UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Yttrium disilicate as environmental barrier coating for silicon nitride-based glow plug Lin, Xin 2017

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

Item Metadata

Download

Media
24-ubc_2018_february_lin_xin.pdf [ 8.17MB ]
Metadata
JSON: 24-1.0362389.json
JSON-LD: 24-1.0362389-ld.json
RDF/XML (Pretty): 24-1.0362389-rdf.xml
RDF/JSON: 24-1.0362389-rdf.json
Turtle: 24-1.0362389-turtle.txt
N-Triples: 24-1.0362389-rdf-ntriples.txt
Original Record: 24-1.0362389-source.json
Full Text
24-1.0362389-fulltext.txt
Citation
24-1.0362389.ris

Full Text

 YTTRIUM DISILICATE AS ENVIRONMENTAL BARRIER COATING FOR SILICON NITRIDE-BASED GLOW PLUG by  Xin Lin  B.Eng., Zhejiang University, 2015  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)  December 2017  © Xin Lin, 2017 ii  Abstract  Silicon-based ceramics undergo severe degradation at temperatures above ~1000 ℃ in the presence of water vapor, which is inevitable in combustion environments. Therefore, Environmental Barrier Coatings (EBCs) are necessary for the protection of Si3N4-based ceramic components in the harsh combustion environments. Rare earth silicates, which have relatively low thermal expansion coefficients, good chemical stability at high temperatures and low recession rates in the presence of water vapor, are promising candidate materials for such EBC application. This study was related to the application of Si3N4 as part of Hot Surface Ignition Systems (“Glow-Plugs”, GP) in High-Pressure Natural Gas Direct Injection engines, currently under development by Vancouver company Westport Fuel Systems Inc. For certain kinds of commercially available Si3N4-based GPs, the use of Y2O3 as sintering additive results in the in-service formation of yttrium silicates on their ceramic pins. Therefore, taking the chemical compatibility into consideration, yttrium disilicate coating was chosen to provide corrosion protection for such GPs. A sol-gel dip-coating route, which is simple, cost effective and industrially applicable, has been developed to apply multi-layer Y2Si2O7 EBCs on the GPs. Selective processing parameters, including the sol aging conditions and the withdrawal speed of the GP substrate during dip coating procedure, were investigated in detail. The thickness and microstructures of the coatings were controlled through the adjustment of these parameters during sol preparation and dip coating processes. To simultaneously achieve sufficient thickness and avoid the formation of cracks, thin layers of Y2Si2O7 coating, each with the thickness of ~1 µm, were successively applied and processed. The 6-layer crack-free coating was able to achieve an average thickness of ~5.5 µm. The microstructures of the coatings were evaluated and their performance was tested at ~1200 ℃ in iii  high concentration water vapor atmosphere and on a natural gas burner rig. Improved corrosion resistance of such EBC-protected glow plugs was observed in these tests. iv  Lay Summary  A glow plug is a heating device used for the ignition of combustion gases in diesel engines. Silicon nitride is a material that has been used to manufacture the heater of some commercial glow plugs. However, it is vulnerable under the attack of water vapor in combustion chambers. Therefore, a protective coating made up of yttrium disilicate, a material with a good resistance to water vapor, was applied on the surface of the silicon nitride component. This study employed a simple method to fabricate such coatings with good control of their microstructures and thickness. Selective processing parameters were investigated in order to make better coatings. Some relevant tests showed that the performance of the glow plugs improved under the protection of these coatings. v  Preface  This research project was a collaborative effort between the ceramic group (UBCeram) in the Department of Materials Engineering at the University of British Columbia and Westport Fuel Systems Inc. (Vancouver, Canada). The synthesis, characterization and High Concentration Water Vapor (HCWV) test of the materials were conducted at UBC. The natural gas burner rig test was carried out at Westport. For the experimental part of this study, I was responsible for the experimental design (process development of the sol-gel dip coating route), processing of the GP specimens and coatings, as well as the XRD analysis of the bulk materials. The SEM and EDX characterizatio presented in this thesis was performed by Mrs. Carmen Oprea. The HCWV test was collaboratively carried out by Mrs. Oprea and me. The burner rig test was conducted by Dr. Hamed Karimi Sharif. I was responsible for the analysis and interpretation of the data collected from the above experimental procedures, characterization methods and tests. This study was supervised by Prof. Tom Troczynski of UBC and co-supervised by Dr. Hamed Karimi Sharif of Westport. vi  Table of Contents  Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ..................................................................................................................................x List of Figures ............................................................................................................................... xi List of Symbols ........................................................................................................................... xiv List of Abbreviations ................................................................................................................. xvi Acknowledgements ................................................................................................................... xvii Dedication ................................................................................................................................. xviii Chapter 1: Introduction and Literature Review .......................................................................... 1 1.1 General background .................................................................................................... 1 1.2 Silicon nitride-based glow plugs ................................................................................. 1 1.2.1 Silicon nitride in water-vapor/oxygen environment ............................................... 3 1.2.2 The self-formation of silicates on glow plugs......................................................... 5 1.3 Environmental barrier coatings ................................................................................... 6 1.3.1 Requirements for EBC ............................................................................................ 6 1.3.2 Candidate materials as EBC for silicon-based ceramics ......................................... 7 1.3.3 Advantages of rare-earth silicates as EBCs ............................................................ 9 1.4 Y2O3-SiO2 system ..................................................................................................... 11 1.4.1 The prospects of Y2Si2O7 and Y2SiO5 as EBCs for Si3N4 .................................... 12 vii  1.4.2 Phase transformations of Y2Si2O7 ......................................................................... 14 1.4.3 Synthesis of yttrium silicates ................................................................................ 17 1.4.4 Recent research on yttrium/ytterbium silicates as EBCs ...................................... 17 1.5 Sol-gel dip coating .................................................................................................... 19 1.5.1 Sol-gel process ...................................................................................................... 19 1.5.2 Dip coating ............................................................................................................ 19 Chapter 2: Scope and Objectives .............................................................................................. 23 2.1 Scope of the research ................................................................................................ 23 2.2 Objectives of the research ......................................................................................... 24 Chapter 3: Experimental Procedures ........................................................................................ 26 3.1 Sol preparation .......................................................................................................... 26 3.2 Preparation of the substrate ....................................................................................... 27 3.2.1 Surface cleaning .................................................................................................... 28 3.2.2 Surface treatment by sanding ................................................................................ 28 3.2.3 Electrical characterization of the glow plug ......................................................... 28 3.3 Dip coating ................................................................................................................ 29 3.3.1 Immersion and withdrawal of the glow plug ........................................................ 29 3.3.2 In-situ heat treatment of the applied coating ......................................................... 30 3.3.3 Fabrication of multi-layer coatings ....................................................................... 30 3.3.4 Summary of the processing conditions of GP specimens ..................................... 31 3.4 Synthesis of monolithic samples ............................................................................... 32 3.5 Sample characterization ............................................................................................ 32 3.5.1 SEM and EDS ....................................................................................................... 32 viii  3.5.2 XRD ...................................................................................................................... 35 3.6 Tests on the performances of the as-received and coated Glow Plugs ..................... 35 3.6.1 Natural gas burner rig test ..................................................................................... 35 3.6.2 High concentration water vapor test ..................................................................... 36 Chapter 4: Self-formed Y2Si2O7 on Le-Mark Glow Plugs ....................................................... 37 4.1 Introduction ............................................................................................................... 37 4.2 As-received Le-Mark GP .......................................................................................... 37 4.3 Le-Mark GP after the electrical characterization ...................................................... 38 4.4 Self-formed yttrium silicates on Le-Mark GP .......................................................... 41 4.5 Discussion and conclusion ........................................................................................ 45 Chapter 5: Sol-gel Processing of Y2Si2O7 Coatings and Bulk Materials .................................. 46 5.1 Introduction ............................................................................................................... 46 5.2 A Speculation on the reaction mechanisms of the sol-gel processed Y2Si2O7 ......... 46 5.3 Effects of sol aging conditions .................................................................................. 47 5.4 Phase analysis of bulk Y2Si2O7 materials ................................................................. 51 5.5 In-situ sintering ......................................................................................................... 53 5.6 Discussion and conclusion ........................................................................................ 55 Chapter 6: Effects of The Dip-coating Withdrawal Speeds ...................................................... 56 6.1 Introduction ............................................................................................................... 56 6.2 The effects of withdrawal speed at 2 mm from GP tip ............................................. 57 6.3 The effects of withdrawal speed at 4 mm from GP tip ............................................. 61 6.4 The effects of withdrawal speed at 6 mm from GP tip ............................................. 63 6.5 Discussion and conclusion ........................................................................................ 66 ix  Chapter 7: Multi-layer Y2Si2O7 Coatings ................................................................................. 68 7.1 Introduction ............................................................................................................... 68 7.2 Multi-layer Y2Si2O7 coatings at 2 mm from the GP tip ............................................ 68 7.3 Multi-layer Y2Si2O7 coatings at 4 mm from the GP tip ............................................ 71 7.4 Multi-layer Y2Si2O7 coatings at 6 mm from the GP tip ............................................ 74 7.5 Thickness and adhesion of the multi-layer Y2Si2O7 coatings ................................... 77 7.6 Discussion and conclusion ........................................................................................ 79 Chapter 8: Tests of the Performance of the Coated Glow Plugs .............................................. 80 8.1 Introduction ............................................................................................................... 80 8.2 Natural gas burner rig test ......................................................................................... 80 8.2.1 Burner rig tests for the uncoated glow plug .......................................................... 80 8.2.2 Burner rig test for the glow plug with Y2Si2O7 coating ........................................ 82 8.3 High concentration water vapor test ......................................................................... 85 8.4 Discussion and conclusion ........................................................................................ 89 Chapter 9: Summary and Conclusions ...................................................................................... 90 Bibliography .................................................................................................................................93 Appendices ..................................................................................................................................102  x  List of Tables  Table 1.1 Summary of the relevant properties of yttrium silicates and Si3N4 .............................. 10 Table 1.2 Crystal structures of polymorphs of Y2Si2O7 ............................................................... 15 Table 1.3 Models for the prediction of dip-coated film thickness [54] ........................................ 21 Table 3.1 Summary of the sols prepared and their synthesis conditions ...................................... 27 Table 3.2 Summary of the processing conditions of GP specimens ............................................. 31 Table 4.1 Composition of as-received Le-Mark glow plug (by EDS) after electrical characterization at ......................................................................................................................... 40 Table 4.2 EDS results of as-received Le-Mark glow plug after in-situ heat treatment at ~ 1260 ℃ for 24 h .......................................................................................................................................... 42 Table 4.3 EDS results of as-received Le-Mark glow plug after in-situ sintering at ~1260 ℃ for 100 h.............................................................................................................................................. 45 Table 6.1 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.1 ....... 60 Table 6.2 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.3 ....... 63 Table 6.3 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.5 ....... 65 Table 7.1 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.1 ...... 70 Table 7.2 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.3 ...... 73 Table 7.3 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.5 ...... 76  xi  List of Figures  Figure 1.1 Micrograph of the cross-section through the length of the ceramic heater of a Le-Mark ceramic glow plug (a) and mode of Direct Injection Combustion and the installation of GP in NGDI engine (b) [7]........................................................................................................................ 