linear oxidation was happening (higher rate) and then the linear oxidation regime ends and a parabolic regime (slower rate) starts when an oxide layer covers surrounding the sample. In this period of exposure time, Si and O2 must diffuse through the bulk material and the initial oxide layer, respectively, to arrive at the SiO2–MoSi2 interface for oxidation reaction to take place.    Figure 6.8 Mass variations versus time of KS-1700 KS-1800 in the temperature range 1000-1600°C 55  Figure 6.9 Mass variations versus time of KS-1800 in the temperature range 1000-1600°C  It has been shown that oxidation mechanism proceeds by inward movement of  oxygen through SiO2 to the MoSi2−SiO2 interface where the reaction Si(solid) + O(gas) → SiO2 takes place [53, 73, 80–82]. The excellent high-temperature oxidation resistance of MoSi2 is attributed to the favorable thermodynamic conditions of silica formation over molybdenum oxides [58]. The SiO2 scale formed on the surface of MoSi2 material is essentially responsible for the protection of the substrate. This amorphous scale layer is preferred to the crystalline state, because the grain boundaries provide short circuit paths for diffusion in the latter [6].  Micrographs of the KS-1700 and KS-1800 samples after 72 and 144h isothermal exposure in the temperature range 1000-1600°C are shown in Figure 6.10 and Figure 6.11. As can be seen in the pictures, at high temperatures, Clay bonded MoSi2 based materials (KS-1700 and KS-1800) form a coherent and adhesive passivation layer of glassy silica (SiO2) that protects it from further oxidation and provides superb oxidation resistance. However, the sample oxidized at 1500 and 1600 °C shows relatively large mass variations, Figure 6.8 Figure 6.9, and large variations in oxide thickness, Figure 6.10 and Figure 6.11. A 56 third phase, Mo5Si3, was observed under the SiO2 scale of the samples of KS-1700 and KS-1800 at almost all temperatures above 1000°C.   The thickness of the silica scale formed on the surface of the KS-1700 after aging for 144 hrs at 1600C was ~30 µm, i.e. 10 times thicker than for the as-received and 8 times thicker than for 72 h exposure at 1000°C, and overall contained 9% impurities, but the bright spots had 15-23 % Fe. The composition through the cross-section was relatively constant and similar to that of the as-received, Table 6.2. Except for the area adjacent to the silica layer, which contained more of the silica-poor Mo5Si3 phase, which is the bright gray phase between the MoSi2 and SiO2 layers, lighter than the other two phases in the back-scattered electron (BSE) micrograph in shown in Figure 6.10 and Figure 6.11. This was a thin layer of molybdenum silicide with a lower silicon content (i.e. Mo5Si3 ),  that was continuous at some regions and remained isolated at other regions. Table 6.2 Elemental composition of KS-1700 through the cross section after 144 hrs exposed to 1600°C (all data in wt%)  O Al Si Mo center 16.51 0.85 36.38 46.26 2 mm from the surface 17.94 0.88 35.81 45.37 1 mm from the surface 20.25 0.43 36.00 43.31  The silica layer of exposed KS-1800 was ~32 µm at 1600°C, 10 times thicker than for the as-received samples and about 8 times thicker than for 72 h exposure at 1000°C, similar what it was for KS-1700. The composition through the cross-section was relatively constant and similar to as-received, except for the area adjacent to the silica layer, which contained more of the silica-poor Mo5Si3 phase, the bright gray phase shown in Figure 6.10 and Figure 6.11. Spallation of the SiO2 scale did not occur at any temperature, except at 1600°C for both MoSi2 based materials, KS-1700 and KS-1800. It was observed that spallation occurred around the circumference but the Mo5Si3 layer was continuous, Figure 6.10f and Figure 6.11f. Since a cohesive layer of Mo5Si3 was 57 observed underneath the oxide silica layer, the possibility of a rate controlling mechanism by diffusion of Si through the Mo5Si3 layer must be considered. No sign of bubbling of the oxide silica scale was detected; therefore, MoO3 did not form beneath the scale under these experimental conditions.        Figure 6.10 SEM section of the scale development formed on KS-1700 at temperature ranges 1000-1600°C  The scale of KS-1800 at 1600°C spalled-off on more than 50% of circumference too. The initial columnar morphology was observed at 1600°C. KS-1700 showed the same Si:Mo =~0.85 ratio at all (d) 1400°C-72h (a) 1000°C-72h (b)1200°C-72h (c) 1300°C-144h (e) 1500°C-144h (f) 1600°C-144h Clay Silica Mo5Si3 MoSi2 Fe Spallation 58 temperatures (except 1600°C) in the centre and at ½ radius; the ratio was slightly lower =~0.8, difference increasing with temperature, in the region at 50 µm from the interface with the grown SiO2 deposit. At 1600°C, the ratio was slightly lower =~0.8 throughout the section, as the scale was thicker and some of it has even spalled off Figure 6.11f. KS-1800 had similar values, except for 1000°C, where no scale formed, so the composition was uniform through the section Figure 6.11.        Figure 6.11 SEM section of the scale development formed on KS-1800 at temperature ranges 1000-1600°C  (a)1000°C-72h (b) 1200°C-72h (c) 1300°C-144h (d) 1400°C-72h (e) 1500°C-144h (f) 1600°C-144h Spallation 59 6.3 Overview of oxidation of MoSi2 based materials (300-1600°C) The mass variation after 72 hrs of exposure from 300°C to 1600°C, every 100°C, is presented in Figure 6.12a,b for KS-1700 and KS-1800. Below 900°C the products of the low temperature oxidation were MoO3 and SiO2. In the intermediate temperature ranges, above 795°C, the evaporation and the formation of MoO3 happened simultaneously while SiO2 remained on the surface.        Figure 6.12 Mass variations after 72 h isothermal exposure versus temperature of (a) KS-1700 and (b) KS-1800  300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600-101234567Temperature CMass gain (mg.cm-2)Isothermal Oxidation of KS-1800 (300:100:1600C for 72h)  low temperature oxidation (LTO)Intermediate temperature oxidation (ITO)high temperature oxidation (HTO)MoO3+SiO2MoO3+SiO2 (-MoO3)Mo5Si3+SiO260 Above 1000°C, SiO2 formed on the surface and poor Mo5Si3 phase underneath the surface. KS-1800 shows a very high resistance against oxidation among other types (those which formed silica on the surface) in temperature range 300-900°C. The peak temperature of pest oxidation for KS-1700 occurred at 700°C while it occurred at 400°C for KS-1800 for similar experimental conditions. This is attributed to different minor additives in KS-1700 and KS-1800.   61  7   Oxidation behavior of (Mo, W)Si2 based materials Chapter 7 Oxidation behavior of (Mo, W)Si2 based materials This part includes an investigation into oxidation behaviors of two (Mo, W)Si2 based materials namely Kanthal Super 1900 (KS-1900) and Kanthal Super HT (KS-HT). The major phase of the materials studied here is (Mo,W)Si2 with islands of SiO2 and (Mo, W)5Si3. 7.1 Low temperature oxidation (LTO) of (Mo, W)Si2 based materials (300-900°C) The oxidation kinetics of (Mo, W)Si2 based materials with boride addition was examined by isothermal experiments over a range of temperatures from 300 to 1600°C. Samples were removed periodically from the furnace and their weight gain measured.  The formation of molybdenum oxides in low temperature experiments is expected to lead to rupture of the oxide film owing to the significant volume change; this  accelerates the rate of oxidation, leading to the pest oxidation. Figure 7.1 shows the mass variation versus time for KS-1900 and KS-HT. The isothermal exposure temperature was from 300°C to 900°C for every 100°C and the time exposure was from 4 to 240 hrs (i.e. up to 10 days for each temperature) in these low temperature oxidation studies. Each point represents average of six measurements of two samples and the bars indicate one standard deviation. Sample preparation and heat treatment schedule here was the same as in the previous experiments. The plots of KS-1900 and KS-HT after 240 hrs exposure at 300°C were remained unchanged and the surface of the samples unaffected. The signs of oxidation started at the temperature of 400°C for both (Mo, W)Si2 materials tested here. The small mass variations of KS-1900 and KS-HT at 300 and 400°C are attributed to slow oxidation, vaporization of the oxidation products being still insignificant. The slow oxidation rate at 400°C results in a thin scale with a microstructure that depends on substrate composition. 