2 Figure 1.2 Profiles of the temperature/voltage/current/resistance of Le-Mark 12V all-ceramic glow plug during heat-up time [7] .................................................................................................. 3 Figure 1.3 Change in the temperature capability of ceramic matrix composites with EBCs [27] . 8 Figure 1.4 The phase diagram of the calculated Y2O3-SiO2 system[35] ...................................... 12 Figure 1.5 Crystal structures of: (a) y-Y2Si2O7, (b) α-Y2Si2O7, (c) β-Y2Si2O7, (d) γ-Y2Si2O7, (e) δ-Y2Si2O7 according to the parameters listed in Table 1.2 [15] ................................................... 16 Figure 1.6 Schematic illustration of the dip-coating method and both capillarity and draining regimes involved at low and fast withdrawal speeds, respectively [55] ....................................... 20 Figure 3.1 Synthesis steps for the fabrication of yttrium disilicate coatings ................................ 26 Figure 3.2 Le-Mark QTJ5-12 glow plug....................................................................................... 27 Figure 3.3 Experimental setup for dip coating with the assistance of Instron testing system ...... 29 Figure 3.4 Experimental setup for in-situ sintering via direct current power supply ................... 30 Figure 3.5 Experimental setup for natural gas burner rig test....................................................... 36 Figure 3.6 Experimental setup for high concentration water vapor test ....................................... 36 Figure 4.1 Micrograph of the surface of the ceramic pin of an as-received Le-Mark QTJ5-12 glow plug ...................................................................................................................................... 38 Figure 4.2 Micrographs of the surface of Le-Mark glow plug after electrical characterization at 12V (~1260 ℃) for 10 min: .......................................................................................................... 39 xii  Figure 4.3 Micrographs of the surface of Le-Mark glow plug after in-situ heat treatment at 12 V (~1260 ℃) for 24 h: ...................................................................................................................... 42 Figure 4.4 Micrographs of the surface of Le-Mark glow plug after in-situ heat treatment at 12 V (~1260 ℃) for 100 h: .................................................................................................................... 44 Figure 5.1 Micrographs of coatings prepared with different sols after drying: ............................ 48 Figure 5.2 XRD patterns of the bulk materials obtained from sol Y75C16h after heat treatment at:................................................................................................................................................... 52 Figure 5.3 Micrographs of coatings prepared with different sols after in-situ sintering: ............. 53 Figure 6.1 Micrographs of the coatings from different withdrawal speeds at the distance of 2 mm from tip: ........................................................................................................................................ 59 Figure 6.2 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.1 ............ 60 Figure 6.3 Micrographs of coatings from different withdrawal speeds at the distance of 4 mm from tip: ........................................................................................................................................ 62 Figure 6.4 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.2 ............ 63 Figure 6.5 Micrographs of coatings from different withdrawal speeds at the distance 6 mm from tip: ................................................................................................................................................. 65 Figure 6.6 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.3 ............ 66 Figure 7.1 Micrographs of LM GPs (2 mm from tip) with different layers of coatings after in-situ sintering: ....................................................................................................................................... 69 Figure 7.2 Elemental contents of the multi-layer coatings at 2 mm from the GP tips .................. 70 Figure 7.3 Micrographs of LM GPs (4 mm from tip) with different layers of coatings after in-situ sintering: ....................................................................................................................................... 73 Figure 7.4 Elemental contents of the multi-layer coatings at 4 mm from the GP tips .................. 74 xiii  Figure 7.5 Micrographs of LM GPs (6 mm from tip) with different layers of coatings after in-situ sintering: ....................................................................................................................................... 75 Figure 7.6 Elemental contents of the multi-layer coatings at 6 mm from the GP tips .................. 76 Figure 7.7 Micrographs of the cross-sections of Le-Mark GPs with different layers of Y2Si2O7 coatings at 2 mm from GP tips: .................................................................................................... 78 Figure 8.1 Micrographs of the surface of the ceramic pin for specimen BRGP0 after BR test for 3 weeks: ........................................................................................................................................... 81 Figure 8.2 Micrographs of the surface of the ceramic pin for specimen BRGP1 before (a) and after the 3-week BR test (b-j): ...................................................................................................... 84 Figure 8.3 Micrographs of the surface of the ceramic pin for specimen HCWVGP0 at 5 mm from tip after HCWV test for 300 h....................................................................................................... 85 Figure 8.4 Micrographs of the surface of the ceramic pin for specimen HCWVGP1 before and after HCWV test for 300 h: ........................................................................................................... 87  xiv  List of Symbols  t oxidation time [h] MW molecular weight [g∙mol-1] kp oxidation parabolic rate constant [mg2/(cm4∙h)] k1 linear volatilization rate constant [mg/(cm2∙h)] ∆w1 weight gain from SiO2 growth [mg∙cm-2] ∆w2 weight loss from SiO2 volatilization [mg∙cm-2] U0 withdrawal speed of the substrate during dip coating [mm∙s-1] h0 entrained film thickness [nm] hf final film thickness (after thermal treatment) [nm] η liquid viscosity [Pa∙s] ρ liquid density [g∙cm-3] g acceleration of gravity [m∙s-2] γLV liquid-vapor surface tension [J∙m-2] E evaporation rate of the solvent [m3∙s-1] ci inorganic precursor solution concentration [mol∙cm-3] Mi molar weight of inorganic material [g∙mol-1] αi fraction of inorganic material in the film [%] ρi density of the inorganic material [g∙cm-3] L width of the film [m] ki material proportion constant  xv  D global constant that combines the physiochemical constants of the precursor solution  xvi  List of Abbreviations (In alphabetical order)  BR Burner Rig CMC Ceramic Matrix Composite CNG Compressed Natural Gas CTE Coefficient of Thermal Expansion DC Direct Current EBC Environmental Barrier Coating EDS Energy Dispersive X-ray Spectroscopy EPMA Electron Probe Microanalysis GP Glow Plug HCWV High Concentration Water Vapor HSI Hot Surface Ignition LM Le-Mark NG Natural Gas NGDI Natural Gas Direct Injection RE Rare Earth SEM Scanning Electron Microscopy WDS Wavelength Dispersive X-ray Spectroscopy XRD X-Ray Diffraction  xvii  Acknowledgements  My biggest gratitude goes to my supervisor Professor Tom Troczynski for his enlightening guidance, continuous encouragement and immeasurable support throughout my program. I am deeply grateful to have a caring, generous and inspiring mentor who supports my exploration in research and in life.  I would like to acknowledge my co-supervisor Dr. Hamed Karimi Sharif for all the guidance, suggestions, and especially his long-term effort on running burner rig tests.  My sincere appreciation goes to Mrs. Carmen Oprea for her lasting help and guidance with sample characterization by SEM and EDS, as well as sharing her life wisdom with me.  And I appreciate my colleagues, Matteo and Shubham, for their help and friendship.  Natural Sciences and Engineering Research Council of Canada (NSERC) and Westport Fuel Systems are greatly appreciated for their financial support.  Special thanks to my father, mother and little brother. I am so grateful to have them by my side in this journey of learning and exploration. It is their love and support that make me keep going, now and in the future.  xviii  Dedication     TO MY FATHER, MOTHER AND BROTHER 致我最亲爱的——爸爸、妈妈和弟弟1  Chapter 1: Introduction and Literature Review  1.1 General background With its environmental and economic advantages, natural gas (NG) has been used as a cleaner alternative to other fossil fuels for vehicles. NG vehicles either use compressed natural gas (CNG) or liquefied natural gas (LNG). In efforts to build improved systems for CNG cars, natural gas direct injection (NGDI) engine is currently under development by Westport Fuel Systems. As NG ignition in the internal-combustion engine is more difficult than gasoline or diesel fuel, the permanent hot surface ignition (HSI) is a critical technology used for NGDI engines [1]. Silicon nitride-based ceramic glow plugs (GPs) have been expected to provide reliable hot surfaces for such application due to their good strength retention at elevated temperatures [1]. However, silicon-based ceramics undergo severe corrosive degradation in the harsh environment of combustion chamber, under the repeated reducing/oxidizing conditions in the presence of water vapor [2]. Externally applied environmental barrier coating (EBC) on GP surface is one of the most commonly suggested solutions for this issue [3].  1.2 Silicon nitride-based glow plugs A glow plug is made up of a pencil-shaped metal body and a heater that has a high melting point. The heating element can be either a metal heating coil (as is the case for metal glow plug and first-generation ceramic glow plug) or a ceramic resistor (for all-ceramic glow plug) [4]. A fully ceramic heater enables higher operation temperature and shorter heat-up time, down to 1-2 seconds from room temperature to 1000 ℃, as claimed by some manufacturers [4][5]. Acclaimed for its 2  excellent durability at high temperatures, silicon nitride has been used to manufacture ceramic glow plugs since 1980s [6]. The silicon nitride-based ceramic heater is typically made up of an electrically conductive ceramic resistor for heating and a Si3N4 body as insulation [4]. The major manufacturers of ceramic glow plugs include Kyocera, Le-Mark, Beru, NGK, Bosch. Different suppliers feature different technologies in the processing of the glow plugs, especially when it comes to the composition of the ceramic heater. Figure 1.1a shows the micrograph of the cross-section through the length of the ceramic heater of a Si3N4-based glow plug (type: QTJ5-12) manufactured by Le-Mark (Chongqing, China), which has been widely used in this study. Figure 1.1b illustrates the installation of glow plug in an NGDI engine [7]. Figure 1.2 presents the profiles of the temperature, voltage, current, resistance of the Le-Mark 12 V all-ceramic glow plug (type: QTJ5-12) during its heat-up time. (a)  (b)  Figure 1.1 Micrograph of the cross-section through the length of the ceramic heater of a Le-Mark ceramic glow plug (a) and mode of Direct Injection Combustion and the installation of GP in NGDI engine (b) [7] 2 mm resistor 3   Figure 1.2 Profiles of the temperature/voltage/current/resistance of Le-Mark 12V all-ceramic glow plug during heat-up time [7]  1.2.1 Silicon nitride in water-vapor/oxygen environment Silicon-based ceramics, such as silicon nitride (Si3N4) and silicon carbide (SiC), with their high-temperature strength and durability, have attracted wide interests for structural applications in high temperature environments [8]. With spontaneously formed protective silica scale on its surface in oxidizing environments, silicon nitride exhibits excellent high temperature oxidation resistance in dry air. However, in moist environments such as combustion chamber, the silica scale reacts with water vapor, forming gaseous silicon hydroxides like Si(OH)4 [9].  Also, combustion environment impurities that react with silica layer can also trigger accelerated oxidation and corrosion/erosion of silicon nitride [10]. 4  Thus, the unprotected silicon nitride is subjected to severe degradation in the presence of water vapor. The overall reactions for the formation and degradation of the surface silica layer can be summarized as: Si3N4(s) + 3O2(g) = 3SiO2(s) + 2N2(g) (1.1) SiO2(s) + 2H2O(g) = Si(OH)4(g) (1.2) Researchers [2][11][12] believe that the rate of the oxidation of the silicon-based ceramics can be described by a parabolic rate constant, kp, while the rate of its reaction with water vapor is given by a volatilization constant, k1. When oxidation and volatilization reactions happen simultaneously, the overall kinetics can be described by a paralinear kinetic model [2][12], which was originally proposed by Tedmon for Cr2O3-forming Fe-Cr alloys [11]: 𝑡 =𝛼2𝑘𝑝2(𝑘12)[−2𝑘1∆𝑤1𝛼𝑘𝑝− 𝑙𝑛 (1 −2𝑘1∆𝑤1𝛼𝑘𝑝)] (1.3) ∆𝑤2 = −𝛽𝑘1𝑡 (1.4)  Where, 𝛼 =𝑀𝑊𝑆𝑖𝑂2𝑀𝑊𝑂2 −23𝑀𝑊𝑁2 𝛽 = 𝑀𝑊𝑆𝑖𝑁1.3/𝑀𝑊𝑆𝑖𝑂2 Refer to List of Symbols for the meanings of the symbols used in Equation 1.3 and 1.4. Fox et al [2] studied three types of silicon nitride in 50% H2O – 50% O2 flowing at 4.4 cm/s and paralinear behaviors were observed for all of them. The researchers then used the linear 5  volatilization rate constant k1 obtained in terms of weight loss to predict the recession rate of the Si3N4 substrate. Under the combustion conditions of 1200 ℃, gas velocity of 21 m/s, total system pressure of 600 kPa (6 atm), and H2O partial pressure of 60 kPa (0.6 atm), the recession rate of AS800 Si3N4 from AlliedSignal Ceramic Components, CA, United States, is predicted to be 2.28 × 10-5 cm/h. As such, 100 h of exposure would subject the silicon nitride substrate to 28.8 µm of material loss, which is a significant degradation and limits the lifetime of such components when used under similar conditions. 1.2.2 The self-formation of silicates on glow plugs A mixture of oxides, especially rare earth oxides, are commonly used as sintering aids for the fabrication of silicon nitride ceramics with high temperature structural applications [13][14][15]. The sintering additives, such as Y2O3, Yb2O3, Al2O3 and MgO, have a tendency of outward migration from the Si3N4 body to the surface silica scale and forming the corresponding silicates. It was reported [16] that two forces drive the diffusion of these additives. The first one is the concentration gradient of the cations between the grain-boundary silicate glass, where the additives are initially present, and the pure silica scale. The second one is the free energy of silicate formation in the surface SiO2. Also, the mobility of additive ions is greatly enhanced at temperatures >1200℃ [17] and under the influence of the electric field within the powered GP [1][18][19][20]. According to Lee et al [10], an ytterbium silicate skin formed on Yb2O3-doped Si3N4 when it was exposed to an oxidizing air atmosphere at temperatures above 1250 ℃. Controlled oxidation at 1450 ℃ for 24 h resulted in an ytterbium silicate layer 3-4 μm thick and an Yb-depleted zone of 6  30-40 μm. The mechanisms behind this phenomenon include the outward diffusion of Yb2O3 through the grain boundaries and its reaction with silica, with the possible reactions as follows: 2Si3N4 (s) + 6O2 (g) = 6SiO2 (s) + 4N2 (g) (1.5) 6SiO2 (s) + 3Yb2O3 (s) = 3Yb2Si2O7 (s) (1.6)  1.3 Environmental barrier coatings Environmental barrier coatings (EBCs) have been studied for the purpose of protecting silicon-based ceramics from the corrosion triggered by water vapor and molten salt [21]. Among the various candidate materials for such EBC application, BSAS, mullite, rare-earth silicates and their composites attracted the most interests in the past few decades [22]. 