62 It is reported that the Mo:W ratio in the oxide formed on the three intermetallic phases, i.e. (Mo, W)Si2, (Mo, W)5Si3, (Mo, W)B, is the same as that of unexposed materials and it is attributed to a simultaneous oxidation of Mo, Wand Si [83]. A higher rate of oxidation of KS-1900 and KS-HT was observed at 500°C and the highest rate of oxidation experienced at 600˚C among all the samples at 300-900˚C, as also shown in Figure 7.1. The steep and roughly linear mass gain curve recorded at 500-600˚C indicates that the oxide scale is non-protective. At higher than peak oxidation temperature, the mass gain drop off. A continuous linear increase in mass gain of KS-1900 and KS-HT was observed at 700°C, but in a lower rate than that of 600°C. The mass gain of KS-1900 and KS-HT at 400-800˚C is attributed to the formation of solid oxide of SiO2 and mass loss due to evaporation of MoO3 and WO3. The mass gain increased linearly at all temperatures except at 800 and 900°C, similarly a continuous mass loss during thermal aging was observed. This is because, in the intermediate temperature range, 800-1000°C, the evaporation and the formation of MoO3 occurred simultaneously, while SiO2 and WO3 remained on the surface. The negative mass variations recorded at 800 and 900°C is attributed to a combination of formation of MoO3, WO3, and SiO2 and evaporation of MoO3. The oxidation behavior of WSi2 is similar to that of MoSi2. They both display accelerated oxidation in the intermediate temperature range of 400–600°C and develop a semi-protective scale of SiO2 at elevated  temperatures at and above 900°C [9]. The simultaneous formation of MoO3, WO3 and SiO2 of KS-1900 and KS-HT was result of more extensive oxidation in LTO as compared to KS-1700 and KS-1800. It has shown that SiO2 is the most stable oxide among them. However, a mixed of MoO3 and SiO2 is formed at temperatures <550°C. At higher temperatures, the MoO3 is volatilized in the form of the (MoO3)n species [84].  63   Figure 7.1 Mass varations versus time of KS-1900 (left) and KS-HT (right) (300-900°C)  The oxidation reactions of (Mo,W)Si2 based materials within the temperature range at which accelerated oxidation takes place is described by the following reaction [84]:  2 (𝑀𝑜,𝑊)𝑆𝑖2 + 7𝑂2  → 2 (𝑀𝑜,𝑊)5𝑆𝑖3 + 4 𝑆𝑖𝑂2 (4)  Figure 7.2c shows the cross-section of the exposed sample of KS-HT after 120 hrs at 500°C.  The elemental analyses on the cross-section after 120 hrs at 500˚C  showed similar compositions for the center and mid-section, and richer in O compositions adjacent to the 80-95 µm thick silica scale, which also contains Mo-W oxides (seen as the light-contrasted  inclusions). There was more O throughout the section 64 (37 wt%) compared to the unexposed specimen (9 wt%), indicating formation of MoO3 and WO3 in the bulk, not only on the surface. After 240 h at 500˚C, Figure 7.2, except for the region adjacent to the silica deposit, which contains the silica-poor (Mo, W)5O3, the composition throughout the section was similar to that of the unexposed specimen.     Figure 7.2 The surface section and microstructure of KS-1900 and KS-HT at 400 and 500°C  The surface of the deposit after 240 hrs at 500°C in KS-HT is presented in Figure 7.4a: The whole surface was covered by oxidation products; the acicular particles are MoO3, silica (dark contrasted), MoO3 and WO3  (bright contrasted). There is also massive cracking of this deposit. The needle shaped microstructure on the surface of the samples identified by SEM/EXD indicates the existence of MoO3 on the surface while the majority of oxidation reactions formed on the surface were WO3 which was identified (a) KS-1900-400°C-240h (b) KS-HT-400°C-240h (d) KS-HT-500°C-240h (c) KS-HT- 500°C-120h 65 by SEM and XRD (Figure 7.4a,b). It is suggested that the vaporization of crystalline MoO3 in the oxide scale makes the oxide scale porous and non-protective, explaining the rapid oxidation [83]. After 240 h at 700°C, the surface of the KS-1900 (Figure 7.3a) had a composition of 41 wt% O, 16 wt% Si, 5 wt% Mo, 38 wt% W, with higher W and lower Mo concentrations compared to the sample aged at 500°C. Both the light agglomerations of small particles and the longer, better formed crystals, are mostly WO3; silica is gray with fine particles of tungsten oxide. This indicates that at higher temperatures, the oxidation of WSi2 is predominant. After 240 h at 700°C, the surface of the KS-HT (Figure 7.3b), shows numerous cracks and large acicular crystals of MoO3, formation of WO3, and small crystals of oxides very close to the surface of silica; the overall composition is 43 wt% O, 16 wt% Si, 15 wt% Mo, 26 wt% W.      Figure 7.3a,b SEM backscattered surface of KS-1900 and KS-HT exposed to 700°C for 240 h  KS-HT 700°C KS-1900 700°C WO3 SiO2 66   Figure 7.4a,b SEM backscattered electron image and X-rays diffraction pattern of the surface of KS-HT exposed to 500°C after 240 hrs (b) KS-HT exposed to 500°C after 240h (a)KS-HT- exposed to 500°C after 240h Intensity 67  Figure 7.3 (a, b) is showing SEM backscattered electron images of  KS-1900 and KS-HT exposed to 700°C for 240 hrs with two different characteristic features: the dark, continuous SiO2 and the bright WO3 crystals (containing small amount of molybdenum). The surface microstructure of KS-HT at 700°C (Figure 7.3) includes silica and mostly WO3, with large inclusions of B2O3; this a low-temperature melting glass (450˚C, which is why it is used as a fluxing agent).  Boron oxide formation is an undesired effect of using B to increase the oxidation resistance by filling out the pores and cracks in the silica scale. The matrix of KS-1900 and KS-HT consists of a solid solution MoSi2-WSi2; this formulation is intended to improve mechanical properties and high temperature oxidation resistance of these materials [8, 18, 24]. KS-HT also contains borides of Mo and W (to improve (Mo, W)5Si3 oxidation resistance), but boron content cannot be quantified (as it is very light element), so it does not appear in the overall elemental composition. The surface microstructure in Figure 7.5 shows region where MoO3 evaporated from the surface of KS-1900 exposed to 800-900°C for 240 hrs. It shows the surface of samples contain two regions: the smooth continuous SiO2 region that appears dark and WO3 crystals that appears bright on SEM images of Figure 7.5. It is reported that the (Mo, W)O3 is replaced by WO3 in the scale, the WO3 crystallites growing with increasing temperature [83]. The lack of molybdenum in the scale in the temperature range 800-900°C is attributed to vaporization of MoO3. The surface discoloration of the KS-1900 and KS-HT samples heat treated for 240 h at 400-900°C is presented in Figure 7.6. The formation of different low temperature oxidation products can be seen even with the naked eye. This kind of oxidation can lead in time to the disintegration of the material.  68   Figure 7.5 The surface microstructure of KS-1900 exposed to 800-900°C for 240 h    Figure 7.6 The surface discoloration of the specimens after 240h at different temperatures  The greenish/yellowish samples signifies the existence of mostly WO3 on the surface and it formed during thermal aging in the temperature range 300-900°C, as it can be seen even with the naked eye in Figure 7.6. The dark and dull color of the specimens indicates the presence of MoSi2 on the surface; the KS-1900-800°C KS-1900-900°C 69 shiny dark indicates amorphous silica; and WO3 and/or MoO3 imparts a yellow/green color.  The whitish color is for a mix of MoO3 and SiO2. 7.2 High temperature oxidation (HTO) of (Mo, W)Si2 based materials (1000-1600°C) This  section reports on  an investigation into high temperature oxidation behaviors of two (Mo, W)Si2 based materials namely Kanthal Super 1900 (KS-1900) and Kanthal Super HT (KS-HT). Figure 7.7 and Figure 7.8 show the mass variations versus time for KS-1900 and KS-HT, respectively. The isothermal exposure temperature was from 1000°C to 1600°C for every 100°C and the time exposure was from 2 to 72 hrs (5 days for each temperature) in these high temperature oxidation studies. Each point represents average of six measurements of two samples and the bars indicates standard deviation. Sample preparation and heat treatment schedule here was the same as previous experiment. The plots of specific weight gain versus time for isothermal oxidation of both KS-1900 and KS-HT at 1000, 1200, 1300, 1400, 1500, and 1600°C, are shown in Figure 7.7 and Figure 7.8.   Rapid weight gain is seen for the first two hours of exposure followed by a slower rate of oxidation, noticeably slower than that of KS-1700 and KS-1800. Clearly, there was an insignificant change for KS-1900 and KS-HT in the rate of oxidation with increasing temperature and time exposure from 1000 to 1600°C, while the rate of oxidation changed significantly for  KS-1700 and KS-1800 with increasing temperature and time. Except at 1600°C,  a relatively large mass variations and oxide scale thicknesses were observed in KS-1900. During the initial part of the high temperature oxidation, when a protective oxide layer had not yet formed, oxygen was readily available to react with silicon. In the first two hours, it appears that linear oxidation was happening (at a higher rate) and then the linear oxidation regime transforms into a parabolic regime (at a slower oxidation rate) starts when an oxide layer covers surrounding the sample referring to Figure 7.7 and Figure 7.8. In this period of exposure time, Si and O   must diffuse through the bulk material and the initial oxide layer, respectively, to arrive at the SiO2–(Mo, W)Si2 interface for oxidation reaction to take place.  70 The oxidation behavior of WSi2 is similar to that of MoSi2. They both display accelerated oxidation in the intermediate temperature range of 400–600°C and develop a semi-protective scale of SiO2 at elevated temperatures at and above 900°C [85]. The simultaneous formation of MoO3, WO3 and SiO2 of KS-1900 and KS-HT was a result of extensive oxidation in LTO as compared to KS-1700 and KS-1800. However, dry oxidation at 1000 and 1200°C has shown evidence of simultaneous formation  of WO3 and SiO2, which is in contrast to minor mass gain observed in the case of MoSi2 under similar conditions of oxidation [9].   