1.3.1 Requirements for EBC There are several key factors to be taken into consideration while selecting EBC materials. First and foremost, excellent environmental durability of the candidate material is the most important criteria. The EBC should have low evaporation and recession rates in the presence of water vapor. For example, in NASA’s endeavor to develop next-generation EBCs for CMC (Ceramic Matrix Composite) airfoils and combustors, the goal is to achieve a recession rate lower than 5 mg/cm2 per 1000 h (under 40-50 atm) [23]. Secondly, mechanical compatibility between the EBC material and the substrate is also of great importance. The EBC should have a Coefficient of Thermal Expansion (CTE) close to that of the substrate to avoid thermal stresses caused by CTE mismatch which could lead to delamination or cracks on the coating. Since the CTE of silicon nitride is in the range of 3-4×10-6 K-1 [22][24][25], the ideal EBC for silicon nitride-based components should 7  have a relatively small CTE that is close to that range. Thirdly, chemical compatibility between the coating and the substrate should be assured to avoid detrimental chemical reaction or interaction. Fourthly, phase stability of the EBC in the temperature range that the substrate undergoes is also critical. For the integrity of the EBC, phase transformation that involves volumetric change should be avoided. In addition, the coating material should have low oxygen permeability to protect the silicon nitride substrate from further oxidation. Lastly, materials that can be prepared through simple and low-cost processes is favored when it comes to industrial scale production. 1.3.2 Candidate materials as EBC for silicon-based ceramics Up to now, the development of environmental barrier coatings on silicon-based ceramics has generally gone through three generations, with respective focuses on mullite/YSZ (Yttria-Stabilized Zirconia), mullite/BSAS (1-xBaO-xSrO-Al2O3-2SiO2, 0≤x≤1) and rare earth silicates [22]. Figure 1.3 presents the change of temperature capability of ceramic matrix composites (CMCs) with the three generations of EBCs. Mullite (3Al2O3∙2SiO2) was extensively studied as EBC for Si3N4 and SiC in the early days of coating development due to its low CTE (5-6×10-6 K-1), low thermal conductivity and chemical compatibility with silicon-based ceramics [22]. However, its key durability limitations [26] include through-thickness cracking, weak bonding between mullite and Si-based ceramics and a relatively high silica activity (0.3~0.4), which results in selective volatilization of silica due to its reaction with water vapor. 8   Figure 1.3 Change in the temperature capability of ceramic matrix composites with EBCs [27]  Yttria-stabilized zirconia (YSZ) top coat was added on the top of mullite to improve water vapor resistance, which made the first generation water-vapor-resistant EBC. This EBC only lasted a few hundred hours at ~1300 ℃ before the large CTE mismatch between YSZ and mullite caused severe cracking and delamination of the coating [26]. For the second generation EBCs, the YSZ top coat was replaced with BSAS (1-xBaO-xSrO-Al2O3-2SiO2, 0≤x≤1), whose lower CTE and lower elastic modulus brought superior crack resistance [28]. A silicon bond coat was also added to improve adherence. Thus, the second generation EBC was made up of a Si bond coat, a mullite or a mullite + BSAS intermediate coat, and a BSAS top coat[22][28]. There are two key durability issues[22] with this EBC: (1) the volatilization of BSAS top coat in high velocity combustion environments (a BSAS recession of ~70 µm was predicted after 1000 h at 1400 ℃, 6 atm total pressure, and 24 m/s gas velocity) [28]; (2) the chemical 9  reaction between BSAS and silica (formed by the oxidation of Si bond), which generates a relatively low-temperature melting (~1300 ℃) glass [28]. After decades of research on mullite coatings, recent research suggested that rare earth disilicates (Re2Si2O7), which have relatively low CTE, good phase stability and low recession rate in water vapor, are promising alternatives to the previous generations of EBCs. The third generation of EBC system NASA designed for SiC/CMC is made up of a Si bond coat, a mullite intermediate layer and a rare earth silicate top coat [27].  1.3.3 Advantages of rare-earth silicates as EBCs Rare earth silicates have attracted interest in the application of EBCs mainly for their excellent durability in high temperature corrosion environments and good CTE match with silicon-based ceramics. The major advantages of yttrium silicates as EBCs for Si3N4-based ceramic glow plugs can be summarized as: A. High melting points As summarized in Table 1.1, the melting point of Y2Si2O7 is 1780 ℃ and Y2SiO5 has a melting point of 1950 ℃. Such high melting points are essential for EBC materials employed in high temperature environments. B. Environmental durability Maier et al [29] predicted the corrosion rate of Y2Si2O7 in water vapor environment to be 1.92×10-3 mg/cm2∙h, which is about 1/7 of the corrosion rate of pure silica (12.7×10-3 mg/cm2∙h). Jacobson 10  [30] estimated the corrosion rate of Y2Si2O7 + Y2SiO5 to be less than 1/3 of the corrosion rate of pure silica in the presence of water vapor. And the corrosion rate for Y2SiO5 + Y2O3 was found to be 1/55 of that of pure silica [30]. Therefore, externally applied yttrium silicate coatings should significantly improve the environmental durability of silicon-based ceramics.  Table 1.1 Summary of the relevant properties of yttrium silicates and Si3N4 Properties Y2Si2O7 Y2SiO5 Si3N4 Density [g/cm3] 4.30 (α) [31]   4.03 (β) [31]   4.04 (γ) [31]   4.11 (δ) [31]   Melting point 1780 ℃ [32] 1950 ℃ [33] 1900 ℃ [34] 2060 K (δ) [35] 2232 K [35]  Average CTE [×10-6/K] 3.9 (γ) [24] 8.36 [24] 3.7 [24] 3.9 (500-1773K) [36] 5-6 [22] 3-4 [22] 8.0 (α) [37] 6.9 (473-1633K) [36] 3.4 [25] 4.1 (β) [37]   3.9 (γ) [37]   8.1 (δ) [37]   Thermal conductivity [W/m∙K] 1.35 (γ, minimum) [24] 1.13 (minimum) [24]   Polymorphs Y, α, β, γ, δ, z, ζ, η [37] X1 (low temp phase) X2 (high temp phase)  Note: The CTE of MoSi2 is 8.2×10-6/K [24].  C. Mechanical compatibility (CTE similarity) As summarized in Table 1.1, the Coefficient of Thermal Expansion of Si3N4 is generally in the range of 3-4×10-6/K. CTE is greatly influenced by the temperature and porosity (relative density) of the material. Although there is a wide variation between the CTE reported by different 11  researchers[37], the CTEs of Y2Si2O7 and Yb2Si2O7 are overall close to that of silicon nitride (see Table 1.1). D. Chemical compatibility Yttrium/Ytterbium silicates are originally present as gain boundary phases in the glow plugs (GPs) as a result of the addition of Y2O3/Yb2O3 sintering aids [18][38][39]. Given the migration of Y2O3/Yb2O3 to the GP surface and the self-formation of corresponding silicates [40], there is desirable chemical compatibility between rare earth silicate EBCs and the GPs with rare earth additives. 1.4 Y2O3-SiO2 system As summarized by Sun et al [37], there are four ternary crystalline phases, Y2Si2O7, Y2SiO5, 2Y2O3∙3SiO2 and Y4.67(SiO4)3O (2.34Y2O3∙3SiO2), that have been reported in the Y2O3-SiO2 binary system. 2Y2O3∙3SiO2 is stable only in the presence of other divalent or monovalent cations [37]. Figure 1.4 shows the phase diagram of Y2O3-SiO2 system that was calculated by CALPHAD (CALculation of PHAse Diagrams) method, using a consistent set of related thermodynamic data [35]. Only Y2Si2O7 and Y2SiO5 are present as ternary compound phases in this diagram. This phase diagram also specifies four polymorphs of the yttrium disilicate, α, β, γ and δ. Single-phase Y2Si2O7 and Y2SiO5 are only present as the two vertical lines in the diagram, meaning that obtaining the single-phase silicates requires strict stoichiometry of the Y/Si ratio. Details about the properties of the silicates, their polymorphs and phase transformation will be discussed later in this chapter. 12   Figure 1.4 The phase diagram of the calculated Y2O3-SiO2 system[35]  1.4.1 The prospects of Y2Si2O7 and Y2SiO5 as EBCs for Si3N4 When it comes to their application as EBCs for silicon nitride, both Y2Si2O7 and Y2SiO5 have advantages and disadvantages. Table 1.1 summarizes the properties of the two yttrium silicates and silicon nitride. On the one hand, Y2SiO5 outperforms Y2Si2O7 in terms of corrosion resistance and stability in water vapor environments. Similar to silica, Y2Si2O7 (equivalent to Y2O3∙2SiO2) is also subjected to degradation caused by the formation of volatile Si(OH)4 in the presence of water vapor [41], according to the reaction: Y2Si2O7(s) + 2H2O(g) = Y2SiO5(s) + Si(OH)4(g) (1.7) 13  The selective volatilization of silica in Y2O3∙2SiO2 results in a porous network of Y2SiO5.  Jocobson [30] developed a procedure for measuring silica activities in the Y2O3-SiO2 system. The silica activity in the Y2SiO5 + Y2O3 biphasic region was calculated to be [30]:  log[a(𝑆𝑖𝑂2)] = −5200.26𝑇+ 0.0567   (1532 < T(K) < 1670)  (1.8) And the modeled silica activity in the Y2Si2O7 + Y2SiO5 biphasic region is [30]: log[a(𝑆𝑖𝑂2)] =4.2252𝑇− 0.5531   (1628 < T(K) < 1747)  (1.9) Based on the above equations, at 1633 K (1360 ℃), the silica activity in the Y2SiO5 + Y2O3 biphasic area is 7.45×10-4 and that in the Y2Si2O7 + Y2SiO5 biphasic area is 0.28. It was also predicted that the flux of Si(OH)4, which corresponds to weight loss rate, for Y2Si2O7 + Y2SiO5 is 0.003 mg/cm2∙h [30]. And the flux of Si(OH)4 for Y2SiO5 + Y2O3 was predicted to be 2×10-4 mg/cm2∙h, more than one magnitude lower than that for Y2Si2O7 + Y2SiO5 [30]. On the other hand, the mechanical compatibility of Y2Si2O7 as EBC for Si3N4 is in general better than Y2SiO5 [36]. Despite the variances between the CTEs of different polymorphs, Y2Si2O7 generally has a CTE that is closer to that of Si3N4 than Y2SiO5 does. The CTEs of β-Y2Si2O7 and γ-Y2Si2O7, which are respectively 4.1×10-6 K-1 and 3.9×10-6 K-1 (see Table 1.1), especially match the CTE of silicon nitride. The CTE mismatch between Y2SiO5 and Si3N4, however, is very likely to cause the formation of cracks and even spallation when the coating is in service at high temperatures. One approach to balance the corrosion resistance and mechanical stability of yttrium silicate coating is to add a certain percentage of Y2Si2O7 to the Y2SiO5 coating [42]. In the case where the 14  silicon-based ceramic substrate is oxidized (and the silica scale not removed) prior to the fabrication of Y2SiO5 coating, the formation of Y2Si2O7 is theoretically inevitable at high temperatures. As can be speculated from the phase diagram in Figure 1.4, the excess silica will cause the composition of the coating to shift from the MS (Y2SiO5) region to the MS + DS (Y2Si2O7) region and even to the DS + SiO2 region, through the following reaction: Y2SiO5(s) + SiO2(s) = Y2Si2O7(s) (1.10) It’s reported that Y2Si2O7 and Y2SiO5 remain as the only stable solid phases after annealing the Y2SiO5/SiC mixture at 1923 K in air for 128 h [43]. In summary, the selection and combination of the two yttrium silicates remains a challenging issue in the application of yttrium silicates as EBCs for silicon-based ceramics. 1.4.2 Phase transformations of Y2Si2O7 Up to six polymorphs of Y2Si2O7 have been widely reported: y, α, β, γ, δ and z, four of which are shown in the phase diagram in Figure 1.4. ζ and η-Y2Si2O7 are two recently reported polymorphs that were discovered as by-products in the synthesis of other compounds [37]. Despite wide variation between transformation temperatures reported by different researchers, the most cited ones are published by Ito and Johnson [44]: α1225𝐶→   β1445𝐶→   γ1535𝐶→   δ Felsche [45] also reported the phase transformation temperatures between these four polymorphs: α1190±50𝐶→       𝛽1350±30𝐶→       𝛾1580±30𝐶→       𝛿 15  However, several studies have obtained results not in accordance with the above findings. The review paper by Sun et. Al [37] attributed this ambiguity to the differences of preparation methods, impurities and heat treatment history by different studies.  Table 1.2 Crystal structures of polymorphs of Y2Si2O7 Polymorph Structure Space group Lattice parameters References y-Y2S2O7 monoclinic P21/m a=7.45 Å, b=8.07 Å, c=5.03 Å,  β =111.11° Heward et al.  [46] α-Y2S2O7 triclinic 𝑃1̅ a=6.59 Å, b=6.63 Å, c=12.03 Å, α =94°, β =89°, γ=88° Kahlenberg et al. [47] β- Y2S2O7 monoclinic C2/m a=6.88 Å, b=8.97 Å, c=4.72 Å, β=101.7° Liddell et al. [48] γ- Y2S2O7 monoclinic P21/c a=4.69 Å, b=10.86 Å, c=5.59 Å, β=96.01° Leonyuk et al. [49] δ- Y2S2O7 orthorhombic Pna21 a=13.81 Å, b=5.02 Å, c=8.30 Å Dias et al. [50]  Figure 1.5 presents the crystal structures of the five polymorphs of Y2Si2O7 [15].  Since the polymorphs have different crystal structures, the phase transformations are of reconstructive character. α, β, γ and δ are the stable polymorphs of Y2Si2O7. The y-phase, also referred to as yttrialite, has often been described as an impurity-stabilized polymorph of yttrium disilicate [46]. According to Ito and Johnson [44], this yttrialite low form is stable up to 1200 ℃ and its possible stabilizers include cations such as H+, Na+, Mg2+, Mn2+, Fe2+, Fe3+, Al3+, Th4+, or Zr4+. According to the study by Alba et al [51], the stability of y-Y2Si2O7 is higher than that of α-Y2Si2O7 at 1100 ℃ and lower than that of β-Y2Si2O7 at 1300 ℃ for sol-gel-derived yttrium disilicate powders, when the pH of the precursor sol is lower than 5. At neutral pH, the obtained yttrium disilicate is reported to consist of 79.1wt% α-Y2Si2O7 and 20.9wt% y-Y2Si2O7 after annealing at 1100 ℃. After the 16  heat treatment at 1300 ℃, 4.27wt% of y-Y2Si2O7 was detected via XRD analysis, alone with β-Y2Si2O7, the stable phase in this temperature range. Therefore, the stability of y-Y2Si2O7 is greatly influenced by pH and the synthesis method[51].             Figure 1.5 Crystal structures of: (a) y-Y2Si2O7, (b) α-Y2Si2O7, (c) β-Y2Si2O7, (d) γ-Y2Si2O7, (e) δ-Y2Si2O7 according to the parameters listed in Table 1.2 [15]  17  1.4.3 Synthesis of yttrium silicates The synthesis and sintering of single-phase yttrium silicates has remained a challenge in related research areas. Based on the phase diagram, strict control of the stoichiometric Y/Si ratio is required to get a single compound. And appropriate temperature is indispensable to avoid the presence of more than one polymorphs.  As summarized by Sun et al1 [37], the common synthesis methods for yttrium silicate bulk materials include: sol-gel, hydrothermal, solid-state reaction, solid-liquid reaction, molten salt fluxing growth, high-pressure/high-temperature synthesis, etc. For the processing of yttrium silicate coatings/films, a number of methods have been investigated, including slurry dip coating, plasma spraying, electrophoretic deposition and CVD [37].  1.4.4 Recent research on yttrium/ytterbium silicates as EBCs Lee [10] reported the development of self-forming ytterbium silicate skin on silicon nitride (type SN362, Kyocera, Inc., Kyoto, Japan) by controlled oxidation. Yb2Si2O7 was formed as an oxidation product on the surface of silicon nitride when it was exposed to air at temperatures above 1250 ℃. Oxidation at 1450 ℃ for 24 h resulted in a dense and uniform ytterbium silicate scale with the thickness of 3-4 µm, with elongated grains of 16 µm in length and 2.5 µm in diameter. It was indicated that Yb2O3 diffused to the surface and reacted with silica during the oxidation process.  Wagner et al [43] prepared yttrium silicate layer on chemical vapor deposition (CVD) SiC-precoated C/C–SiC via low pressure plasma spray (LPPS). The thickness of the yttrium silicate layer, which consisted of Y2SiO5 and SiO2, was in the range of 100-200 µm. Plasma wind tunnel tests between 1623 and 1923 K showed the LPPS yttrium silicate coating provided good oxidation 18  and erosion protection for the substrate. The relative mass loss of the coating system was less than 0.6%. But a sudden coating failure, caused by the formation of blisters at the coating/substrate interface, was observed at temperatures above 1923 K. And the occurrence of blisters was attributed to the gaseous phase (mainly CO) resulting from the reactions between Y2SiO5 and SiC. Therefore, the interface stability is of vital importance to rare earth silicate coatings in the EBC application. García et al [41] used thermal spraying to deposit yttrium silicate (Y2Si2O7+Y2SiO5) coating on Si-precoated SiC coupon. Isothermal heat treatment at 1350 ℃ under a water vapor flux of ~4 mL/min in a box furnace for 12 h caused the formation of ridges on the surface of the coating. And the presence of the micro-ridged Y2SiO5 zones was attributed to the selective volatilization of the SiO2 from the less stable Y2Si2O7 phase. The Y2SiO5 phase, which remained stable under the attacks from water vapor, was suggested to be more suitable for the EBC application in the high-temperature water vapor environments. Wang et al [52] used slurry spraying to in-situ fabricate a multilayer γ-Y2Si2O7 EBC on porous Si3N4 substrate. The coating system was made up of a Y-Si-Al-O glass top layer, the γ-Y2Si2O7 intermediate layer and the bottom bond layer. The bond layer was formed by the infiltration of the Y-Si-Al-O liquid into the porous Si3N4 through the γ-Y2Si2O7 interlayer. Sintering at temperatures above 1400 ℃ transformed the coating into two layers by completely infiltrating the Y-Si-Al-O liquid into γ-Y2Si2O7. The coating sintered at 1450 ℃ for 1 h in N2 atmosphere demonstrated excellent thermal shock resistance, surviving the water-bath quenching from 1000 ℃ to room temperature without forming any defects. The bond coat with a thickness of ~200 µm created a 19  strong interface bonding between the γ-Y2Si2O7 layer and the substrate, which greatly contributed to the strength of the coating during thermal shock test. 1.5 Sol-gel dip coating Among the common preparation methods for yttrium silicate coatings, the use of CVD is limited by its high cost. Plasma spray, with its high-velocity droplets, could generate damage on the substrate through mechanical erosion [52]. Sol-gel dip coating is advantageous for its easy composition control through precursors, low synthesis temperature, availability of reaction apparatus and suitability for the coating applications. 