Figure 7.7 Mass variations versus time of KS-1900 in the temperature range 1000-1600°C 71  Figure 7.8 Mass variations versus time of KS-HT in the temperature range 1000-1600°C  At higher temperatures, (above 1000°C) the oxidation reaction is described by following reaction [84]:  10 (𝑀𝑜,𝑊)𝑆𝑖2 + 6𝑂2  → 2 (𝑀𝑜,𝑊)5𝑆𝑖3 + 6 𝑆𝑖𝑂2 (4)  It was observed, see Figure 7.7 and Figure 7.8, that oxidation kinetics of (Mo, W)Si2 based materials (KS-1900 and KS-HT ) involved  loss in mass with time of exposure at all temperatures studied here (except KS-1900 at 1500 and 1600°C). The mass loss is attributed to volatilization of WO3 (which is in contrast to minor mass gain observed in the case of MoSi2 under similar conditions of oxidation) during oxidation at 1200 and 1500°C in air. Formation of W5Si3 by reaction of WSi2 phase with oxygen (in solid solution alloy of MoSi2 and WSi2 of KS-1900 and KS-HT) has also been confirmed. The oxidation resistance of W5Si3 has  been found to be much poorer, characterized by a large  loss in mass at 1200°C [9]. Micrographs of the KS-1900 sample after 72 and or 144 hrs isothermal exposure in the temperature range 1000-1600°C are shown in Figure 7.9.  72       Figure 7.9 SEM section of the scale development formed on KS-1900 at temperature ranges 1000-1600°C  The silica layer was ~45 µm at 1600°C, 10 times thicker than for the as-received samples and ~20 times than for 72 h exposure at 1000°C. The composition through the cross-section was relatively constant and similar to as-received, except for the area adjacent to the silica layer, which contained more of the silica-poor Mo5Si3 phase, the bright gray phase shown in Figure 7.9. After the high temperature oxidation, KS-1900 (Figure 7.9) had slightly higher Si:(Mo+W) ratios =~0.5 at 1000°C and 1200°C than at the higher temperatures; this is probably due to evaporation of WO3, which starts to form just above 800°C and starts evaporating at over 900°C [86]. The scale at 1300°C was (a) 1000°C-72h (b) 1200°C-72h (c) 1300°C-144h (d) 1400°C-72h (e) 1500°C-144h (f) 1600°C-144h Mo5Si3+W5Si3 Silica 73 very irregular (this image in Figure 7.9c is a good example): the minimum value was 4 µm; many regions had 8-10 µm, and a few places had 20-25 µm. Also, specific to this temperature, regions of SiO2 (dark-contrasted) with inclusions of small (tenths of microns) bright particles (incipient Mo5Si3+W5Si3) started to appear at 0.6-0.7 mm from the scale interface, and closer to it developed into larger particles. Micrographs of the KS-HT sample after 72 and or 144h isothermal exposure in the temperature range 1000-1600°C are shown in Figure 7.10. KS-HT is designed for the application of thermal cycling and also use for making micro heating elements. Therefore the oxidation resistance of KS-HT is improved at higher temperatures (above 1200°C). The silica layer was only 18 µm at 1600°C, which was considerably thinner than that of KS-1900. The composition through the cross-section was relatively constant and similar to as-received, except for the area adjacent to the silica layer, which contained more of the silica-poor (Mo, W)5Si3 phase, the bright gray phase shown in Figure 7.10. For KS-HT at 1000°C, the ratio Si:(Mo+W) =~0.5 was dramatically higher than at the other temperatures; the EDX analysis detected only minor amounts of W (0-3 wt%, instead of 28-29 wt% in the rest of the samples; Figure 7.10 also shows practically none of the bright-contrasted particles seen at the other temperatures. This indicates that at 1000°C, both tungsten silicides were oxidized to WO3, which had already evaporated. The surface microstructure of KS 1900 and KS-HT bars after 144h at 1300°C and 1500°C are shown in Figure 7.11. 1000°C-72h 1200°C-144h 1300°C-72h 1500°C-144h 1600°C-144h  Figure 7.10 SEM section of the scale development formed on KS-HT at temperature ranges 1000-1600°C 74    Figure 7.11 The surface appearance of KS 1900/HT bars after 144h at 1300 & 1500°C     Figure 7.12 The microstructure section of KS-HT a) 1200°C after 72h b) 1500°C after 144h   As shown in Figure 7.12, after exposure different temperature in high temperature oxidation experiments, more pocket of silica observed near the surface, but less silica far from the surface. 7.3 Overview of oxidation of (Mo, W)Si2 based materials (300-1600°C) The mass variation after 72 hrs of exposure from 300 to 1600°C, every 100°C, is presented in Figure 7.13 for (a) KS-1900 and (b) KS-HT. In this investigation, it has been shown that the oxidation of KS-1900 and KS-HT, (Mo, W)Si2 based materials , is similar to that of KS-1700 and KS-1800, MoSi2 based (a) KS-1900 1300°C-144h (b) KS-1900 1500°C-144h (c) KS-HT 1500°C-144h Less silica More silica near the surface (d) KS-HT 1200°C-72h (e) KS-HT 1500°C-144h More silica Less silica 75 materials. However, the peak oxidation temperature of (Mo,W)Si2 material was shifted to 600°C and the low temperature oxidation was extensive. Below 900°C the products of the low temperature oxidation were MoO3, WO3, and SiO2. In the intermediate temperature ranges, above 795°C, the evaporation and the formation of MoO3 happened simultaneously while SiO2 and WO3 remained on the surface. In KS-1900, and KS-HT, about a third of the Mo content is substituted by W. It is known that the oxidation behavior of this compound varies depending on the amount of W [87]. Oxidation behavior also varies according to the oxidation temperature and time exposure.     Figure 7.13 Mass variations after 72 h isothermal exposure versus temperature of (a) KS-1900 and (b) KS-HT  300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600-101234567Temperature CMass gain (mg.cm-2)Isothermal Oxidation of KS-1900 (300:100:1600C for 72h)  low temperature oxidation (LTO)Intermediate temperature oxidation (ITO)high temperature oxidation (HTO)MoO3+SiO2+WO3SiO2 +/-- (WO3 & MoO3) Mo5(Si,W)3+SiO2+W5Si3300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600-101234567Temperature CMass gain (mg.cm-2)Isothermal Oxidation of KS-HT (300:100:1600C for 72h)  low temperature oxidation (LTO)Intermediate temperature oxidation (ITO)high temperature oxidation (HTO)MoO3+SiO2+WO3SiO2 +/- (WO3 & MoO3)Mo5(Si,W)3+SiO2+W5Si3 (-MoO3)76 Above 1000°C, SiO2 formed on the surface and poor Mo5Si3+W5Si3 phase underneath the surface. (Mo, W)Si2 based materials were more susceptible to LTO (300-900°C) as compared to MoSi2 and Mo(Si, Al)2 based materials. But they showed a better oxidation resistance in higher temperatures (above 1000°C)  as compared to MoSi2 and Mo (Si, Al)2 based materials.    77  8   Oxidation behavior of Mo(Al, Si)2 based materials Chapter 8 Oxidation behavior of Mo(Al, Si)2 based materials In the formulation of Kanthal Super ER, Mo(Si, Al)2, part of Si in MoSi2 is replaced by aluminum, so the main compounds are molybdenum disilicide and molybdenum aluminate; 15% alumina is also added, as well as <5% Mo5(Si, Al)3. According to the manufacturer's specifications, at high temperatures of 1000 to 1580°C, this material should form an alumina scale on the surface, which is much more resistant to reducing conditions and water vapors than the silica scale formed on the other materials studied here [10]. In addition, this composite is much more resistant to the active oxidation at lower temperatures than the silica-forming materials [10, 75, 76]. This part includes an investigation into oxidation behavior of Mo(Si, Al)2-based materials namely Kanthal Super ER (KS-ER). The phase of the matrix of the material studied here is Mo(Si, Al)2 with islands of Mo5(Si, Al)3 phase and Al2O3 phase.  8.1 Low temperature oxidation (LTO) of Mo(Al, Si)2 based materials (300-900°C) The oxidation kinetics of MoSi2-based materials have been determined by isothermal oxidation experiments over a range of temperatures from 300 to 1600°C. Figure 8.1 shows the mass variation versus time for KS-ER. The isothermal exposure temperature was from 300 to 900°C for every 100°C and the time exposure was from 4 to 240 hrs (i.e. up to 10 days for each temperature). Sample preparation and heat treatment schedule here was the same as in the previous experiments. Experimental results have indicated that KS-ER also exhibits pest disintegration after oxidation at low temperatures, but the rate of oxidation is insignificant in compared to other types of MoSi2 based materials. At 400°C  KS-ER which starts  to show  early signs of oxidation , as shown in Figure 8.2. The surface of the exposed samples contains MoO3 and 78 SiO2 in different proportions, and in some areas pockets of impure silica can be seen, Figure 8.2. Similar to MoSi2 ( refer to  Chapter 2), the maximum low temperature oxidation of Mo(Al, Si)2 occurred at 500°C.   Figure 8.1 Mass variations versus time of KS-ER (300-900°C)    Figure 8.2 SEM surface of KS-ER exposed to 300-400°C for 240 h (a) The major grey region is Mo(Al,Si)2 (I), the blac regions are Al2O3 (II) and the white regions are Mo5(Al,Si)3 (III). (b) Elongated MoO3 crystals formed on the surface (I)  The EDX analysis of the silicide regions reported by Ingemarsson et al., [76] a Mo:Si:Al ratio of 1:1:1 for Mo(Si,Al)2 and 10:5:1 for Mo5(Si,Al)3, approximately. The surface and cross-sections of KS-ER after (a) KS-ER 300°C-240h (b) KS-ER 400°C-240h I II III I 79 240 hrs exposure at 500˚C is presented in Figure 8.3a,b.  