1.5.1 Sol-gel process  The typical stages in sol-gel processing include hydrolysis, condensation, gelation, gel drying and further heat treatment [53]. Hydrolysis and condensation, which can be accelerated through the addition of acidic or basic catalysts, result in the gradual formation and linkage of colloidal particles. Coating preparation is carried out before the sol-gel transition of the sol, where the viscosity of the sol increases sharply. The control of sol viscosity is critical to achieve satisfactory coatings.  1.5.2 Dip coating  Dip coating basically consists of five stages: immersion, start-up, deposition, drainage and evaporation. Apart from the composition and properties of the sol used, several operation parameters of dip coating process are critical to the tailoring of coating properties. For example, 20  the withdrawal speed of the substrate is one of the important factors that control the coating/film thickness [54][55]. The control of coating thickness is vitally important for EBC application. Coating with a thickness of less than 1 µm is not desirable as it may not be sufficiently dense to be protective. But coating thicker than 5 µm is undesirable as well. High thickness comes with high thermal stresses, which results in the formation of cracks. And thick coating acts as a thermal barrier, preventing heat flowing from the heating element. 1.5.2.1 The correlation between coating thickness and withdrawal speeds Based on the range of withdrawal speed U0 and liquid viscosity η, a few models have been proposed to predict dip-coated film thickness, which have been summarized by Brinker [54] and shown in Table 1.3.  Figure 1.6 Schematic illustration of the dip-coating method and both capillarity and draining regimes involved at low and fast withdrawal speeds, respectively [55] 21  Table 1.3 Models for the prediction of dip-coated film thickness [54] No. Scope of application Model Mechanism 1 η and U0 are high enough to lower the curvature of the gravitational meniscus ℎ0 = 𝑐1(𝜂𝑈0𝜌𝑔)12 (1.11) The viscous drag (ηU0) and gravity force (ρg) reach a balance 2 η is low and U0 is in the typical range for sol-gel film deposition (~1-10 mm/s); for a Newtonian and non-evaporating liquid ℎ0 = 0.94(𝜂𝑈0)𝛾𝐿𝑉16(𝜌𝑔)1223 (1.12) The effect of sol drainage is in charge 3 U0 is ultra-slow (0.01 mm/s to ~0.1 mm/s) that the “capillarity regime” is in charge and modeled by semi-experimental equations ℎ𝑓 =𝑐𝑖𝑀𝑖𝐸𝛼𝑖𝜌𝑖𝐿𝑈0 (1.13) The solvent evaporation is faster than the movement of the drying line, which results in the capillary rise of the precursor sol 4 U0 is of intermediate values (~0.1-1 mm/s), both regimes (“draining” and “capillarity”) are taken into account ℎ𝑓 = 𝑘𝑖 (𝐸𝐿𝑈0+ 𝐷𝑈023) (1.14)   Where, c1 = constant (about 0.8 for Newtonian liquids) D = global constant that combines the physiochemical constants of the precursor solution 22  (D is calculated from experimental data and found to be roughly constant if U0 is in the range of 1 to ~10 mm/s) And he meanings for all the other symbols used in the equations in Table 1.6 are listed in List of Symbols. Model No.2 is the classic Landau-Levich model [56], which often describes the entrained thickness of dip-coated films relatively well when draining is the dominant effect during substrate withdrawal. However, recent study revealed that this model doesn’t work in the case of ultra-slow withdrawal speeds [55]. It was found that the film thickness reaches a minimum value at certain withdrawal rate. To describe the relationship of dip-coated film thickness vs. withdrawal speed at such conditions, the “capillarity regime” (shown in Figure 1.6) was introduced and experimental data was used to derive the semi-experimental models [55] (Model No.2 and No.3) in Table 1.3. It was proposed that the presence of two opposite film thickness evolution regimes, i.e. “draining regime” and “capillarity regime”, enables a good control of the thickness when using the same precursor solution [54]. Further discussion on the effects of withdrawal speeds and the comparison between the two regimes will unfold along with the experimental investigation by this study in Chapter 6. 23  Chapter 2: Scope and Objectives  2.1 Scope of the research In this research project, rare earth silicates have been employed as Environmental Barrier Coatings (EBC) for the corrosion protection of Si3N4-based glow plugs (GPs) in high temperature combustion environments. This selection was made due to their excellent environmental durability, low Coefficients of Thermal Expansion (CTEs), nearly matching that of Si3N4, and excellent chemical compatibility. Y2Si2O7 and Yb2Si2O7 were respectively chosen as the candidate materials for such EBC application on two types of commercially available GPs, which have either Y2O3 or Yb2O3 as sintering additive. Due to the limited time frame of this project and the lower cost of the Y2O3-doped glow plug, however, most work has been focused on yttrium disilicate coatings. The substrate used in this study, the ceramic pin of the Le-Mark Si3N4-based glow plug, is a dynamic system that undergoes continuous change when the GP is powered. Therefore, it is necessary to investigate the as-received GP as well as its behaviors under the influence of electrical field. In the path to develop Y2Si2O7 EBCs on Le-Mark GP, the understanding and control of sol preparation process and dip coating procedure have been proven to be critical. Specifically, the aging conditions of the precursor solution (sol) and the withdrawal rate of the substrate are the two controllable parameters that have significant impacts on the properties of the sol-gel-derived coatings. 24  One of the challenging issues for the sol-gel processing of coatings is the formation of cracks caused by volumetric shrinkage when a large amount solvent is removed during thermal treatment. The approach taken to address this problem was to successively apply and process few micrometers-thin layers of coating, which not only avoids the formation of cracks but also allows for the tailoring of coating thickness. The natural gas burner rig and the high concentration water vapor test are two testing methods that have been employed to evaluate the performance of the Y2Si2O7-coated glow plugs, especially their corrosion resistance, in harsh environments. The real NG engine tests of the coated GP are planned by Westport (industrial sponsor of the program) at a later stage. The phase compositions of the obtained materials are characterized by the XRD analysis of the monolithic powder samples. The surface morphology and microstructure of the coatings are examined under SEM. And EDS are used to investigate the elemental distribution across the ceramic pins of the coated and uncoated GP specimens.  2.2 Objectives of the research This research aims at developing rare earth silicates as Environmental Barrier Coatings (EBCs) for the protection of commercial glow plugs in high temperature combustion environments. During different stages of this research project, the various objectives can be specified as: 1. Establish a sol-gel dip coating route to develop multi-layer Y2Si2O7 coatings as EBCs for Le-Mark Si3N4-based glow plugs. 25  2. Study the as-received glow plug, its response to electrical field and oxidation behaviors in ambient environment. 3. Investigate the effects of the aging conditions (time and temperature) of the sol on the properties of the sol-gel-derived Y2Si2O7 coating. 4. Investigate the correlation between the withdrawal speed of the GP substrate during dip-coating procedure and the properties of Y2Si2O7 coating, especially its thickness. 5. Determine the coating thickness evolution of the multi-layer Y2Si2O7 coating at different distances across the ceramic pin of the GP, with the increase of the number of layers. 6. Evaluate the performance of the coated and uncoated GP specimens in different environments, including natural gas burner rig and high concentration water vapor atmosphere. 7. Explore the degradation mechanism of the applied Y2Si2O7 coatings during the performed tests.       26  Chapter 3: Experimental Procedures  3.1 Sol preparation The following chemicals have been used as the starting materials: Y(NO3)3∙6H2O (99.9% by Alfa Aesar, 99.8% by Sigma-Aldrich), TEOS (Tetraethyl orthosilicate, 98%, Sigma-Aldrich), Hydrochloric acid (1.0 N, VWR analytical), Ethanol (absolute).  Figure 3.1 presents a flow chart that summarizes the steps of the preparation of precursor solution (sol), dip coating process and further heat treatment for yttrium disilicate coatings. To start sol preparation, Y(NO3)3∙6H2O and ethanol of the molar ratio as listed in Table 3.1 were added to a sealed glass container (reactor). The mixture was stirred at room temperature (~25 ℃) for 1 h and at ~75 ℃ on hot plate for 3 h. TEOS was added over the above solution with the final TEOS:Y(NO3)3 molar ratio equivalent to 1:1. The obtained mixture was stirred at room temperature for 1 h. A small amount of hydrochloric acid (~2% of the weight of ethanol [57]) was added to the above mixture. A set of different sols were prepared via further aging processes whose conditions are as specified in Table 3.1.  The equipment used included KA C-MAG HS7 digital hot plate.  Figure 3.1 Synthesis steps for the fabrication of yttrium disilicate coatings Note: temp = temperature 27  Table 3.1 Summary of the sols prepared and their synthesis conditions Sol YNH : ethanol YNH : TEOS Sol aging conditions YRT24h 1: 15 1: 1 Stirring at RT for 24 h Y75C8h 1: 15 1: 1 Stirring at RT for 40 h and 75 ℃ for 8 h Y75C16h 1: 15 1: 1 Stirring at RT for 32 h and 75 ℃ for 16 h Y75C16h-2weeks 1: 15 1: 1 Stirring at RT for 40 h and 75 ℃ for 8 h & Kept still at RT for 2 weeks Notes: RT = Room Temperature (approximately 25±5 ℃, depending on the season that the experiments were carried out); YNH = Y(NO3)3∙6H2O  3.2 Preparation of the substrate Figure 3.2 presents the glow plug, supplied by Le-Mark, whose ceramic pin was used as substrate for the coating: Le-Mark all-ceramic glow plugs (Type: QTJ5-12) Chongqing Le-Mark Ceramic Technology Co., Chongqing, China  Figure 3.2 Le-Mark QTJ5-12 glow plug Since only the ceramic pin (shown in Figure 3.2) of the glow plug is of interests to this study, unless otherwise indicated, the GP discussed in this thesis only refers to its exposed Si3N4-based ceramic pin. For example, the GP tip equals to the tip of the ceramic pin as demonstrated by Figure 3.2. And the GP surface is equivalently referred to as the surface of the ceramic pin. ceramic pin tip of the ceramic pin 28  3.2.1 Surface cleaning Before dip coating, the ceramic pins of glow plugs were rinsed with distilled water and absolute acetone respectively for 15 s to remove dirt and to degrease. 3.2.2 Surface treatment by sanding The GP specimens investigated in Chapter 5 and Chapter 7 were sanded with ultra-fine sandpaper (CAMI Grit designation number = 800) before dip coating. This was aimed to remove the previous coating on the GPs so that they can be recycled and reused for further coating application. As specified in Table 3.2, the glow plugs used in all the other experiments were used in as-received state without sanding or any other surface treatments. 3.2.3 Electrical characterization of the glow plug Due to the minor internal variances in GP geometry due to variability of the manufacturing process, there are slight differences in currents and temperatures of the same type of glow plugs even under the same voltage. Therefore, all the glow plug specimens were connected to Direct Current power supply ( Xantrex DC power supply HPD 30-10 )  and electrically characterized at certain voltages (typically 10 to 12 V) prior to the dip coating procedure. The currents and surface temperatures of the ceramic pins in GP were recorded. Pyrometer ( Ircon UX-20P infrared thermometer ) was used for the measurement of the surface temperature of the hot zone (the area with the highest temperature) on the ceramic pin. 29  3.3 Dip coating 3.3.1 Immersion and withdrawal of the glow plug During dip coating process, Instron 3369 mechanical testing system ( Instron 3369 Dual Column Tabletop Testing System ) was used to control the immersion and withdrawal of the glow plug. As shown in Figure 3.3, the glow plug was gripped by the Instron tester and vertically dipped into the aforementioned precursor solution (sol). Only the first 10-15 mm of the ceramic pin was immersed in the sol. The residence time of the pin in the sol was kept at nearly zero for all specimens. Then the glow plug was vertically pulled upward and withdrawn from the sol at a certain rate, which was controlled by the Instron tester. The withdrawal rate was the major parameter to be investigated during this dip coating process.   Figure 3.3 Experimental setup for dip coating with the assistance of Instron testing system  Instron 3369 glow plug sol 30  3.3.2 In-situ heat treatment of the applied coating The coated glow plugs were placed vertically (with the ceramic pin pointing to the ground) at room temperature for 1 h. Then the specimens were dried in the oven dryer at 70-80 ℃ for 1 h ( Lab-Line oven, Xantrex DC power supply HPD 30-10 ). Finally, the coatings on the GPs were in-situ sintered by applying power via the experimental setup as shown in Figure 3.4. The GP was first powered at 2 V and 3V for 1 min. And the increasing of the voltage from 3 to 12 V followed a rate of 2 V/min. The in-situ sintering of each layer of the coatings was carried out at 12 V for 1.5 h.   Figure 3.4 Experimental setup for in-situ sintering via direct current power supply  3.3.3 Fabrication of multi-layer coatings Multi-layer coatings were prepared by repeating the dip coating, drying and in-situ sintering processes explained in Section 3.3.1 and 3.3.2 for each layer of the coatings. For instance, the 2nd layer was applied after the 1st layer was dried and in-situ sintered. Then the 3rd layer was built upon ceramic pin of glow plug DC power supply 31  the 2nd one after it went through the same drying and in-situ sintering steps. And further layers can be fabricated successively as such. 3.3.4 Summary of the processing conditions of GP specimens During different stages of this research project, different processing conditions have been used to prepare coated and uncoated GP specimens. The specimens that will be discussed in the following chapters and their processing conditions, including the type of sols, the withdrawal speeds during dip-coating and the numbers of fabricated layers of the coatings, are as summarized in Table 3.2. Table 3.2 Summary of the processing conditions of GP specimens GP Chapter Sol Surface treatment Withdrawal speed Number of layers Heat treatment Test as specified 4 N/A N/A N/A N/A 12 V for 100 h N/A as specified 5 varies sanding 20 mm/min 1 Drying at 70-80 ℃ N/A #5~#500 6 Y75C16h N/A varies 1 12 V for 1.5 h N/A 1L~6L 7 Y75C16h-2weeks sanding 20 mm/min varies 12 V for 1.5 h per layer N/A BRGP0 8 N/A N/A N/A N/A N/A BR BRGP1 8 YRT24h N/A 20 mm/min 3 12 V for 1.5 h per layer BR HCWVGP0 8 N/A N/A N/A N/A N/A HCWV HCWVGP1 8 Y75C8h N/A 20 mm/min 3 12 V for 1.5 h per layer HCWV   32  3.4 Synthesis of monolithic samples Monolithic powder samples were prepared for XRD characterization. In order to accelerate the gelation process, the obtained sol from Section 3.1 was placed at ~40 ℃ under ambient atmosphere until the gel was formed. The container of the sol was covered by parafilm on which 6-10 holes had been pierced by toothpick. Then the gel was dried at 75 ℃ for 24 h and 200 ℃ for 1 h. After the dried material was grinded into fine powders, the calcination was carried out at different temperatures (1200-1300 ℃) in furnace with a heating rate of 10 ℃/min. Equipment used: Lab-Line oven, Eciton furnace, Micropyretics Heaters International box furnace M18-40 with BPAN-O-PLUS control panel 3.5 Sample characterization A variety of characterization methods have been employed to characterize the morphology, microstructure, phase composition, elemental distribution and coating thickness of the specimens. 3.5.1 SEM and EDS Hitachi S-3000N Variable-Pressure Scanning Electron Microscope (SEM) was used to observe the surface morphology and microstructure of the coating, as well as identifying the coating thickness from the cross-sectioned samples. The SEM was operated under low vacuum mode, with a variable pressure in the range of 20-40 Pa. In variable pressure mode, SEM was capable of imaging of the insulating glow plugs specimens, which were characterized without conductive coatings. 33  Energy Dispersive X-ray Spectroscopy (EDS) was conducted with SEM through Advanced Analysis Technologies silicon lithium detector, which had a resolution of 133 eV at Mn Kα.  The operating voltage was 20 kV and working distance was set at 15 mm. EDS was used in this study as a semi-quantitative microanalysis method for the determination of elemental compositions of both the substrate and the coatings. It is therefore necessary to discuss the accuracy and precision of EDS in such application. The Electron Probe Microanalysis (EPMA) has a long history in the quantitative analysis of the chemical composition of solid materials. There are two X-ray intensity measurement techniques in EPMA, Wavelength Dispersive X-ray Spectroscopy (WDS) and Energy Dispersive X-ray Spectroscopy (EDS). The former technique is generally acknowledged as a more precise method with lower detection limits (CDL). Nevertheless, it was demonstrated by Ritchie [58] that EDS with a silicon drift detector can match the accuracy and precision of WDS, particularly in the detection of major (>0.10 mass fraction) and minor (>0.01 mass fraction) elements, by following the “k-ratio” protocol. The “k-ratio” measurement protocol was first proposed by Castaing[59] for the quantitative microanalysis via WDS, 𝑘 =  𝐼𝑠𝑝𝑒𝑐𝐼𝑠𝑡𝑑 (3.1) where Ispec and Istd respectively stand for the intensities of a specific characteristic X-ray measured for the specimen (“spec”) and a standard (“std”). Through continuous contributions by later researchers, the formula for converting the k-ratios into mass concentrations now has the form [60][61], 34  𝐶𝑠𝑝𝑒𝑐𝐶𝑠𝑡𝑑= 𝑘𝑍𝐴𝐹𝑐 (3.2) where Cspec and Cstd are the mass concentrations of the element of interest measured in the specimen and the standard, Z represents the “atomic number correction”, A stands for the “absorption correction”, F and c are the “secondary fluorescence corrections” for characteristic (F) and continuum (c) X-rays. The standardless EDS analysis, however, is subject to broader relative errors than the standards-based analysis following the k-ratio/matrix correction protocol by a factor as high as five [60]. The coating system (including the substrate) investigated by EDS in this study is a complex system consisting of five or more elements, making the matrix correction quite difficult and complicated. Due to this characterization dilemma, the EDS results presented in this work were acquired by the standardless analysis and should be handled with caution. Thus, this work does not use the obtained EDS data to quantify the chemical formula of the specimen but to study the evolution trend of certain elements, which should allow for a more generous error budget. In addition, the two element of most interest to this study, Yttrium and Molybdenum, are both heavy elements that have higher detection precision by EDS than light elements present in the investigated materials, in particular nitrogen (which was thus not quantified in the EDS measurements). Therefore, it is expected that the introduction of Yttrium and Molybdenum as semi-quantitative indicators should reveal useful information about the coating properties (especially its thickness), as will be discussed in the following chapters. 35  3.5.2 XRD X-ray Diffraction (XRD) was used for phase identification of the monolithic powder samples, on Rigaku MultiFlex XRD operating at 40kV, Cu Kα radiation. Phase identification was carried out by Match! (CRYSTAL IMPACT GbR, Bonn, Germany), which is a software for phase identification from powder diffraction data.  3.6 Tests on the performances of the as-received and coated Glow Plugs Natural gas burner rig test and high concentration water vapor test were employed to evaluate the performance of coated and uncoated GP specimens. 3.6.1 Natural gas burner rig test The natural gas burner rig (BR) facility, which is made up of 2 Xantrex XHR 20-50 power supplies, was built in-house at Westport Fuel Systems (Vancouver, Canada). The air and fuel were pre-mixed at the base of the Fisher Blast Burner. And the gas mixture was controlled by the pressure regulators and rotameter flow meters [18]. BR testing conditions used in this study are: Volumetric air/fuel = 10, Voltage = 12 V, Current = 4 A, Estimated temperature = ~1200 ℃. 36   Figure 3.5 Experimental setup for natural gas burner rig test 3.6.2 High concentration water vapor test This test was carried out by placing glow plugs in water vapors generated by household humidifier, believed to provide close to 100% relative humidity at room temperature. The GPs were powered via direct current with the voltage of 12 V. The initial temperature read on the GP surface close to tip was in the range of 1220-1250 °C.  Figure 3.6 Experimental setup for high concentration water vapor test glow plug humidifier glow plugs 37  Chapter 4: Self-formed Y2Si2O7 on Le-Mark Glow Plugs  4.1 Introduction The substrate for the environmental barrier coating (EBC) in this study is the ceramic pin of the commercially available Le-Mark QTJ5-12 glow plug (GP) as shown in Figure 3.2, which possesses the approximate chemical composition as listed in Table 4.1. This ceramic pin is made up of a silicon nitride body with Y2O3 as sintering additive, and MoSi2 providing electrical conductivity of the plug. And it’s also a dynamic system that is subject to continuous change when the GP is in service. Therefore, the first part of this research focused on the investigation of the as-received GP, its response to electrical field and the self-formation of yttrium silicates on its surface through oxidation of the plug material in air. 4.2 As-received Le-Mark GP Figure 4.1 shows the surface of a Le-Mark QTJ5-12 glow plug under SEM. The EDS characterization revealed that the brighter regions are rich in molybdenum while the darker areas are silicon-rich. EDS results, as demonstrated in Table 4.1, indicate the EDS-analyzed area (same as Figure 4.1) of this as-received Le-Mark glow plug consists of 38wt% O, 34wt% Si, 22wt% Mo, 3wt% Y and 3wt% Al. The nitrogen content of the substrate was not included in this chapter, as well as in the EDS analysis performed on the specimens in Chapter 6, 7 and 8. The detection of nitrogen from the surface of the glow plug is difficult since the self-formed silica scale on the GP surface prevents N from being detected. It’s also due to the light atomic mass of nitrogen, which makes it harder to be detected by the EDS detector than heavy elements like Y and Mo from the substrate. 38  Molybdenum is a relatively heavy element that can be easily detected by EDS even when a thin layer of EBC is externally applied on the ceramic pin of the GP. Given the penetration depth of the electron beam used for the EDS system, EBC or self-formed surface scale with a thickness of less than ~5 µm is unlikely to be able to entirely prevent the substrate Mo from being detected by EDS. Generally speaking, the lower the Mo content detected by EDS on the surface of GP’s ceramic pin, the thicker the applied EBC or the self-formed scale. Therefore, the elemental content of Mo will be conveniently used as a semi-quantitative indicator for the thickness of the self-formed silica scale and the applied EBC on this type of GP, as previously mentioned in Section 3.5.1.  Figure 4.1 Micrograph of the surface of the ceramic pin of an as-received Le-Mark QTJ5-12 glow plug Despite the fact that the molybdenum and yttrium contents of different glow plugs may vary, the Mo contents detected by EDS at the ceramic pins of as-received Le-Mark GPs are generally in the range of 15-25wt%. 4.3 Le-Mark GP after the electrical characterization The micrographs in Figure 4.2 present the surface of a Le-Mark glow plug after electrical characterization at 12 V (~1260 ℃) for 10 min. During the electrical characterization, the glow 50 µm 39  plug was connected to direct current power supply and powered at 12 V for 5-10 min so that its surface temperature could be measured before external coating was applied on the surface.     Figure 4.2 Micrographs of the surface of Le-Mark glow plug after electrical characterization at 12V (~1260 ℃) for 10 min: (a) 2mm from tip; (b) 2mm from tip; (c) 4mm from tip; (d) 6mm from tip  As can be seen on the higher magnification micrograph (Figure 4.2b), some gray acicular crystals appeared on the surface at 2 mm from the GP tip. These crystals are the self-formed yttrium silicates that resulted from the reaction between the Y2O3 additive and the surface SiO2 scale under high temperature, which will be discussed more later in this chapter. EDS analysis was performed on this GP and the results are as demonstrated in Table 4.1. According to the EDS analysis performed on the same area as shown in Figure 4.2a, the molybdenum and yttrium contents are 17wt% and 3wt%, respectively, at 2 mm from the GP tip. And 18wt% of molybdenum and 3wt% of yttrium were detected on the area shown in Figure 4.2c, which was located at 4 mm from the B A a b 200 µm 50 µm 200 µm 200 µm c d 40  GP tip. Spot analysis on the bright area (spot A) indicated its Mo content was much higher than the Mo content from the whole area analysis. Thus, it’s confirmed that the composition of the irregular bright areas is MoSi2, which is a major additive used in this type of GP. And spot analysis on the dark area (spot B) indicated Y content was slightly higher than areal average, which should be caused by the migration of Y3+ to the GP surface under the influence of electrical field [1][18]. Therefore, the self-formation of yttrium silicates on GP surface is triggered and accelerated even under a short period of electrical power supply. The SEM and EDS results of this electrically characterized GP will be used as a baseline to compare to the GP specimens with externally applied coatings, which will be discussed in the following chapters.  Table 4.1 Composition of as-received Le-Mark glow plug (by EDS) after electrical characterization at  ~1260 ℃ for 10 min Content (wt%) Location O Na Mg Al Si K Ca Fe Y Mo as-received GP 38   3.0 34    2.6 22 IS10min-2 mm 46   4.1 29    3.1 17 IS10min-2 mm-A 43   2.4 25    1.0 29 IS10min-2 mm-B 49 0.7 0.6 7.0 27 0.5 2.1 3.7 4.7 4.9 IS10min-4 mm 46 0.4  3.8 30    2.8 18 Notes:  1. IS stands for in-situ heat treatment. 2. The EDS results presented in the above table have been rounded to 1-2 significant figures in order to better represent the accuracy and precision of the EDS system. Values greater than 1wt% have been rounded to 2 significant figures and those smaller than 1wt% have been rounded to 1 significant figure. As such, the sum of relative weight percentages in one row doesn’t always add up to exactly 100wt%. Nitrogen content was not determined. This rounding rule applies to all the EDS results presented in the body of this thesis.  3. The unrounded raw data are attached in Appendices.  41  4.4 Self-formed yttrium silicates on Le-Mark GP Figure 4.3 shows the micrographs of the surface of a Le-Mark glow plug after in-situ heat treatment at 12 V (~1260 ℃) for 24 h. As can be seen from the micrographs, the surface of the ceramic pin was covered with bright needle-shaped yttrium silicate crystals at up to 8 mm from the GP tip. The irregular bright areas as shown in Figure 4.1 and 4.2, which have been confirmed to be MoSi2, became harder to distinguish due to the enhanced formation of those needle-shaped crystals after 24 h of in-situ heat treatment. According to these micrographs, it can also be seen from the surface morphology that the distribution of self-formed yttrium silicate was not uniform. Table 4.2 provides a summary of the EDS analysis performed at the areas shown by Figure 4.3. Compared with the results from the as-received GP, the molybdenum content significantly decreased while the yttrium content increased at all distances of the GP surface after 24 h of in-situ heat treatment. The change of elemental composition was also due to the migration of Y3+ to the GP surface and the self-formation of yttrium silicate. Similar to Lee’s research [10] about the self-formed Yb2Si2O7 on Si3N4-based substrate, these needle-shaped yttrium silicate crystals were formed as oxidation products during the high temperature treatment. The oxidation process involves the diffusion of Y2O3, which was used as sintering additive for Le-Mark glow plugs, to the surface and its reaction with the silica scale spontaneously formed on Si3N4 surface [40]. The possible reactions are: 2Si3N4 (s) + 6O2 (g) = 6SiO2 (s) + 4N2 (g)  (4.1) 6SiO2 (s) + 3Y2O3 (s) = 3Y2Si2O7 (s)  (4.2)  (a) (b) 42         Figure 4.3 Micrographs of the surface of Le-Mark glow plug after in-situ heat treatment at 12 V (~1260 ℃) for 24 h:  (a)(b) 2 mm from tip; (c)(d) 4 mm from tip; (e)(f) 8 mm from tip  Table 4.2 EDS results of as-received Le-Mark glow plug after in-situ heat treatment at ~ 1260 ℃ for 24 h Content (wt%) Location O Al Si Y Mo As-received GP 38 3.0 34 2.6 22 IS24h-2 mm 50 3.7 31 6.2 9.3 IS24h-4 mm 46 4.2 32 9.1 8.0 IS24h-8 mm 49 2.9 32 3.1 13 c d e f 200 µm 50 µm 200 µm 50 µm 200 µm 50 µm a b 43  The phase diagram of the calculated Y2O3-SiO2 system is shown in Figure 1.3. Since the surface silica is of a much larger quantity than the Y2O3 diffused to the GP surface, the right bottom region of the phase diagram should be of interest here. When the temperature is in the range of ~1200-1400 ℃, the stable substances that remain in this binary system should be α-Y2Si2O7 and tridymite SiO2. Therefore, the acicular yttrium silicate crystals that were produced by the reactions between Y2O3 and SiO2 should be Y2Si2O7 rather than Y2SiO5. The self-formed yttrium disilicate crystals at 4 mm from GP tip, as shown in Figure 4.3c and 4.3d, are of the largest size compared to those at other distances. The EDS results also indicate the highest yttrium content at 4 mm. This is due to the fact that the hot zone of Le-Mark glow plug, where the temperature is of the highest cross the ceramic pin, lies at ~3-4 mm from tip. The high temperature at hot zone facilitated the migration of yttrium cations to the surface and their reactions with silica, which resulted in the promoted formation and growth of those needle-shaped crystals. Figure 4.4 shows the micrographs of the surface of the same Le-Mark glow plug as shown by Figure 4.3 after in-situ heat treatment at 12 V (~1260 ℃) for 100 h. The bright MoSi2 that used to present on the GP surface (see Figure 4.1-4.3) was no longer visible under SEM. The surface of the ceramic pin was covered with the mixture of silica deposit, which appeared to be dark on the SEM images, and the bright acicular yttrium disilicate. It became more obvious that the self-formed yttrium disilicate crystals were of the largest size and quantity at ~4 mm from the GP tip.  Table 4.3 summarizes the results of the EDS analysis performed at the areas shown by Figure 4.4. Comparing the EDS results in Table 4.1 and Table 4.2, yttrium contents kept increasing as a result of longer in-situ heat treatment. Compared with in-situ heat treatment for 24 h, the yttrium weight 44  percentages increased significantly at 2 mm, 4 mm, and 8 mm from GP tip. Meanwhile, Mo contents decreased at all those distances across the glow plug.        