The as-received samples had the overall surface deposit composition of w 53 wt% O, 34 wt% Al, 1 wt % Si, 1 wt% Mo, so it was basically only alumina. After 240 h at 500˚C, the surface composition was 45 wt% O, 6 wt% Al, 12 wt % Si, 37 wt% Mo; this indicates massive oxidation of the superficial MoSi2, which now was not protected anymore by alumina. However, this happened only on the surface, as the composition in different positions (even close to the MoO3 layer) was constant, and similar to that of the new material (17 wt% O, 18 wt% Al, 19 wt % Si, 47 wt% Mo). The EDX analysis of the silicide regions of unexposed samples of KS-ER showed a Mo:Si:Al ratio of approximately 1:1:1 for Mo(Si, Al)2 and 0:5:1 for Mo5(Si ,Al)3. The surface deposit after the exposure at 600-700°C is shown in Figure 8.3c,d; the MoO3 does not form readily at this temperature, as it does at 500°C, so the surface consists of silica with variable concentrations of Mo(Si, Al)2.     Figure 8.3 (a, b) The SEM surface and section of KS-ER exposed to 500°C and (c, d) the SEM surface of KS-ER exposed to 600, 700°C for 240 h (a) KS-ER-500°C (b) KS-ER-500°C (c) KS-ER-600°C (d) KS-ER-700°C 80 The surface discoloration of the samples heat treated for 240 hrs at 400-900°C is presented in Figure 8.4. KS-ER had a dull black surface with white oxidation products. White spots were observed on the free surface after thermal aging at all different temperatures used in this experiment. Except a little (~0.5 mg. cm-2) mass gain at 500˚C after 240 hrs, there was no significant mass changes during thermal aging of the KS-ER samples.  Figure 8.4 The surface discoloration of the specimens after 240h at different temperatures  8.2 High temperature oxidation (HTO) of Mo(Al, Si)2 based materials (1000-1600°C) At temperatures higher than 1000°C, compact scales of alumina and silica provide the best resistance to oxygen diffusion because of their high thermodynamic stability and low diffusivities for both cations  and anions [6, 9]. For high temperature applications, it is desirable to have a continuous and coherent scale of SiO2 or Al2O3, the formation of which often comparable with the oxides of other alloying elements or composites [9]. Alumina is  an excellent barrier to oxygen at temperatures below  1300 °C, but at higher temperatures oxygen permeation  through silica occurs at a lower rate [6]. High temperature oxidation product of four types of materials studied here (KS-1700, KS-1800, KS-1900, and KS-HT) is silica and the scale forms at high temperature on the surface of KS-ER is alumina. In the formulation of Kanthal Super ER, Mo(Si, Al)2, part of Si in MoSi2 is replaced by aluminum, so the main compounds are molybdenum disilicide and molybdenum aluminate; 15% alumina is also added, 81 as well as <5wt% Mo5(Si, Al)3. At high temperature, this material forms an alumina scale on the surface, which is much more resistant to reducing conditions and water vapor than the silica scale formed on the other materials studied here. In addition, this composite is much more resistant to the active oxidation at relatively low temperatures (< 1000°C) than the silica-forming materials. The mass variation after 72 hrs of exposure from 1000 to 1600°C, every 100°C, is presented in Figure 8.5 for KS-ER. The scale that forms on KS-ER at high temperature is alumina, as in this material part of Mo is replaced by Al; the basic formula becomes Mo(Si, Al)2. The plots of specific weight gain versus time for isothermal oxidation of both KS-ER at 1000, 1200, 1300, 1400, 1500, and 1600°C started with a rapid weight gain followed by a slower rate of oxidation. During the initial part of high temperature oxidation, when a protective oxide layer had not yet formed, oxygen was readily available to react with alumina, and there was a change in the rate of oxidation with increasing temperature and time exposure from 1000 to 1500°C. The rate of oxidation at 1500°C was higher than that of other temperature studies here (except 1600°C). In the first two hours, it is assumed the linear oxidation was happening (higher rate) and then the linear oxidation regime ends and a parabolic regime (slower rate) starts when an oxide layer covers surrounding the sample. In this period of exposure time, Al and O2 must diffuse through the bulk material and the initial oxide layer, respectively, to arrive at the Al2O3–Mo (Si, Al)2 interface for oxidation reaction to take place. The alumina scale imparts this material a high resistance to alternating oxidizing and reducing atmospheres.  In our experiments, KS-ER showed excellent resistance against the low temperature oxidation by formation of a 5 to 100 µm thick and about 100 % dense alumina scale on the surface of the samples at high temperatures, above 1200°C (Figure 8.6). The rate of alumina formation on KS-ER was relatively higher than the glass film formation on the other types of composite MoSi2 materials. Alumina scales, shown in the Figure 8.6, exhibit considerable thickness variations above 1300°C.  82  Figure 8.5 Mass variations versus time of KS-ER in the temperature range 1000-1600°C  KS-ER had the highest mass variation (Figure 8.5) and highest thickness scale variation (Figure 8.6), as the alumina scale was much thicker (4 to 100 µm) at every temperature than the SiO2 formed on the other formulations in the same conditions (4 to 50 µm). Also, the “depletion layer” from which alumina grows, had the same thickness of 20 to 50 µm at 1400°C and 1500°C as the alumina layer. Our tests consisted in isothermal oxidation of KS-ER bars at 1500°C in air, for different durations, and tracking the deposit thickness and the Al variation through the cross-sections. The alumina forms from an adjacent later, which becomes depleted in Al, and in turn, slightly reduces the Al next to it. Figure 8.6 shows the thickness of the scale and Al-depletion layer, and the Al variation through the section. Ingemarsson et al. [88] reported a thickness of 32 µm after 72h for KS-ER in similar conditions in 2013.  At 1600°C, which is above the maximum temperature specified for KS-ER, the alumina layer was over 160 µm thick, and the bulk showed massive material loss during the preparation for SEM, probably due to some loss of the binding phase, see Figure 8.5, and Figure 8.7. The spallation at 1600°C may be attributed to the CTE mismatch between thick scale alumina (160 µm) and substrate (Mo (Si, Al)2  based materials). Massive loss at 1600°C 83     Figure 8.6a,b,c,d,e the section and the scale appearance formed on KS-ER at temperature ranges 1000-1500°C  A recent work conducted by Hellstrom et al. [89] shown that after 1000 h in the furnace at 1580°C, the thickness of the alumina layer was ~110 µm, but it was just 120 µm for the same type of element, which was heated at the same temperature by alternating current for 1 year (8760 hrs). The authors attribute the thinner Al2O3 scale under AC to the slower Al diffusion caused by different thermal gradients than in the case of furnace exposure. The initial phases in the bulk of the element (Mo5(Si, Al)3 and Al2O3) agglomerate differently after longer durations: randomly in the furnace exposure and in a circular shape after AC exposure, which indicates that the current influences the structural modifications in a manner not yet understood. After one year, the alumina scale grown under AC was still compact, continuous and protective. (a) 1000°C-72h (b) 1200°C-72h (c) 1300°C-144h (d) 1400°C-72h (e) 1500°C-144h Al2O3 Mo5(Si, Al)3+ Al2O3  Mo(Si, Al)2+ Al2O3+ Mo5(Si, Al)3  84       Figure 8.7 The microstructure section KS-ER exposed to 1600°C  The unexposed KS-ER material had a 2 µm alumina layer formed on the surface, and the same thickness Al-depleted layer adjacent to it. After 12 h at 1500˚C, the alumina layer was 15-35 µm, so not uniform, but quite compact. The Al-depleted layer of (Mo, Al)Si2 contained only 4% Al; the region next to it had 10% Al, compared to the 18 wt% Al in the rest of the section, which had the same composition as the unexposed KS-ER (17 wt% O, 18 wt% Al, 19 wt% Si, 47 wt% Mo). The KS-ER surface composition corresponds to alumina with some MoSi2, which indicates that the alumina layer is thin (confirmed by the image of the cross-section). The EDX after 12 h at 1500°C showed only alumina. After 72 h exposure at 1500˚C, the EDX analysis showed that the oxidation layer was alumina only, and its thickness, as well as of the adjacent the Al-depleted layer, was 30-38 µm. These results were confirmed by published works (15, 16), which showed 35 µm after 72 h at 1500°C). Away from the surface, the composition was constant through the cross-section, and similar to as-received samples. Even after exposures longer than 144 h, the kinetics of the Al2O3 formation remained parabolic.   1600°C-144h 1600°C-144h Al2O3 oxide scale 85 8.3 Overview of oxidation of Mo(Al, Si)2 based materials (300-1600°C) The mass variation after 72 hrs of exposure from 300 to 1600°C, every 100°C, is presented in Figure 8.8 for KS-ER. In the formulation of Kanthal Super ER, Mo(Si, Al)2, part of Si in MoSi2 is replaced by aluminum, so the main compounds are molybdenum disilicide and molybdenum aluminate; 15% alumina is also added, as well as <5% Mo5(Si, Al)3. At high temperatures, this material forms an alumina scale on the surface, which is much more resistant to reducing conditions and water than the silica scale formed on the other materials studied here. At high temperatures, this material is covered by a continuous alumina film. The substrate beneath the scale is enriched in Mo5(Si, Al)3. This is attributed to aluminum depletion of the Mo(Si, Al)2 matrix due to alumina formation [76]. Mo(Si, Al)2 is more resistant to the active oxidation at low temperatures than the silica-forming materials. Below 900°C the products of the low temperature oxidation were MoO3 and SiO2. KS-ER showed the highest resistant against pest oxidation in temperature range of 300-900°C. In the intermediate temperature ranges, above 795°C, the evaporation and the formation of MoO3 happened simultaneously while SiO2 remained on the surface.  In contrast to the clay bonded MoSi2 based material, the Mo(Si,Al)2 based materials exhibits insignificant MoO3 volatilization. Above 1000°C, Al2O3, and Mo5(Si, Al)3 formed on the surface and poor Mo5Si3 phase underneath the surface.  Figure 8.8 Mass variations after 72 h isothermal exposure versus temperature of Mo(Si, Al)2 (KS-ER)  86  9   Comparison of the oxidation of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials Chapter 9 Comparison of the oxidation of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials  9.1 A proposed failure criterion for MoSi2 based materials Since the chemical and physical modifications of the sample coupons under the given conditions were so diverse, in order to be able to compare and rank the materials studied here, a new failure criterion was proposed. This is a failure criterion signified by the accumulation of the degradation via summation of the absolute values of mass gain and mass loss at different temperatures. The consideration of the absolute values was dictated due to simultaneous mass gain due to the materials oxidation, and mass loss due to evaporation of some of the oxidation products. The bar chart in Figure 9.1(a) shows the final (total) accumulated mass gain or mass loss for all five types of Kanthal Super materials studied, after isothermal exposure for 240 hrs in the temperature range of 300-900°C (every 100°C), with the pre-oxidation layer removed. The materials are subsequently ranked in Figure 9.1(b). To be able to rank the materials based on the oxidation rate, the summation of the absolute values of mass gain and mass loss per unit area at each temperature was calculated in the temperature range of 300˚C-900˚C. As shown in the Figure 9.1 (a, b), Kanthal-1800 and Kanthal-ER showed an excellent resistance against low temperature oxidation, while Kanthal-HT and Kanthal-1900 were extensively degraded in low temperature oxidation. In summary, the role that the low temperature oxidation can play in the life-in-service of these materials is remarkable; if subjected long enough to the 300-900°C temperature range, they can suffer structural loss, and even disintegrate. 87   Figure 9.1 (a) Low temperature oxidation study-summary, (b) Summation of the absolute values of mass gain and mass loss per unit area, at different temperatures (300˚C-900˚C), for the materials studied in this work  9.2 Microstructural analysis of MoSi2 based materials after HTO (1000-1600°C) Alumina is  an excellent barrier to oxygen at temperatures below  1300 °C, but at higher temperatures oxygen permeation  through silica occurs at a lower rate [6]. High temperature oxidation product of four (a) (b) 88 types of materials studied here (KS-1700, KS-1800, KS-1900, and KS-HT) is silica and the scale forms at high temperature on the surface of KS-ER is alumina. This part includes microstructural analysis of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials in the temperature range 1000-1600°C. The variation of the major elements through the cross-sections was tracked in the center (A), at ½ radius (B), and at 50 µm from interface (C). The EDX analysis was performed on the whole field of these regions, at 1000x magnification, and the summary of the results are presented in Table 9.1. The lower values for  Si:Mo or Si(Mo+W) ratio mean that more of the low-Si silicides are present (i.e. in the form of Mo5Si3 and Mo5(Si, W)3);. However, these ratios could be slightly eschewed by the formation and evaporation of MoO3, even though precautions were taken to not allow that, by placing the samples in the already hot furnace. For KS-ER, the lower ratio Al:Mo is found closer to the surface, where an Al-poor Mo5(Si, Al)3 phase form, from which the Al2O3 scale grows.  Table 9.1 Composition variation through the cross-sections of samples in temperature range 1000-1600°C °C KS-1700 Si:Mo KS-1800 Si:Mo KS-1900 Si:(Mo+W) KS-HT Si:(Mo+W) KS-ER Al:Mo A B C A B C A B C A B C A B C 1000 0.80 0.80 0.79 0.80 0.80 0.80 0.52 0.52 0.51 0.70 0.70 0.73 0.42 0.39 0.38 1200 0.80 0.80 0.78 0.80 0.80 0.78 0.51 0.51 0.49 0.44 0.43 0.43 0.42 0.39 0.38 1300 0.80 0.80 0.77 0.80 0.80 0.77 0.48 0.48 0.47 0.44 0.43 0.42 0.40 0.39 0.35 1400 0.80 0.80 0.77 0.80 0.80 0.77 0.48 0.47 0.47 0.44 0.41 0.41 0.39 0.37 0.33 1500 0.80 0.80 0.77 0.80 0.79 0.77 0.48 0.48 0.46 0.43 0.40 0.40 0.37 0.36 0.31 1600 0.78 0.78 0.77 0.80 0.79 0.77 0.47 0.47 0.45 0.43 0.40 0.40 0.35 0.34 0.30  The morphology and the oxide scale thickness formed on the exposed samples of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials in the temperature range 1000-1600°C were studied in the previous sections of this study. The scales were measured on the SEM images in the same 12 positions around the circumference; the 1st value (µm) was measured in most positions; for the cases with very irregular scales, minimum and maximum values found in a few other positions are given in parenthesis. 89 9.3 Determining the activation energy of diffusion in MoSi2 based materials A special interest was paid to calculate the oxidation activation energies of  most materials studied here. Since oxidation is a thermally activated process, the oxide growth rate may be described by Arrhenius equation. With the result of this investigation, the activation energy through Arrhenius equation were determined following the procedure described in [90]. The Arrhenius equation, Eq. (1),  𝐾 = 𝐴𝑒_𝐸𝑎𝑅𝑇⁄  (1) Where Ea is the activation energy for oxidation, R is the gas constant, T is the absolute temperature, K is parabolic constant, and A is a constant. The Eq. (1) can be written in a non-exponential form, Eq. (2), that is more convenient to use and to interpret graphically. Taking the logarithms of both sides and separating the exponential and pre- exponential term yields.  ln 𝐾 = 𝐿𝑛 𝐴 −−𝐸𝑎𝑅𝑇 (2) Which is the equation of a straight line whose slope is −𝐸𝑎𝑅. This affords a simple way of determining the activation energy from values of parabolic constant 𝐾𝑝 observed at different temperatures, by plotting ln 𝐾𝑝 as a function of  1𝑇⁄  [90]. The kinetics of high-temperature oxidation of MoSi2 based materials may be studied by measuring either the increase in the thickness of the oxide layer or the weight gain as a result absorption of oxygen into the sample versus time at any given temperature. However, the parabolic rate constants may be calculated by either method. In this investigation, the parabolic rate constants were determined from mass variations versus time curves from the data measured in this study. Figure 9.2 shows the oxidation kinetics of the KS-1800 samples oxidized in air at 1000-1600°C and Figure 9.3 illustrates the square of the weight gain varies linearly with time indicating parabolic oxidation kinetics. The plots of specific weight gain versus time of all materials studied here should be parabolic because the oxidation mechanism is diffusion-90 controlled. The parabolic oxidation rate, Kp, may be calculated from the slope of the line obtained by plotting weight gain square (mass gain)2 versus time by 2 (time), as shown in Figure 9.3. The parabolic growth rate of the oxide scales is based on Wagner’s oxidation theory [91]. Based on this theory, the oxide scale is assumed compact perfectly adherent scale and the migration of ions or electrons across the oxide scale is the rate-controlling process. Parabolic rate constants (Kp) were determined at each temperature by obtaining the slopes of plots in Figure 9.3. Similar procedure has been done for Kanthal Super 1700, and Kanthal Super ER. Based on Eq. (2) with using the calculated parabolic rate constants (Kp), the activation energies were determined from the slope of logarithm of parabolic rate constant, ln (Kp), versus inverse absolute temperature,  1 𝑇⁄ , as it can be seen in Figure 9.4, and Table 9.2.   