Figure 4.4 Micrographs of the surface of Le-Mark glow plug after in-situ heat treatment at 12 V (~1260 ℃) for 100 h:  (a)(b) 2 mm from tip; (c)(d) 4 mm from tip; (e)(f) 8 mm from tip    e a b c d 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm f 45   Table 4.3 EDS results of as-received Le-Mark glow plug after in-situ sintering at ~1260 ℃ for 100 h Content (wt%) Location O Al Si Y Mo IS100h-2 mm 50 4.3 32 8.7 4.8 IS100h-4 mm 47 5.3 33 12 2.2 IS100h-8 mm 51 3.8 34 5.7 5.0  4.5 Discussion and conclusion Despite the continuous self-formation of yttrium silicate on the ceramic pin of the glow plug when the GP is powered, it is not feasible to use this self-grown scale as environmental barrier coating for corrosion protection. On one hand, the reactions did not yield pure Y2Si2O7 but the mixture of yttrium silicate and silica, which remains vulnerable in the presence of water vapor at elevated temperatures.  On the other hand, the continuous migration of Y3+ cations to the GP surface will result in an yttrium-depleted layer under the surface. According to the research [10] performed on Yb2O3-doped Si3N4, a self-formed ytterbium silicate layer of 3-4 μm thick brought about an Yb-depleted zone of 30-40 μm. Likewise, the deficiency of Y3+, which is a component of the glassy grain boundary phase of the Le-Mark GP, in the depletion layer may cause structural deterioration of the silicon nitride body. Therefore, it has been determined that externally applied environmental barrier coatings that are corrosion resistant and compatible with the substrates are necessary for the protection of silicon nitride-based GPs.  46  Chapter 5: Sol-gel Processing of Y2Si2O7 Coatings and Bulk Materials  5.1 Introduction As illustrated by the flow chart in Figure 3.1, the sol-gel processing of Y2Si2O7 coatings on Le-Mark glow plugs (GPs) consists of the two main stages: the preparation of the precursor solution (sol) and the fabrication of yttrium disilicate coatings. The preparation of the sol can be divided into mixing of the precursors and the aging stage. The fabrication of coatings includes dip-coating, drying and in-situ sintering (i.e. by using surface temperature of powered GP). Y2Si2O7 monolithic samples were also synthesized through the preparation procedure as specified in Section 3.4. 5.2 A Speculation on the reaction mechanisms of the sol-gel processed Y2Si2O7 The preparation of the sol starts by adding Y(NO3)3∙6H2O to ethanol with the molar ratio of 1:15. During dissolution (assisted by stirring the mixture at room temperature and 75℃), Y(NO3)3∙6H2O loses its water of crystallization and dissolves into ethanol at the elevated mixing temperature. After adding TEOS to the above solution at 1:1 molar ratio to Y(NO3)3∙6H2O, TEOS starts reacting with the water of hydration released from Y(NO3)3∙6H2O. The hydrolysis and condensation reactions can be summarized as[62]:  Si(OC2H5)4 + 2H2O → SiO2 + 4C2H5OH (5.1) As such, for each mole of TEOS, the overall complete reaction theoretically consumes 2 moles of water and produces 4 moles of ethanol. In the meantime, yttrium nitrate keeps dissolving in ethanol at elevated temperature if the previous dissolution (i.e. at room temperature and 75℃) is not complete. Assuming complete hydrolysis and condensation according to reaction (5.1), the 47  composition of the sol (mole : mole) after the reactions can be theoretically described as Y(NO3)3 : SiO2 : H2O : ethanol = 1 : 1 : 4 : 19. The control of solvent evaporation, which is ethanol in this case, is vitally important to the control of sol properties, especially viscosity. Since 75 ℃ was used to promote the dissolution and aging process of the sol, it’s critical to prevent ethanol from evaporation at this elevated temperature. Therefore, the dissolution and aging steps took place in a sealed glass container which served as a reactor. With the aim of testing the air-tightness of this simple reactor, weight loss of the sol was monitored during dissolution stage and aging process. According to the weight measurement, the dissolution stage, which took place at room temperature for 1 h and at 75 ℃ for 3 h, caused the sol to lose ~1wt% of its ethanol. Aging at room temperature for 48 h brought about the loss of <1wt% of the ethanol. Aging at 75 ℃ for 8 h caused the sol to lose ~3wt% of the ethanol. And aging at 75 ℃ for 16 h resulted in the loss of 8-10wt% of ethanol, which accounts for less than 1wt% of the total mass of the sol. Therefore, the ethanol content was kept at a relatively stable level during the dissolution and aging process. 5.3 Effects of sol aging conditions Aging of the precursor solution has been reported as a critical stage in the sol-gel processing of various coatings [63][64][65]. In terms of the sol-gel-derived rare earth silicate coatings and powders, different aging conditions have been used by different researchers [25][32][66]. But few of them have addressed the detailed effects of aging conditions. In this study, sols prepared with 4 different aging conditions have been investigated with the aim of exploring the aging effects on the sol-gel-derived Y2Si2O7 coatings. 48  Figure 5.1 presents the micrographs of the surfaces of the GPs applied with coatings from different sols (preparation conditions summarized in Table 3.1), all of which were captured at 2 mm back from the tips. All the coatings were prepared via the withdrawal speed of 20 mm/min. Unlike the coatings that will be discussed in Chapter 6-8, the coatings in Figure 5.1 have only been dried at 75 ℃ to remove the solvent (ethanol) without going through any further heat treatment. This aims to investigate the microstructure of the dried sol without introducing the effect of sintering.           Figure 5.1 Micrographs of coatings prepared with different sols after drying: (a)(b)(c) YRT24h; (d)(e) Y75C8h; (f)(g)(h) Y75C16h; (i)(j) Y75C16h-2weeks 14% N 9.9% Y 7.1% Mo  13% N 9.2% Y 8.4% Mo  16% N 18% Y 0.6% Mo  1  200 µm 50 µm 10 µm 10 µm 200 µm 50 µm 200 µm 50 µm 200 µm 50 µm 16% N 17% Y 1.2% Mo  a b c d e f g h j i a b c 49  As can be observed from Figure 5.1a and 5.1b, dark grey blobs whose sizes range from ~40 µm to ~60 µm are present on the coating prepared with sol YRT24h. EDS spot analysis of the blobs indicates higher nitrogen and yttrium contents than their surrounding areas. Spot 1 on Figure 5.1b, for instance, has 17wt% of nitrogen and 17wt% of yttrium, both of which are higher than the contents acquired by area analysis shown on Figure 5.1a. Thus, the blobs are determined to be the undissolved yttrium nitrate particles that have been dispersed and suspended within the sol. As such, sol YRT24h is speculated to be a colloidal system consisting of a continuous liquid phase, ethanol/water with dissolved yttrium nitrate and TEOS, and a dispersed phase of the yttrium nitrate particles.  At higher magnification, Figure 5.3c shows that the coating is porous with most pores of ~1 µm. The pores were introduced as a result of solvent evaporation during drying. The bright spots that are visible under the coating are MoSi2 present on the surface of the ceramic pin of the GP. After aging the sol at room temperature for 40 h and at 75 ℃ for 8 h under continuous stirring, less undissolved yttrium nitrate can be observed on the coating from sol Y75C8h (Fig. 5.1d and 5.1e), compared with the coating from YRT24h. And the coating became much less porous as well. However, Figure 5.1e shows that there are drying cracks as wide as several microns on the coating. On the one hand, aging at elevated temperature (75 ℃ for 8 h) reduces the porosity and yttrium nitrate blobs on the coating. On the other hand, however, it also seems to promote the formation of cracks by connecting the pores. Figures 5.1f-5.1h present the micrographs of the coating prepared with sol Y75C16h, which was aged at room temperature for 32 h and at 75 ℃ for 16 h. Blobs of the dispersed yttrium nitrate as large as 60~80 µm can be observed on Figure 5.1f. Figure 5.1g and 5.1h show that the coating has 50  a low porosity and few cracks. The slightly inter-connected pores on the coating are of small sizes. Compared with YRT24h and Y75C8h, aging at 75 ℃ for 16 h yields the best coating, with improved compactness, lower porosity and better uniformity.  The coating prepared with sol Y75C16h-2weeks is as shown by Figure 5.1i and 5.1j. This sol was aged without stirring at room temperature for additional 2 weeks, after the high temperature aging at 75 ℃ for 16 h as experienced by sol Y75C16h. Blobs of the dispersed yttrium nitrate, whose sizes range from ~ 5 µm to ~30 µm, are still present on the coating. Thus, the sol remained a stable colloidal system during the two-week aging process. By comparing Figure 5.1i to Figure 5.1f, the quantity of the blobs increased while their average size decreased after the two-week aging period. And the difference between Figure 5.1j and Figure 5.1g shows that the size of the pores has decreased and their redistribution contributed to the uniformity of the coating. The changes on the microstructure of the coating should be attributed to the ongoing hydrolysis and condensation reactions during the aging period. In general, aging of the sol is critical to the microstructural tailoring of the yttrium disilicate coatings. Both aging at 75 ℃ for 16 h and at room temperature for 2 weeks significantly contributed to achieving the more desirable size and distribution of the pores in the dried coating. The aging effects on the sintered coatings will be further discussed in Section 5.4.  51  5.4 Phase analysis of bulk Y2Si2O7 materials The service temperatures on the first 10 mm of the ceramic pins of the powered GPs, where the yttrium disilicate coating are applied, are in the range of 1200-1300℃. With the purpose of phase identification, bulk Y2Si2O7 materials heat-treated at 1200℃ and 1300℃ were prepared and characterized by XRD analysis. Figure 5.2a presents the XRD pattern of the bulk sample prepared from sol Y75C16 after heat treatment at 1200℃ for 1.5 h. As can be seen from the identified peaks, the bulk material consists of two polymorphs, y-Y2Si2O7 and α-Y2Si2O7. The peaks from the y phase have much stronger intensities than those from the α phase do, indicating that y-Y2Si2O7 is the dominant phase at 1200℃. The phase identification software, Match!, determined the contents of the y phase and α phase to be approximately 73wt% and 27wt%, respectively. y-Y2Si2O7, the low-temperature polymorph, remained the major phase under such heat treatment conditions, which was likely due to the stabilizing effect of H+ in the sol. H+, introduced by the hydrochloric acid added as the catalyst, is one of the possible stabilizers for y-Y2Si2O7 [44]. The pH is ~1.1 for sol Y75C16h. It was reported that the stability of y-Y2Si2O7 is higher than α-Y2Si2O7 for sol-gel-derived yttrium disilicate when the pH of the sol is lower than 5 [51]. The XRD pattern in Figure 5.2b was obtained from the bulk sample heat treated at 1200℃ for 1.5 h followed by additional heat treatment at 1300 ℃ for 1.5 h. Given the high similarity between Figure 5.2a and 5.2b, the additional heat treatment at 1300 ℃ for 1.5 h didn’t trigger the phase transformation from y-Y2Si2O7 to α-Y2Si2O7 or β-Y2Si2O7. This contradicts some of the commonly cited studies [35][44][45] about the polymorphs and phase transformations of Y2Si2O7. The 52  ambiguity regarding the stability of different polymorphs of Y2Si2O7, as has also been reported by other researchers, is mainly attributed to the different processing methods, impurities and heat treatment history [37].   Figure 5.2 XRD patterns of the bulk materials obtained from sol Y75C16h after heat treatment at: (a) 1200 ℃ for 1.5 h; (b) 1200 ℃ for 1.5 h followed by 1300 ℃ for 1.5 h 53  5.5 In-situ sintering The in-situ sintering of the coatings (as specified in Section 3.3.2) was carried out by powering the coated glow plugs at 12 V for 1.5 h via DC power supply in ambient environment. Through controlled ramping of the voltage, the surface temperature of the ceramic pin gradually rises from room temperature to ~1250 ℃ within ~10 min, providing the temperature needed for the sintering of the applied coatings.       Figure 5.3 Micrographs of coatings prepared with different sols after in-situ sintering: (a)(b) Y75C8h; (c)(d) Y75C16h; (e)(f) Y75C16h-2weeks 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm a b c d e f 54  Figure 5.3 presents the micrographs of the yttrium disilicate coatings prepared from different sols after the heat treatment via in-situ sintering. The surface morphologies of the sintered coatings show that the blobs as shown in Figure 5.1 are no longer present on all three samples. The in-situ heat treatment of the dried coating involved complex reactions, including the thermal decomposition of yttrium nitrate [67] and TEOS [68] (if its hydrolysis is incomplete in the aging stage), as well as the sintering between Y2O3 and SiO2. The disappearance of the yttrium-rich blobs may be attributed to the migration of Y3+ cations and their reactions with the surrounding silica. Not only was silica produced through the hydrolysis/decomposition of TEOS, but also it was present at the coating/substrate boundary due to the oxidation of silicon nitride. Therefore, the abundance of silica should facilitate its reaction with the yttrium-rich blobs, leading to the disappearance of the blobs and eventually an even surface morphology of the coating. Figure 5.3b shows that there are some areas that are not well-sintered on the coating from sol Y75C8h. In Figure 5.3d and 5.3f, the coatings from Y75C16h and Y75C16h-2weeks are mostly well-sintered but both subject to the presence of micro-cracks. Overall speaking, the coating from Y75C16h-2weeks demonstrates the best microstructure among the three specimens, with the best compactness and uniformity. The in-situ sintering of the coatings is quite different from the furnace heat treatment of the bulk materials. The different heating rate and temperature distribution at different locations across the surface of the ceramic pin can possibly have significant impacts on the crystal structure of the coatings. Therefore, the XRD analysis of the bulk samples may not be entirely indicative of the phase constitution of the coatings. The phase characterization on the coatings, which wasn’t 55  performed during this study due to the required techniques and costly expense, will be needed in order to further the knowledge about the in-situ sintered yttrium disilicate coatings. 5.6 Discussion and conclusion The investigation of the aging stage revealed that the aging conditions play an important role in the control of the uniformity, compactness, size and distribution of the pores and cracks of the sol-gel-derived yttrium disilicate coatings. XRD characterization of the powder samples indicated that the bulk materials heat treated at 1200 ℃ and 1200 ℃+1300 ℃ in furnace were both made up a mixture of y-Y2Si2O7 and α-Y2Si2O7. Given the different heating rate and temperature distribution on GP surface for the in-situ heat treatment, phase characterization of the coatings on the glow plug specimens is indispensable for future research. 56  Chapter 6: Effects of The Dip-coating Withdrawal Speeds  6.1 Introduction The typical dip-coating process consists of five stages: immersion, start-up, deposition, drainage and evaporation (refer to Section 3.3 for the description of the dip-coating process used in this study). The deposition of the coating was carried out by vertically withdrawing the substrate (Le-Mark glow plug) from the solution reservoir at a constant speed. The gravitational draining of the sol occurs simultaneously with the deposition. Among the various processing parameters during sol-gel dip-coating, the withdrawal speed is critical in controlling the coating properties, especially the coating thickness. Therefore, investigating the effects of withdrawal speeds is essential to have a better control and understanding of the sol-gel dip-coating processing of yttrium disilicate coatings in this research project. The typical range of withdrawal rates for sol-gel dip-coating is 1-10 mm/s (60-600 mm/min). In this study, the range under investigation is expanded to 5-500 mm/min, with the aim of exploring the coatings prepared by a broader range of speeds. A set of Y2Si2O7 coatings were prepared on Le-Mark glow plugs (GPs) using 6 different withdrawal speeds from the sol coded YRT75C16h (as listed in Table 3.1), whose solid content (weight percentage of the nonvolatile species of the sol) was 37.3wt%.  In this chapter, we look into the influence of the withdrawal speeds on the surface morphology, microstructure and the thickness of the coatings based on the distances of the coating regions from the GP tip.  The research on withdrawal speeds by the group of Grosso [55] will be also introduced 57  during the discussion of the experimental results to facilitate the explanation and comprehension of our case. It should be noted that the discussion on the withdrawal speeds will follow a decreasing order, starting from the higher end at 500 mm/min, which is within the typical range of withdrawal speeds used in classic sol-gel dip-coating process, and reaching the lower end at 5 mm/min, which is regarded as an ultra-slow withdrawal speed in the practice of dip-coating. 6.2 The effects of withdrawal speed at 2 mm from GP tip Figure 6.1 shows the micrographs of the coatings prepared via 6 different withdrawal speeds in the range of 5-500 mm/min, all of which were captured at the distance of 2 mm back from the GP tips. It was found that the coating is usually of the greatest thickness at around the first 2 mm from the GP tip. As can be seen from Figure 6.1a and 6.1b, the coating from the withdrawal speed of 500 mm/min exhibits massive cracks at 2 mm from the GP tip. The length of the cracks can reach several hundred microns. This kind of cracks usually result from the volumetric shrinkage of thick coatings during the drying process, when the removal of a large amount of solvent brings about the residual tensile stress within the coating [69]. When the withdrawal speed decreases to 100 mm/min, the corresponding coating appears to be crack-free, as demonstrated in Figure 6.1c and 6.1d. At 50 mm/min, the obtained coating is mostly compact but with a few cracks of several microns (Figure 6.1e and 6.1f). The coating yielded from 20 mm/min shows cracks of ~10 µm at 2 mm (Figure 6.1g and 6.1h). At 10 mm/min, the corresponding coating, based on its darker appearance in Figure 6.1i and 6.1j, seems to be thinner and less compact than the coatings from other withdrawal rates. 58  At 5 mm/min, the obtained coating exhibits massive inter-connected cracks (Figure 6.1k and 6.1l), which can be an indication that it is thicker than the coating from 10 mm/min. Table 6.1 summarizes the results of the EDS analysis performed at the areas shown by the 200X micrographs in Figure 6.1. As discussed in Chapter 3-4, The yttrium contents and molybdenum contents are used as semi-quantitative indicators of the coating thickness in this study. Typically, the thicker the coating, the stronger the signal for yttrium from the coating and weaker the signal for molybdenum from the GP substrate detected by the EDS. Therefore, higher Y content and lower Mo content generally correspond to thicker coating.        200 µm 200 µm 200 µm 50 µm 50 µm 50 µm a b c d e f h 59           Figure 6.1 Micrographs of the coatings from different withdrawal speeds at the distance of 2 mm from tip: (a)(b) 500 mm/min; (c)(d) 100 mm/min; (e)(f) 50 mm/min; (g)(h) 20 mm/min; (i)(j) 10 mm/min; (k)(l) 5 mm/min The line chart in Figure 6.2 shows the change of the coatings’ elemental compositions with the change of withdrawal speeds. Starting from 500 mm/min, with the decrease of the withdrawal speed, the Y content first goes down and then goes up, with the coating from 10 mm/min showing the lowest Y weight percentage. Correspondingly, with the decrease of withdrawal speed, the Mo content first increases, reaches its highest value at 10 mm/min, and then decreases. On the basis of this trend, it is speculated that the coating thickness first decreases and then increases, when the 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm g i j k l h 60  withdrawal speed goes from 500 mm/min to 5 mm/min. And the coating obtained from 10 mm/min should be the thinnest amongst the coatings from the six withdrawal rates explored here. Table 6.1 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.1 Location O  Al Si Y  Mo #500-2mm 40 2.6 21 36 0.1 #100-2mm 42 3.8 20 33 0.9 #50-2mm 39 3.2 22 33 3.2 #20-2mm 43 2.6 22 29 3.3 #10-2mm 45 2.8 24 19 8.6 #5-2mm 42 3.6 23 27 3.9 Notes:  1. the format of the location name in all the tables of this chapter is: #[withdrawal speed(mm/min)]-[distance from the GP tip]. For example, #500-2mm means the EDS result is collected on the coating prepared with the withdrawal speed of 500 mm/min, at 2 mm from the GP tip. 2. All EDS data are in wt% of the elements, but excluding nitrogen as discussed in Chapter 3-4.   Figure 6.2 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.1 61  6.3 The effects of withdrawal speed at 4 mm from GP tip Figure 6.3 shows the micrographs of the same coatings as demonstrated by Figure 6.1, but at the distance of 4 mm from the GP tips. At 4 mm from the GP tips, only the coatings from 500 mm/min and 5 mm/min are subject to the presence of massive cracks, the sizes of which can be greater than 100 µm. The coatings from the other four withdrawal speeds all appear to be crack-free. The speed of 20 mm/min turns out to yield the coating with the most satisfying uniformness out of the 6 withdrawal speeds. But the coating is not entirely compact in some areas. The speed of 100 mm/min, on the other hand, produces the coating with the best compactness and coverage of the substrate.       200 µm 200 µm 200 µm 50 µm 50 µm 50 µm a b c d e f 62        Figure 6.3 Micrographs of coatings from different withdrawal speeds at the distance of 4 mm from tip: (a)(b) 500 mm/min; (c)(d) 100 mm/min; (e)(f) 50 mm/min; (g)(h) 20 mm/min; (i)(j) 10 mm/min; (k)(l) 5 mm/min  The EDS results of the coatings at 4 mm from the GP tips are as demonstrated in Table 6.2 and Figure 6.4. The trend of the elemental composition evolution at 4 mm from GP tips is a bit different from that at 2 mm. With the withdrawal speed decreasing from 500 mm/min to 5 mm/min, the yttrium content first goes down, then fluctuates in the range of 10-50 mm/min, and finally increases again when the speed goes below 10 mm/min. Meanwhile, the molybdenum content exhibits the completely opposite trend. 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm g h i j k l 63  Table 6.2 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.3 Location O  Al Si Y  Mo #500-4mm 40 2.7 22 35 0.2 #100-4mm 42 4.3 22 30 2.0 #50-4mm 42 3.3 28 17 10 #20-4mm 40 3.6 28 20 8.2 #10-4mm 44 3.0 26 16 11 #5-4mm 38 2.2 24 31 4.6   Figure 6.4 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.2 Based on these EDS results, it is again speculated that the coating thickness decreases with the decrease of withdrawal speeds in the range of 50-500 mm/min. 20 mm/min produces thicker coating than 10 mm/min and 50 mm/min do. And the decrease of withdrawal speed from 10 mm/min to 5 mm/min witnesses a great increase of the coating thickness. 6.4 The effects of withdrawal speed at 6 mm from GP tip Like the micrographs at 4 mm from the GP tips, the micrographs captured at 6 mm from the tips in Figure 6.5 show that the withdrawal speeds of 500 mm/min and 5 mm/min result in massive 64  cracks on the coatings. The coating from 50 mm/min is too thin that even the MoSi2 on the substrate surface, which appears to be the irregular bright areas under SEM, can be observed through the coating. And 20 mm/min again produced the coating (Figure 6.5g and 6.5h) that is fairly uniform and compact, as well as free of cracks, at this distance from the tip.         200 µm 200 µm 200 µm 50 µm 50 µm 50 µm 200 µm 50 µm a b c d e f g h i 65      Figure 6.5 Micrographs of coatings from different withdrawal speeds at the distance 6 mm from tip: (a)(b) 500 mm/min; (c)(d) 100 mm/min; (e)(f) 50 mm/min; (g)(h) 20 mm/min; (i)(j) 10 mm/min; (k)(l) 5 mm/min  Table 6.3 EDS results (wt% of non-nitrogen elements) of the coating shown in Figure 6.5 Location O  Al Si Y  Mo #500-6mm 38 2.1 22 37 0.4 #100-6mm 42 3.9 24 26 4.1 #50-6mm 41 2.7 30 14 12 #20-6mm 43 3.0 28 15 11 #10-6mm 43 2.3 27 14 13 #5-6mm 38 2.2 27 24 8.0 Table 6.3 and Figure 6.6 summarize the EDS analysis results of the coated GPs at 6 mm from the tip. The trends of the evolution of yttrium and molybdenum contents at 6 mm are similar to those at 4 mm, except that the Y contents of the coatings from 10, 20 and 50 mm/min are almost the same. Again, the speculation is that the evolution of coating thickness and the change of yttrium contents are to a great extent alike. 200 µm 200 µm 50 µm 50 µm j k l i 66   Figure 6.6 Line chart (X-axis on the log10 scale) of the EDS results shown in Table 6.3  6.5 Discussion and conclusion Generally speaking, it is observed that there are two opposite coating thickness evolution regimes in the range of the withdrawal speeds explored here (5-500 mm/min). Starting from 500 mm/min, the coating thickness decreases with the decrease of withdrawal speed. When the speed is in the range of 10-50 mm/min, the coating thickness evolution trend varies depending on the distance of the coating areas from the GP tip. When the withdrawal speed keeps going down from 10 mm/min to 5 mm/min, the coating thickness increases.  This phenomenon can be explained by the research conducted by Grosso et al [55], which showed that two regimes of film formation independently exist at extreme conditions during sol-gel dip coating processes. Figure 1.6 schematically illustrates the dip-coating method and the two coating evolution regimes explored in Grosso’s study. 67  At higher withdrawal speeds, the film thickness is governed by gravity-induced viscous drag, which is the well-known “draining regime”. The classic Landau-Levich model [56], demonstrated in the second row of Table 1.3, describes this regime as where the coating thickness is proportional to U02/3 (U0 is the withdrawal speed). The second regime is the “capillary regime” observed by the group of Grosso during the dip-coating of reactive sol-gel solutions at lower withdrawal speeds [55]. At ultra-slow withdrawal speeds, the coating thickness is governed by capillarity rise, the continuous feeding of the upper part of the gravitational meniscus of the sol to the dried material above the drying line (vapor-liquid-solid, three phases frontier). The semi-experimental model proposed by Grosso [55], also shown in Table 1.3, describes the coating thickness in this regime to be inversely proportional to the withdrawal speed. The two regimes co-exist and compete with each other at intermediate withdrawal speeds, for which the coating thickness was modelled by adding the contributions of both regimes as shown in Table 1.3. And the counterbalance of the two regimes results in a minimum film thickness at a certain speed. This critical speed uc is found to be around 0.2-0.8 mm/s in Grosso’s research.  Analogically, the co-existence of the “draining regime” and the “capillary regime” should control the mechanism behind the coating thickness evolution studied in our system, where the withdrawal speed range of 50-500 mm/min is dominated by the draining regime while the capillary regime is at the dominant position within 5-10 mm/min. The speed of 20 mm/min, which is out of the typical range of withdrawal speeds used for traditional sol-gel dip-coating practice, turns out to yield in the coating with the best uniformity and is thus selected for the fabrication of most coatings in this study. 100 mm/min also provides the coating with excellent compactness and coverage. These empirical findings are helpful as we look to tailor the coating properties in further endeavors. 68  Chapter 7: Multi-layer Y2Si2O7 Coatings  7.1 Introduction Judging from the findings reported in Chapter 6, a single layer of Y2Si2O7 coating is either too thin to provide satisfactory protection capacity or too thick to be crack-free. To simultaneously achieve sufficient thickness in the range of 5-10 µm and to avoid the formation of cracks, it is decided to successively apply and process thin layers of coating, each with the thickness of ~1 µm. In this chapter, Y2Si2O7 coatings with different numbers of layers on glow plugs (GPs) is going to be investigated. 7.2 Multi-layer Y2Si2O7 coatings at 2 mm from the GP tip Figure 7.1 shows the micrographs of Le-Mark(LM) glow plugs applied with different layers of yttrium silicate coatings, all of which were taken at 2 mm from the GP tips. All the layers of the coatings were prepared using the withdrawal speed of 20 mm/min during dip-coating. It should be noted that these coatings were applied on different GP specimens, except the 6-layer coating was prepared by applying an additional layer on top of the 5-layer coating.    200 µm 50 µm a b 69              Figure 7.1 Micrographs of LM GPs (2 mm from tip) with different layers of coatings after in-situ sintering: (a) (b) 1 layer; (c)(d) 2 layers; (e)(f) 3 layers; (g)(h) 5 layers; (i)(j) 6 layers In Figure 7.1a and 7.1b, which belong to a single-layer coating, cracks as long as 10 µm are present at 2 mm from tip. The presence of cracks in this area is expected as the first 2 mm of the coating is usually of the greatest thickness and therefore harder to be rid of cracks. The 2-layer coating is 200 µm 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm 50 µm c d e f g h i j 70  mostly crack-free and dense, as shown in Figure 7.1c and 7.1d. For the coating with 3 layers, the micrographs in Figure 7.1e and 7.1f indicate a high level of uniformity and compactness of its surface morphology at 2 mm from tip. It is not only free of cracks and voids but also well-sintered that the grain boundaries can no longer be distinguished. The coatings with 5 layers and 6 layers, both of which belong to the same GP specimen, resemble each other in terms of surface morphology under SEM, except that the 6-layer coating is a bit denser than the 5-layer one in some areas where the crystal grains can still be distinguished.  Table 7.1 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.1  O  Al Si Y  Mo 1L-2mm 38 1.1 25 30 5.6 2L-2mm 34 0.5 19 47 0.0 3L-2mm 37 1.2 18 44 0.0 5L-2mm 37 1.3 18 44 0.0 6L-2mm 34 1.1 19 46 0.0   Figure 7.2 Elemental contents of the multi-layer coatings at 2 mm from the GP tips 71  Table 7.1 and the line chart in Figure 7.2 summarize the EDS area analysis results of the same areas as captured by the micrographs (with the magnification of 200X) in Figure 7.1. There is a 17wt% increase of the yttrium content when the number of layers goes from 1 to 2. The elemental weight percentages of silicon, oxygen and molybdenum all decrease simultaneously. And the molybdenum content detected from the ceramic pin of GP reaches zero when two layers were applied on its surface. This indicates the 2-layer coating has achieved a thickness greater than the resolution depth of the EDS analysis performed here that it prevents the Mo on the substrate surface from being detected. What is unexpected is that the yttrium content decreases by 3wt% as the number of layers goes from 2 to 3. The elemental compositions of the 3-layer coating and 5-layer coating are almost identical, while the 6-layer coating possesses higher yttrium content than the 5-layer one does. This unexpected trend of yttrium content evolution may be due to the minor variances between the coatings prepared on different GP specimens. Despite the small error, it is observed that the detected elemental compositions of the coatings basically remain stable after the number of coating layers reach 2, with yttrium contents within the range of 44-47wt%. Therefore, at 2 mm from the GP tip, it is hard to tell the evolution of the coating thickness from the EDS results after the number of layers goes beyond 2.  7.3 Multi-layer Y2Si2O7 coatings at 4 mm from the GP tip Figure 7.3 presents the micrographs of Le-Mark glow plugs applied with multiple layers of yttrium silicate coatings, all of which were taken at 4 mm from the GP tips. The coating with only 1 layer is too thin that even the substrate beneath it can be seen in Figure 7.3a and 7.3b. The 2-layer coating is made up of yttrium silicate crystals whose sizes range from submicron to several microns. The 72  sintering of these crystals is mostly inadequate. The 3-layer coating shows good compactness and uniformity at 4 mm from the tip like it does at 2 mm. The coatings with 5 layers and 6 layers again have very similar surface morphology at 4 mm, where the yttrium silicate crystals are not well-sintered in some areas.             