Figure 9.2 Plots of specific weight gain versus time of KS-1800 in the temperature range 1000-1600°C    91  Figure 9.3 Plots of specific weight gain square versus time by 2 at five test temperatures during oxidation of KS-1800   Figure 9.4 Plots of logarithm of parabolic rate constant versus inverse absolute temperature determining activation energy for oxidation of MoSi2-based materials studied here   The parabolic rate constants calculated from the slopes of the lines in Figure 9.3 are 0.001, 0.0063, 0.002, 0.0241 and 0.0827 (mg2.cm-4.0.5h-2) at 1200, 1300, 1400, 1500 and 1600°C, respectively. Such a 92 change in the parabolic rate constants has been reported, and it is believed to be a result of transformation from one silica phase to another silica phase [33, 92]. The activation energies of Kanthal Super 1700, Kanthal Super 1800, and Kanthal Super ER were calculated as 236, 305.5 and 313.1 in kJ.mol-1. For the Kanthal Super 1900, there was a significant change in behavior at 1400°C, so two values were calculated for the activation energy: 72.4 kJ.mol-1 at 1200-1400°C, and 752.4 kJ.mol-1 for the temperature range of 1400-1600°C. This is very close to the activation energy reported by Ingemarsson et al., [88] and Maruyama et al. [93] for a similar Mo(Si,Al)2 materials 310 and 307 kJ.mol-1, respectively. We attribute this to the very late formation, only around 1400°C, of the surface. SiO2 layer, whose growth was hindered by the evaporation of MoO3, along the formation of WO3, followed by its evaporation. The activation energy could not be calculated for KS-HT, because as explained above, it is formulated to form less silica than the other materials, so after the removal of the pre-oxidized film, it could not re-form a SiO2 protective layer properly.  Table 9.2 Thermal activation energies calculated by performing Arrhenius Plots Materials Temperature range Activation Energy (kJ.mol-1) KS-1700 (1200-1600°C) 236 KS-1800 (1200-1600°C) 305 KS-1900-1 (1200-1400°C) 72 KS-1900 -2 (1400-1600°C) 752 KS-ER (1200-1500°C) 313  9.4 The oxide scale thickness of MoSi2 based materials (1000-1600°C) The parabolic rate constants (Kp) calculated in the previous section were used to estimate the time to form a certain oxide thickness or the life expectancy of a part made from MoSi2 or its alloys at a given temperature in an oxidizing environment.   93 The predicted oxide scale thickness were calculated by Eq. (3)  Assuming same density, i.e. 𝜌1 = 𝜌2, therefore:  𝑚2𝑚1= 𝑉2𝑉1= ℎ2ℎ1  (3)  Where: m1 and m2 represent the mass values, V1 and V2 are the volumes, and h1 and h2 are the scale thickness. Form the previous section, the parabolic oxidation rate, Kp, was calculated, see Figure 9.3, from the slope of the line obtained by plotting weight gain square (mass gain)2 versus time by 2 (time). Therefore, we have:  𝐾𝑝 =(𝑚𝑎𝑠𝑠)22(𝑡𝑖𝑚𝑒)                          →        𝐾𝑝 =  𝑚222 × 𝑡2 (4)  Where: Kp represents the parabolic rate constant at each temperature, m2 is the mass value, t2 is the time. The calculations were performed based on the scale being 100% dense and consisting of pure silica or alumina. The oxide scale thickness in high temperature isothermal oxidation of MoSi2 based materials was measured experimentally by SEM and modeled using the values of mass gain at different time exposures over temperature ranges (1000-1600°C). By using Eq. (3) and (4), the predicted thicknesses at different exposure time were obtained. The Table 9.3 shows the oxide scale thickness (in µm), in both measured by SEM (indicated in red fonts) and modeled using the values of mass gain (indicated in black fonts). Each number represents average of measurements of the oxide thickness at six different locations per specimens and it was used the average values of mass gain for the prediction of oxide scale thicknesses.   94 Table 9.3 The oxide scale thickness (all in µm) in high temperature isothermal oxidation Temperature Materials 72 h 144 h 1000 h 10000 h 1000°C KS-1700 4 … … … KS-1800 … … … … KS-1900 3 … … … KS-ER 1 … … … KS-HT 4 … … … 1200°C KS-1700 4 … … … KS-1800 4 … … … KS-1900 4 … … … KS-ER 4 … … … KS-HT 4 … … … 1300°C KS-1700 … 12 12 13 KS-1800 … 10 10 12 KS-1900 … 8 8 8 KS-ER … 14 15 21 KS-HT … 4 … … 1400°C KS-1700 16 16 16 17 KS-1800 12 12 12 16 KS-1900 10 10 10 10 KS-ER 20 20 21 35 KS-HT 7 … … … 1500°C KS-1700 … 28 29 35 KS-1800  30 31 37 KS-1900 … 25 25 27 KS-ER … 50 53 80 KS-HT … 10 … … 1600°C KS-1700 … 34 … … KS-1800  50 51 64 KS-1900 … 45 46 57 KS-ER … … … … KS-HT … 18 … …   95 9.5 Comparative oxidation kinetics of MoSi2, (Mo, W)Si2 and Mo(Al, Si)2 based materials (300-1600°C) The mass variation after 72 hrs of exposure from 300 to 1600°C, every 100°C, is presented in Figure 9.5 for all material studied in this investigation. There is no unified published data on the oxidation of various MoSi2 based materials over such a broad range of use temperature. The scale of the materials were removed to hasten the onset of oxidation during the laboratory exposures, and to make sure that all materials start in the same condition. The peak oxidation temperature of the accelerated oxidation, below 1000°C, for each composition was obtained. This graph enables us to compare the oxidation resistance of MoSi2 based materials in a quantitative manner over a broad temperature range 300 to 1600°C. The plots of mass variation showed that the oxidation behavior greatly depends on the composition, and the exposure time and temperature.  (Mo, W)Si2 based materials (KS-1900, and KS-HT) were extensively degraded, while KS-1800 and KS-ER showed an excellent resistance against low temperature oxidation. The rate of alumina formation on KS-ER was relatively higher than the amorphous silica film formation on the other types of composite MoSi2 materials. KS-HT exhibited better high temperature oxidation resistance among all the materials studied here.  Figure 9.5 Mass variations after 72h isothermal exposure versus temperature of the materials studied in the temperature range 300-1600°C  96  10   Summary and conclusions Chapter 10 Summary and conclusions In the previous research carried out in our lab, we had found that there was extensive degradation of the Kanthal Super HT terminals during thermal cycling. The root cause analysis carried out on the failed samples indicated that the main cause of the failure is the deleterious oxidation of MoSi2 materials at temperatures ranging from 300-900°C. Considering the importance of this issue, an investigation was initiated for all available Kanthal Super materials (KS-1700, KS-1800, KS-1900, KS-HT, and KS-ER). Studies were performed on a broad range of temperatures, between 300°C and 1600°C. The experiments in this study were divided into two main categories: low temperature oxidation LTO (300 to 900°C; high oxidation rate expected) and high temperature oxidation HTO (1000 to 1600°C; lower oxidation rate expected due to rapid formation of the protective oxide films). The samples shapes selected for the investigation were close to the real applications, i.e. in the form of cylinders. The specimens were cut into 21 mm long and 6 mm in diameter using a high-speed diamond saw. The pre-oxidized layer of the samples was removed uniformly from the samples before the tests to hasten the onset of oxidation during the laboratory exposures, and to make sure that all materials start the tests in the same condition. Scanning Electron Microscopy (SEM), Energy-dispersive X-ray spectroscopy (EDS), and X-ray diffraction (XRD) analyzed the microstructure, chemical composition and phase composition of the oxidized samples. It was found that the oxidation behavior of the different materials under investigation depended strongly on their chemical and phase composition, exposure time and temperature. The mass variation (versus both time and temperature separately) of exposure from 300 to 1600°C was presented for all material studied in this investigation. Peak oxidation temperature of the accelerated oxidation, below 1000°C, for each composition were obtained. (Mo, W)Si2 based materials (KS-1900, and KS-HT) were 97 extensively degraded, while KS-1800 and KS-ER showed an excellent resistance against low temperature oxidation. The rate of alumina formation on KS-ER was relatively higher than the amorphous silica film formation on the other types of composite MoSi2 materials. KS-HT exhibited better high temperature oxidation resistance among all the materials studied here. The greenish/yellowish samples signifies the existence of mostly WO3 on the surface and it formed during thermal aging in the temperature range 300-900°C. The dark and dull color of the specimens indicates the presence of MoSi2 on the surface; the shiny dark indicates amorphous silica; and WO3 and/or MoO3 imparts a yellow/green color.  The whitish color is for a mix of MoO3 and SiO2. Another finding of this investigation was formation MoO3 at 300°C after 240 hrs. There is no publication reporting active oxidation of MoSi2 at a temperature as low as 300°C. The negative mass variations recorded at 800 and 900°C is attributed to a combination of formation of MoO3, WO3, and SiO2 and evaporation of MoO3. In high temperature oxidation experiments, a dense barrier alumina film (1.5 μm thick at 1000°C to 50 μm thick at 1500°C for up to 144 hrs) formed on KS-ER samples. A dense glassy SiO2 film (3 μm thick at 1000°C to 50 μm thick at 1600°C for up to 144 hrs) formed on the other types of samples. The glass scale on the surface of KS-1700, KS-1800 and KS-HT was significantly thicker (~3 times) than that on KS-HT over the temperature range of 1200°C to 1600°C after 144 hrs. The rate of alumina formation of KS-ER was relatively higher than the glass film formation of the other types of composite MoSi2 materials. The differences in the oxidation behavior of various MoSi2-based materials were linked to their chemistry and phase compositions.  Since the chemical and physical modifications of the sample coupons under the given conditions were so diverse, in order to be able to compare and rank the materials studied here, a new failure criterion was proposed. This is a failure criterion signified by the accumulation of the degradation via summation of the absolute values of mass gain and mass loss at different temperatures. The consideration of the absolute values was dictated due to simultaneous mass gain due to the materials oxidation, and mass loss due to 98 evaporation of some of the oxidation products. The parabolic rate constants calculated from the slopes of the line obtained by plotting weight gain square (mass gain)2 versus time by 2 are 0.001, 0.0063, 0.002, 0.0241 and 0.0827 (mg2.cm-4.0.5h-2) at 1200, 1300, 1400, 1500 and 1600°C, respectively. A special interest was also paid to calculate the oxidation activation energies of most materials studied here. The oxidation reaction activation energy and the oxide film thicknesses of each of the MoSi2 based material were obtained from the mass change vs temperature data and the cross section SEM image analysis. The values are: 236.0 kJ.mol-1 for KS-1700; 305.5 for KS-1800; 313.1 kJ.mol-1  for KS-ER; dual values for KS-1900 (72.4 up to 1400°C and 752.4 kJ.mol-1  above this temperature). The thickness of the oxide film after very long periods of time (up to 10,000 hrs) was measured and/or predicted. This extensive investigation covered five types of Kanthal Super materials, over the very broad range of temperatures, between 300°C and 1600°C. In summary, the most important findings of the investigation are as follows:  The oxidation behavior of MoSi2 based materials has been studied with emphasis on the constitution of the oxide scale and the kinetics of oxidation under different temperature regimes.    The oxidation behavior of MoSi2 - based materials greatly depends on their composition, and the exposure time and temperature.   Kanthal-1800 and Kanthal-ER showed an excellent resistance against the low temperature oxidation.  KS-HT and KS-1900 have both shown extensive degradation by low temperature oxidation.  High temperature oxidation product of four types of materials studied here (KS-1700, KS-1800, KS-1900, and KS-HT) is silica and the scale forms at high temperature on the surface of KS-ER is alumina.  A dense barrier alumina film formed on the KS-ER specimens and a dense glassy SiO2 film formed on the other types of samples in the high temperature oxidation.  The rate of alumina formation on KS-ER was relatively higher than the amorphous silica film formation on the other types of composite MoSi2 materials. 99  Spallation of the SiO2 scale did not occur at any temperature; the Al2O3 scale spalled at the temperature of 1600°C but did not occur at any other temperature below this temperature.  The oxidation activation energy was calculated (Arrhenius equation) for most materials studied.  The thickness of the thermally grown surface deposit was measured and or predicted up to 10,000 hrs of temperature/time exposure.  100 References [1] M. L. Willi and B. G. Richards, “Design and development of a direct injected, glow plug ignition-assisted, natural gas engine,” J. Eng. Gas Turbines Power, vol. 117, no. 4, p. 799, 1995. [2] O. Honigschmid, “About the molybdenum silicide MoSi2, the tungsten and tantalum WSi2 TaSi2,” vol. 1069, no. 8, pp. 1017–1028, 1907. [3] H. Mehrer, H. E. Schaefer, I. V. Belova, and G. E. Murch, “Molybdenum disilicide-diffusion, defects, diffusion correlation, and creep,” Defect Diffus. Forum, vol. 322, pp. 107–128, 2012. [4] J. J. Petrovic and a. K. Vasudevan, “Key developments in high temperature structural silicides,” Mater. Sci. Eng. A, vol. 261, no. 1–2, pp. 1–5, 1999. [5] R. . Aikin, “On the ductile-to-brittle transition temperature in MoSi2,” Scr. Metall. Mater., vol. 26, no. 7, pp. 1025–1030, 1992. [6] T. A. Kircher and E. L. Courtright, “Engineering limitations of MoSi2 coatings,” Mater. Sci. Eng. A, vol. 155, no. 1–2, pp. 67–74, 1992. [7] Z. Yao, J. Stiglich, and T. S. Sudarshan, “Molybdenum silicide based materials and their properties,” J. Mater. Eng. Perform., vol. 8, no. 3, pp. 291–304, 1999. [8] K. Vasudévan and J. J. Petrovic, “A comparative overview of molybdenum disilicide composites,” Mater. Sci. Eng. A, vol. 155, no. 1–2, pp. 1–17, 1992. [9] R. Mitra, “Mechanical behaviour and oxidation resistance of structural silicides,” Int. Mater. Rev., vol. 51, no. 1, pp. 13–64, 2006. [10] Kanthal. Ab, “Kanthal Super electric heating element handbook,” 2015. [Online]. Available: http://kanthal.com/en/products/furnace-products-and-heating-systems/electric-heating-elements/molybdenum-disilicide-heating-elements/. [11] G. B. Brook, Smithells Light Metals Handbook. 1992. [12] R. Keiffer, K. Konopicky, F. Benesovsky. “Austrian Patent,”, 179, 100, 1951. [13] J. M. Herbert, Electroceramics. Materials, Properties, Applications. Chapman and Hall, 1991. [14] E. Fitzer, “Heat and corrosion resistant sintered materials,” in Proc. 2nd Plansee Semin., Springer, Vienna, 1953, p. 1953. [15] E. Erdoes, “Investigation of surface layers of gas turbine alloys,” Internal Rep., 1971. [16] P. J. Meschter, “Oxidation of Nb particulate-reinforced MoSi2,” Scr. Metall. Mater., vol. 25, pp. 521–524, 1991. [17] P. J. Meschter, “Oxidation of MoSi2-TiB2 and MoSi2- Al2O3 mixtures,” Scr. Metall. Mater., vol. 25, pp. 1065–1069, 1991. 101 [18] Y. L. Jeng and E. J. Lavernia, “Processing of molybdenum disilicide,” J. Mater. Sci., vol. 29, no. 10, pp. 2557–2571, 1994. [19] D. E. Alman, K. G. Shaw, N. S. Stoloff, and K. R, “Fabrication, structure and properties of MoSi2-base composites,” Mater. Sci. Eng. A, vol. 155, pp. 85–93, 1992. [20] M. Koizumi, M., Nishihara, Isostatic pressing: technology and applications, vol. 50, Springer; no. 12. 1991. [21] Atkinson, H.V., Rickinson, B.A., "Hot Isostatic Processing. Springer" 1991 edition (December 31, 1991). [22] J. Gao, L. Wang, and W. Jiang, “Enhancement of properties and performance of MoSi2-based heating elements via low temperature sintering,” Int. J. Appl. Ceram. Technol., vol. 10, pp. E234–E239, 2013. [23] K. Bundschuh, M. Schuze, C. Muller, P. Greil, and W. Heider, “Selection of materials for use at temperatures above 1500°C in oxidising atmospheres,” J. Eur. Ceram. Soc., vol. 18, no. 1998, pp. 2389–2391, 1999. [24] R. B. Schwarz, S. R. Srinivasan, J. Petrovic, and C. J. Maggiore, “Synthesis of molybdenum disilicide by mechanical alloying,” Mater. Sci. Eng. A, vol. 155, no. 1–2, pp. 75–83, 1992. [25] Costa e Silva and M. J. Kaufman, “Applications of in situ reactions to MoSi2-based materials,” Mater. Sci. Eng. A, vol. 195, pp. 75–88, 1995. [26] S. Jayashankar and M. J. Kaufman, “Tailored MoSi2/SiC composites by mechanical alloying,” J. Mater. Res., vol. 8, no. 06, pp. 1428–1441, 1993. [27] E. K. Nyutu, M. a. Kmetz, and S. L. Suib, “Formation of MoSi2-SiO2 coatings on molybdenum substrates by CVD/MOCVD,” Surf. Coatings Technol., vol. 200, no. 12–13, pp. 3980–3986, 2006. [28] X. Fei, Y. Niu, H. Ji, L. Huang, and X. Zheng, “A comparative study of MoSi2 coatings manufactured by atmospheric and vacuum plasma spray processes,” Ceram. Int., vol. 37, no. 3, pp. 813–817, 2011. [29] Y. Jeng, J. Wolfenstine, E. J. Lavernia, D. E. Bailey, and A. Sickinger, “Low-pressure plasma deposition of SiC-reinforced MoSi2,” Scr. Metall. Mater., vol. 28, no. c, pp. 453–458, 1993. [30] E. J. L. and J. W. Y.-L. Jeng, “Creep behavior of plasma-sprayed SiC-reinforced MoSi2,” Scr. Metall. Mater., vol. 29, no. c, pp. 107–111, 1993. [31] C. H. Henager, Jr., J. L. Brimhall, “Synthesis of a MoSi2-SiC composite in situ using a solid state displacement reaction between Mo2C and Si,” Scr. Metall. Mater., vol. 26, no. Figure 1, pp. 585–589, 1992. [32] Henager, J. Brimhall, and J. Hirth, “Synthesis of a composite in situ using a solid state displacement reaction between Mo2C and Si,” Scr. Metall. Mater., vol. 26, no. 4, pp. 585–589, 1992. 102 [33] A. Sharif, “High-temperature oxidation of MoSi2,” J. Mater. Sci., vol. 45, no. 4, pp. 865–870, 2010. [34] J. Petrovic, “Mechanical behavior of MoSi2 , and MoSi2 composites,” Mater. Sci. Eng. A, vol. 193, pp. 31–37, 1995. [35] U. V Waghmare, E. Kaxiras, V. V. Bulatov, and M. S. Duesberry, “Effects of alloying on the ductility of MoSi2 single crystals from first-principles calculations,” Model. Simul. Mater. Sci. Eng., vol. 6, pp. 493–506, 1998. [36] M. Meyer, M. Kramer, and M. Akinc, “Boron-doped molybdenum silicides,” Adv. Mater., vol. 8, no. 1, pp. 85–88, 1996. [37] E. L. Courtright, “A comparison of MoSi2 matrix composites with other silicon-base composite systems,” Mater. Sci. Eng. A, vol. 261, no. 1–2, pp. 53–63, 1999. [38] K. Natesan and S. C. Deevi, “Oxidation behavior of molybdenum silicides and their composites,” Intermetallics, vol. 8, no. 9–11, pp. 1147–1158, 2000. [39] S. Köbel, J. Plüschke, U. Vogt, and T. J. Graule, “MoSi2-Al2O3 electroconductive ceramic composites,” Ceram. Int., vol. 30, no. 8, pp. 2105–2110, 2004. [40] D. Mazzoni and E. F. Aglietti, “Mechanism of the carbonitriding reactions of SiO2 – Al2O3 minerals in the Si – Al – O – N system,” Appl. Clay Sci., pp. 447–461, 1998. [41] J. Subrahmanyam and R. M. Rao, “Combustion synthesis of MoSi2-WSi2 alloys,” Mater. Sci. Eng. A, vol. 183, pp. 205–210, 1994. [42] H. J. Grabke and G. H. Meier, “Accelerated oxidation, internal oxidation, intergranular oxidation, and pesting of intermetallic compounds,” Oxid. Met., vol. 44, no. 1–2, pp. 147–176, 1995. [43] T. C. Chou and T. G. Nieh, “Kinetics of MoSi2 pest during low-temperature oxidation,” J. Mater. Res., vol. 8, no. 07, pp. 1605–1610, 1993. [44] T. C. Chou and T. G. Nieh, “Pest disintegration of thin MoSi2 films by oxidation at 500°C,” J. Mater. Sci., vol. 29, no. 11, pp. 2963–2967, 1994. [45] S. Chevalier, F. Bernard, E. Gaffet, S. Paris, Z. a. Munir, and J. P. Larpin, “Effect of microstructure on the high temperature oxidation and pesting behaviour of MoSi2,” Mater. Sci. Forum, vol. 461–464, pp. 439–446, 2004. [46] C. G. McKamey, P. F. Tortorelli, J. H. DeVan, and C. A. Carmichael, “A study of pest oxidation in polycrystalline MoSi2,” J. Mater. Res., vol. 7, no. 10, pp. 2747–2755, 1992. [47] T. C. Chou, T. G. Nieh, and P. Alto, “Comparative studies on the pest reactions of single and poly crystalline MoSi2,” Scr. Metall. Mater., vol. 27, pp. 19–24, 1992. [48] K. Yanagihara, K. Przybylski, and T. Maruyama, “The role of microstructure on pesting during oxidation of MoSi2 and Mo(Si,Al)2 at 773K,” Oxid. Met., vol. 47, no. 3–4, pp. 277–293, 1997. 103 [49] T. C. Chou and T. G. Nieh, “New observations of MoSi2 pest at 500°C,” Scr. Metall. Mater., vol. 26, no. 10, pp. 1637–1642, 1992. [50] S. Knittel, S. Mathieu, and M. Vilasi, “Oxidation behaviour of arc-melted and uniaxial hot pressed MoSi2 at 500°C,” Intermetallics, vol. 18, no. 12, pp. 2267–2274, 2010. [51] F. Zhang, L. Zhang, A. Shan, and J. Wu, “Oxidation of stoichiometric poly- and single-crystalline MoSi2 at 773K,” Intermetallics, vol. 14, no. 4, pp. 406–411, 2006. [52] K. Kurokawa, H. Houzumi, I. Saeki, and H. Takahashi, “Low temperature oxidation of fully dense and porous MoSi2,” Mater. Sci. Eng. A, vol. 261, no. 1–2, pp. 292–299, 1999. [53] R. Chen, J., Li, C., Fu, Z., Tu, X., Sundberg, M., Pompe, “Low temperature oxidation behavior of a MoSi2 -based material,” Mater. Sci. Eng. A, vol. 261, pp. 239–244, 1999. [54] J. Cook, A. Khan, and N. Air, “Oxidation of MoSi2-based composites,” Mater. Sci. Eng. A, vol. 155, pp. 183–198, 1992. [55] P. J. Meschter, “Low-temperature oxidation of molybdenum disilicide,” Metall. Trans. A, vol. 23, no. 6, pp. 1763–1772, 1992. [56] J. H. Westbrook, “Applications of Intermetallic Compounds,” MRS Bull., May, p. 26, 1996. [57] J. Yan, H. Xu, H. Zhang, and S. Tang, “MoSi2 oxidation resistance coatings for Mo5Si3/MoSi2 composites,” Rare Met., vol. 28, no. 4, pp. 418–422, 2009. [58] Y. Liu, G. Shao, and P. Tsakiropoulos, “On the oxidation behaviour of MoSi2,” Intermetallics, vol. 9, pp. 125–136, 2001. [59] K. Hansson, M. Halvarsson, J. E. Tang, R. Pompe, M. Sundberg, and J. E. Svensson, “Oxidation behaviour of a MoSi2-based composite in different atmospheres in the low temperature range (400-550°C),” J. Eur. Ceram. Soc., vol. 24, no. 13, pp. 3559–3573, 2004. [60] Y. Liu, G. Shao, and P. Tsakiropoulos, “Thermodynamic reassessment of the Mo-Si and Al-Mo-Si systems,” Calphad Comput. Coupling Phase Diagrams Thermochem., vol. 8, 2000. [61] S. Hou, Z. Liu, D. Liu, and B. Li, “Effect of alloying with Al on oxidation behavior of MoSi2 coatings at 1100°C,” Surf. Coatings Technol., vol. 206, no. 21, pp. 4466–4470, 2012. [62] B. Hallstedt, “Thermodynamic assessment of the Silicon — Oxygen system,” Calphad, vol. 16, no. 1, pp. 53–61, 1992. [63] S. S. Bygden J, Sichen Du, “Thermodynamic activities of FeO in CaO-FeO-SiO2 slags.” 1994. [64] A. T. Dinsdale, “SGTE data for pure elements,” Calphad, vol. 15, no. 4, pp. 317–425, 1991. [65] a. Sharif, “Effects of Nb-alloying on high-temperature oxidation of MoSi2,” Acta Phys. Pol. A, vol. 125, no. 2, pp. 563–566, 2014. 104 [66] K. Hansson, J. E. Tang, M. Halvarsson, R. Pompe, M. Sundberg, and J. E. Svensson, “The beneficial effect of water vapour on the oxidation at 600 and 700°C of a MoSi2-based composite,” J. Eur. Ceram. Soc., vol. 25, no. 1, pp. 1–11, 2005. [67] K. Hansson, J. E. Svensson, M. Halvarsson, J. E. Tang, M. Sundberg, and R. Pompe, “The influence of water vapor on the oxidation of MoSi2 at 450°C,” Mater. Sci. Forum, vol. 369–372, pp. 419–426, 2001. [68] Z. Basinski, J. Dugdale, A. Howie, P. Mag, P. Thornton, S. Electron, J. B. Berkowitz-mattuck, and I. June, “High temperature oxidation,” J. Electrochem. Soc., vol. 112, no. 6, pp. 583–589, 1964. [69] Y. A. Chang, “Oxidation of molybdenum disilicide,” J. Mater. Sci., vol. 4, no. 7, pp. 641–643, 1969. [70] G. B. Cherniack and A. G. Elliot, “High temperature behavior of MoSi2 and Mo5Si3,” J. Am. Ceram. Soc., vol. 91, no. published 1956, 1960. [71] P. Feng, X. Wang, Y. He, and Y. Qiang, “Effect of high-temperature preoxidation treatment on the low-temperature oxidation behavior of a MoSi2-based composite at 500°C,” J. Alloys Compd., vol. 473, no. 1–2, pp. 185–189, 2009. [72] A. Ibano, K. Yoshimi, A. Yamauchi, R. Tu, K. Maruyama, K. Kurokawa, and T. Goto, “High temperature oxidation behavior of MoSi2 under low pressure atmosphere,” Mater. Sci. Forum, vol. 561–565, pp. 427–430, 2007. [73] C. E. Ramberg, “High-temperature oxidation studies of several silicon-based systems,” (Doctoral dissertation) University of Pennsylvania, 2001. [74] S. K. Ramasesha and K. Shobu, “Oxidation of MoSi2 and MoSi2 based materials,” Bull. Mater. Sci., vol. 22, no. 4, pp. 769–773, 1999. [75] M. Sundberg, G. Malmqvist, a. Magnusson, and T. El-Raghy, “Alumina forming high temperature silicides and carbides,” Ceram. Int., vol. 30, no. 7, pp. 1899–1904, 2004. [76] L. Ingemarsson, M. Halvarsson, J. Engkvist, T. Jonsson, K. Hellström, L. G. Johansson, and J. E. Svensson, “Oxidation behavior of a Mo (Si, Al)2-based composite at 300-1000°C,” Intermetallics, vol. 18, no. 4, pp. 633–640, 2010. [77] M. Halvarsson, T. Jonsson, L. Ingemarsson, M. Sundberg, J.-E. Svensson, and L. G. Johansson, “Microstructural investigation of the initial oxidation at 1450°C and 1500°C of a Mo(Si,Al)2 based composite,” Mater. High Temp., vol. 26, no. 2, pp. 137–143, 2009. [78] L. Ingemarsson, K. Hellström, L. G. Johansson, J. E. Svensson, and M. Halvarsson, “Oxidation behaviour of a Mo(Si,Al)2 based composite at 1500°C,” Intermetallics, vol. 19, no. 9, pp. 1319–1329, 2011. [79] S. Becker, “Oxidation of TiSi2 and MoSi2,” Solid State Ionics, vol. 53–56, pp. 280–289, 1992. 105 [80] A. Mueller, G. Wang, R. a. Rapp, E. L. Courtright, and T. a. Kircher, “Oxidation behavior of tungsten and germanium-alloyed molybdenum disilicide coatings,” Mater. Sci. Eng. A, vol. 155, no. 1–2, pp. 199–207, 1992. [81] K. Maruyama, T., Yanagihara and K. Nagata, “High temperature oxidation of intermetallic compounds of Mo(Siy, Alx)2,” Corros. Sci., vol. 35, pp. 939–944, 1993. [82] K. Yanagihara, T. Maruyama, and K. Nagata, “High temperature oxidation of Mo-Si-X intermetallics (X=Al, Ti, Ta, Zr and Y),” Intermetallics, vol. 3, no. 3, pp. 243–251, 1995. [83] L. Ingemarsson, M. Halvarsson, K. Hellström, T. Jonsson, M. Sundberg, L. G. Johansson, and J. E. Svensson, “Oxidation behavior at 300-1000°C of a (Mo,W)Si2-based composite containing boride,” Intermetallics, vol. 18, no. 1, pp. 77–86, 2010. [84] K. Maria, K. M. N, and C. T. Hogskola, “On the oxidation behaviour of molybdenum silicide and ( molybdenum , tungsten ) silicon : The influence of oxygen and water vapour,” 2003. [85] H. S. Kim, J. K. Yoon, G. H. Kim, J. M. Doh, S. I. Kwun, and K. T. Hong, “Growth behavior and microstructure of oxide scales grown on WSi2 coating,” Intermetallics, vol. 16, no. 3, pp. 360–372, 2008. [86] J. K. Yoon, K. W. Lee, S. J. Chung, I. J. Shon, J. M. Doh, and G. H. Kim, “Growth kinetics and oxidation behavior of WSi2 coating formed by chemical vapor deposition of Si on W substrate,” J. Alloys Compd., vol. 420, no. 1–2, pp. 199–206, 2006. [87] Produced, R. Electron, and T. Nonaka, “High temperature oxidation of MoSi2 -WSi2 solid solutions library’s Catalog ( OPAC ) Similar Items :,” Trans Tech Publ., no. 111, pp. 10–11, 1997. [88] L. Ingemarsson, K. Hellström, S. Canovic, T. Jonsson, M. Halvarsson, L. G. Johansson, and J. E. Svensson, “Oxidation behavior of a Mo(Si, Al)2 composite at 900-1600°C in dry air,” J. Mater. Sci., vol. 48, no. 4, pp. 1511–1523, 2013. [89] K. Hellström, P. Persson, and E. Ström, “Oxidation behaviors and microstructural alterations of a Mo(Si, Al)2-based composite after heating at 1580°C either in a furnace (ex-situ) or via alternating current (in-situ),” J. Eur. Ceram. Soc., vol. 35, no. 2, pp. 513–523, 2015. [90] S. Prasad and A. Paul, “Growth mechanism of phases by interdiffusion and atomic mechanism of diffusion in the molybdenum-silicon system,” Intermetallics, vol. 19, no. 8, pp. 1191–1200, 2011. [91] N. Birks, G. H. Meier, and F. S. Pettit, “Introduction to the high temperature oxidation of metals,” Engineering, May, 2006. [92] T. Narushima, T. Goto, and T. Hirai, “High-temperature passive oxidation of chemically vapor deposited silicon carbide,” J. Am. Ceram. Soc., vol. 72, no. 8, pp. 1386–1390, 1989. [93] T. Maruyama and K. Yanagihara, “High temperature oxidation and pesting of Mo (Si, Al )2,” Mater. Sci. Eng. A, vol. 240, pp. 828–841, 1997.   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0166781/manifest

Comment

Related Items