50 µm 50 µm 50 µm 50 µm 200 µm 200 µm 200 µm 200 µm a b c d e f g h 73     Figure 7.3 Micrographs of LM GPs (4 mm from tip) with different layers of coatings after in-situ sintering: (a)(b) 1 layer; (c)(d) 2 layers; (e)(f) 3 layers; (g)(h) 5 layers; (i)(j) 6 layers Table 7.2 and the line chart in Figure 7.4 demonstrate the EDS area analysis results of the same areas as captured by the micrographs (with the magnification of 200X) in Figure 7.3. With the increase of the number of layers, the yttrium contents at 4 mm from the tip gradually goes up. The greatest increase is 18wt%, from the comparison between the 1-layer coating and the 2-layer coating. Correspondingly, the Mo contents gradually goes down with the increase of the number of layers. At this region on the coated GPs, the detected Mo content doesn’t reach zero until more than 3 layers of coating were applied. Therefore, at 4 mm from the GP tips, it takes more than 3 layers of coating to reach a thickness larger the resolution depth of the EDS analysis system, which is due to the fact that the coating is much thinner at 4 mm than it is at 2 mm from the tip. Table 7.2 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.3  O  Al Si Y  Mo 1L-4mm 39 2.2 31 15 12 2L-4mm 37 2.1 25 33 3.4 3L-4mm 41 3.1 19 37 0.3 5L-4mm 38 1.8 19 42 0.0 6L-4mm 37 1.5 18 43 0.0  50 µm 200 µm i j 74   Figure 7.4 Elemental contents of the multi-layer coatings at 4 mm from the GP tips  7.4 Multi-layer Y2Si2O7 coatings at 6 mm from the GP tip Figure 7.5 shows the micrographs of LM glow plugs applied with different layers of yttrium silicate coatings, all of which were captured at 6 mm from the GP tips. In Figure 7.5a and 7.5b, the substrate is visible through the coating, indicating the 1-layer coating is quite thin at 6 mm from the tip, just like at 4 mm. For the coatings with 3 layers, 5 layers and 6 layers, the micrographs in Figure 7.5e-7.5j suggest their surface morphologies very much resemble each other at 6 mm from the tips. All the coatings are free of cracks, with outstanding compactness and uniformity.    200 µm 50 µm a b 75              Figure 7.5 Micrographs of LM GPs (6 mm from tip) with different layers of coatings after in-situ sintering: (a) (b) 1 layer; (c)(d) 2 layers; (e)(f) 3 layers; (g)(h) 5 layers; (i)(j) 6 layers Table 7.3 and the line chart in Figure 7.6 exhibit the EDS area analysis results of the same areas as captured by the micrographs (with the magnification of 200X) in Figure 7.5. The evolution of 200 µm 200 µm 200 µm 200 µm 50 µm 50 µm 50 µm 50 µm c d e f g h i j 76  the elemental composition of the coatings at 6 mm from the GP tip, especially the yttrium and molybdenum contents, is similar to that at 4 mm. The difference lies in the greater increase of Y content and decrease of Mo content when the number of layers goes from 3 to 5, comparing 6 mm to 4 mm. The Mo content at 6 mm from tip also reaches zero when 5 layers of coating were fabricated on GP, like at 4 mm.  Table 7.3 EDS results (wt% of non-nitrogen elements) of the coatings shown in Figure 7.5  O  Al Si Y  Mo 1L-6mm 39 2.2 31 14 14 2L-6mm 39 1.9 25 29 5.4 3L-6mm 39 2.6 22 34 2.6 5L-6mm 36 1.5 19 44 0.0 6L-6mm 33 1.2 20 46 0.0   Figure 7.6 Elemental contents of the multi-layer coatings at 6 mm from the GP tips 77  With 1-3 layers, the coatings should be thinner at 6 mm from the tip than at 4 mm, based on their lower Y contents and higher Mo contents at 6 mm. But it is hard to tell the difference between different regions of the 5-layer and 6-layer coatings since the EDS results are no longer associated with the evolution of the coating thickness once the thickness exceeds the resolution depth of the EDS system.  7.5 Thickness and adhesion of the multi-layer Y2Si2O7 coatings With the aim of taking a more straightforward approach toward the measurement of the coating thickness, the coated GP specimens were cross-sectioned and examined under SEM. Figure 7.7 presents the micrographs of the cross-sections of the coated GP specimens at 2 mm from the GP tips.  Figures 7.7a and 7.7b show that a single layer of coating is too thin that it can barely be observed with the magnification of 1000X under SEM. The coating appears to be discontinuous since nearly half of it has been pulled apart by the mounting resin, which was used to facilitate the preparation of the cross-section specimen. This indicates that the single layer of coating, which was sintered for 1.5 h, does not provide sufficient adhesion to the GP surface. In general, the coating thickness of the single layer does not exceed 1 µm. As demonstrated in Figure 7.7c and 7.7d, there is obvious improvement in the adhesion of the 3-layer coating, although a small portion of the coating has still been pulled apart by the mounting resin. The micrographs indicate the thickness of the 3-layer coating to be in the range of 1-2.5 µm. 78          Figure 7.7 Micrographs of the cross-sections of Le-Mark GPs with different layers of Y2Si2O7 coatings at 2 mm from GP tips: (a)(b) 1 layer; (c)(d) 3 layers; (e)(f) 6 layers With 6 layers, the coating shown in Figure 7.7e and 7.7f was able to remain undamaged and uniform from the attack of the mounting resin. The 6-layer coating has a high level of compactness, with very few pores and voids. Also, the coating is well-sintered that the boundaries between different layers can no longer be distinguished in the micrographs. Based on the cross-section micrographs, the thickness of the 6-layer coating is ~5.5 µm on average. The enhancement of the adhesion of the 6-layer coating is likely due to the increased thickness and the longer total sintering time. As each layer of the coating was in-situ sintered for 1.5 h, the 6-layer coating had a total 50 µm 50 µm 50 µm 50 µm 50 µm 50 µm resin Si3N4 (MoSi2, Y2O3) coating coating coating coating coating coating a b c d e f 79  sintering time of 9 h, which was 6 times of the single-layer coating. Longer heat treatment facilitated the sintering between the externally applied yttrium disilicate coating and the self-formed yttrium silicate on the GP surface, which led to enhanced bonding between the coating and the GP substrate.  To sum up, compared with the single-layer and 3-layer coatings, the 6-layer Y2Si2O7 coating was able to achieve not only a more desirable thickness but also much enhanced adhesion to the substrate. 7.6 Discussion and conclusion With the increase of the number of layers, the evolution of the coating thickness and surface morphology in different regions have been discussed in detail. The increase of the number of layers leads to gradual increase of the coating thickness at all distances from the GP tip. The 6-layer crack-free coating was able to achieve an average thickness of ~5.5 µm and much enhanced adhesion to the GP, both of which could not be done with 3 layers let alone single layer. As such, successively applying and processing thin layers of coating is proved to be an effective approach to prepare the sol-gel-derived Y2Si2O7 coatings with customizable thickness and enhanced adherence on the surface of Le-Mark glow plug. 80  Chapter 8: Tests of the Performance of the Coated Glow Plugs  8.1 Introduction The performance of the glow plugs (GPs) coated with Y2Si2O7 coatings was tested at ~1200 ℃ on a natural gas burner rig and in high concentration water vapor atmosphere. The protection capability of such coatings was evaluated by comparing the coated GPs with the uncoated ones. 8.2 Natural gas burner rig test The natural gas burner rig (BR) test was employed as an accelerated degradation test for the coated glow plug (GP) specimens. The estimated surface temperature of the coated Le-Mark GP was ~1200 ℃ and other testing parameters are as specified in Section 3.6.1. The BR test involves the combined effects of the direct current (DC) electric field, oxidizing environment, reducing environment and corrosion of the surface of the ceramic heater by combustion gases. Therefore, the BR provided the closest simulation of the in-cylinder environment of NG heat engine we could achieve in laboratory environment (the engine tests for the coated GPs will be performed by Westport at a later stage). Both the as-received GP (BRGP0) and the GP with applied Y2Si2O7 coating (BRGP1) were tested in the natural gas burner rig. 8.2.1 Burner rig tests for the uncoated glow plug Figure 8.1 shows the micrographs of the surface of BRGP0, an as-received Le-Mark GP after 3 weeks of continuous natural gas burner rig test, which was a control sample to the coated GP specimen. Thick porous silica layer, among which isolated acicular yttrium silicate crystals were present, was developed on GP surface after the test. There are more self-formed yttrium silicates 81  observed on the micrograph at 4 mm back from the GP tip, where the service temperature was higher, compared to 2 mm from the tip. However, the EDS results suggest lower yttrium content detected at 4 mm than at 2 mm, which might be due to thicker silica deposit at 4 mm that facilitates the detection of Si while inhibiting the detection of Y.         Figure 8.1 Micrographs of the surface of the ceramic pin for specimen BRGP0 after BR test for 3 weeks: (a)(b) at 2 mm from tip; (c)(d) at 4 mm; (e)(f) at 6 mm; (g) at 3 mm (cross section); (h) at 6 mm (cross section) 200 µm 50 µm 200 µm 200 µm 50 µm 50 µm 36% Si 7.2% Y 0.6% Mo 39% Si 3.4% Y 0.9% Mo 36% Si 6.5% Y 0.6% Mo 50 µm 50 µm a b c d e f g h 82  The cross-section micrographs in Figure 8.1g and 8.1h show that the thickness of the silica deposits at 3 mm and 6 mm from the GP tip is in the range of 10~30 µm. The deposit at 3 mm from the tip is thicker than that at 6 mm, as a result of the faster oxidation promoted by the higher temperature at the hot-zone (3~4 mm back from the GP tip). The jagged boundary between the substrate and the deposit indicates there was significant material loss on the surface of the GP. The hypothesis on the degradation mechanisms of the uncoated GP in BR test includes the oxidation of the Si3N4 substrate, the erosion of the porous silica deposit caused by the flame and the combustion gases, as well as the reduction of the silica into volatile species by the reducing gases. Therefore, it is concluded that the silica scale is not protective of the Si3N4 substrate against oxidation and further degradation in the gas burner rig test. 8.2.2 Burner rig test for the glow plug with Y2Si2O7 coating Figure 8.2 shows the micrographs of BRGP1, a Le-Mark glow plug with applied Y2Si2O7 coating before and after 3 weeks of continuous gas burner rig test. The surface of the coated GP before BR test is as presented in Figure 8.2a. Comparing Figure 8.2a to 8.2b, both of which were taken at 1~2 mm from the tip of the Si3N4 heater of the glow plug, the coating exhibits better coverage after the BR test, with the cracks being sealed by self-grown yttrium silicates from the substrate. EDS analysis suggests zero molybdenum content, lower yttrium content and higher silicon content after the BR test. Base on the cross-section micrographs, Figure 8.2g and 8.2h, the coating at 1 mm from tip is quite porous but adheres well to the substrate after the BR test. The coating thickness is approximately 8~12 µm at this spot. The silica layer formed between the coating and the GP surface can barely 83  be distinguished. The boundary between the substrate and the coating is smooth and the GP edge retained its original shape. No material loss like that observed on the uncoated specimen is observed here, indicating that the applied yttrium disilicate coating was effective in protecting the Si3N4 substrate from further oxidation and degradation caused by the burner rig.         200 µm 20 µm 50 µm 50 µm 50 µm 50 µm 50 µm 50 µm a b c d e f g h 84    Figure 8.2 Micrographs of the surface of the ceramic pin for specimen BRGP1 before (a) and after the 3-week BR test (b-j): (a) at 1~2 mm, before test; (b) at 1~2 mm, after test; (c) at 4 mm, before test; (d) at 4 mm, after test;  (e) at 6 mm, before BR test; (f) at 6 mm, after BR test; (g)(h) at 1 mm, cross section, after test; (i)(j) at 4 mm, cross section, after test; As can be seen from Figure 8.2c and 8.2d, however, there is obvious spallation on the coating at 4 mm after BR test, which was likely to be caused by the mechanical impact by gas erosion. Both silica and yttrium silicates were detected on the dark area (indicated by the yellow arrow on Figure 8.2d). The cross-section micrographs (Figure 8.2i,j) indicate the coating thickness to be ~2 µm in this area. Figure 8.2e and 8.2f show that the spallation of the coating also occurred at the 6mm-from-tip areas but in a smaller scale than at 4 mm. In general, the coating shows good coverage yet poor adhesion to the substrate on most parts of the ceramic pin. The oxidation product, silica, is mostly present in the areas where the yttrium disilicate coating does not adhere well to the GP surface. It is therefore speculated that the formation of silica induced by the oxidation of the Si3N4 substrate at the coating-substrate boundary was a possible cause of the spallation of the coating. Oxygen penetrated the coating through the porous/thin areas, where the coating was unable to provide sufficient protection, and oxidized the Si3N4 substrate to produce silica deposit at the coating-substrate boundary. And the growth of silica caused the detachment and failure of the coating. 200 µm 50 µm i j 85  The bench characterization of the coated sample at 12 V indicates no obvious degradation of BRGP1, in terms of current and surface temperature, before and after the burner rig test. There are 2% degradation in current and 1% in temperature for the uncoated reference sample BRGP0, which, though, are within measuring error range. In conclusion, the burner rig test suggested that the applied yttrium disilicate coating has the capacity of protecting LM glow plugs from oxidation and corrosion, compared with the uncoated specimen. However, the poor adhesion and the relatively low thickness of the coating are major issues to deal with. Future work on the improvement of the adherence and the thickness is indispensable. 8.3 High concentration water vapor test High concentration water vapor (HCWV) test was employed to investigate the behavior of the coated GP in high temperature water vapor environment (the experimental setup for this test is described in Section 3.6.2).  Figure 8.3 Micrographs of the surface of the ceramic pin for specimen HCWVGP0 at 5 mm from tip after HCWV test for 300 h 50 µm 86   Very thin incipient yttrium silicates on cracked and porous silica scale was developed on the uncoated Le-Mark glow plug after the HCWV test. And there was 3% drop in current and surface temperature at 12V for bench characterization after the exposure. In comparison, the coated GP showed no drop in current and surface temperature. Figure 8.4 shows the micrographs of the surface of the GP’s ceramic pin, which is coated with Y2Si2O7, before (Figure 8.4a, 8.4c, 8.4e) and after (Figure 8.4b, 8.4d, 8.4f, 8.4g and 8.4h) the high concentration water vapor test. The EDS results of Figure 8.4a-8.4f are acquired by area analysis of the areas presented on the micrographs while the results of Figure 8.4g are acquired by spot analysis of the marked spots A, B and C. Given the EDS-measured low yttrium content (19wt% at 2 mm, 24wt% at 4 mm, 28wt% at 6 mm) and the high molybdenum content (6.5wt% at 2 mm, 3.9wt% at 4 mm, 3.0wt% at 6 mm) detected before the test, the fabricated yttrium disilicate coating is relatively thin. Compared with the micrographs before the HCWV test, the micrographs after the test suggest the exposure to high concentration water vapor causes the appearance of dark areas as marked by the yellow arrow in Figure 8.4b. The degradation of Y2Si2O7 in the presence of water vapor has been reported to be due to the selective volatilization of silica in Y2O3∙2SiO2, which results in a porous network of Y2SiO5 [41]: Y2Si2O7(s) + 2H2O(g) = Y2SiO5(s) + Si(OH)4(g)  87           Figure 8.4 Micrographs of the surface of the ceramic pin for specimen HCWVGP1 before and after HCWV test for 300 h: (a) at 2 mm, before test; (b) at 2 mm, after test; (c) at 3 mm, before test; (d) at 3 mm, after test;  (e) at 5 mm, before test; (f) at 5 mm, after test;  (g) at 1 mm (cross section), after test; (h) at 4 mm (cross section), after test; 19% Y 26% Si 6.5% Mo 21% Y 32% Si 2.0% Mo 24% Y 25% Si 3.9% Mo 29% Y 30% Si 0.8% Mo 28% Y 24% Si 3.0% Mo 32% Y 28% Si 1.0% Mo 31% Y 20% Si 2.3% Mo  0.2% Y 41% Si 51% O 0.2% Mo  28% Si 72% Mo  A B C 50 µm 50 µm 50 µm 50 µm 50 µm 50 µm 50 µm 50 µm resin Si3N4 (MoSi2, Y2O3) a b c d e f g h 88  EDS results indicate that the yttrium content increases by 2wt%, 5wt% and 4wt% at 2 mm, 3 mm and 5 mm back from the tip, respectively. The increase of yttrium contents is in accordance with the aforementioned degradation mechanism since Y2SiO5 has a higher yttrium content than Y2Si2O7. However, the characteristic micro-ridged structure of Y2SiO5 is not observed from the micrographs after the HCWV test. In addition, the increase of Y content could be due to the ongoing formation of self-grown yttrium silicates from the substrate as well. EDS results also indicate the decrease of Mo contents at all distances after the test, which should be due to the thickening of the silica interlayer that causes less Mo to be detected by EDS. Figures 8.4g and 8.4h show the micrographs of the cross-sections of the specimen at 1 mm and 4 mm from the GP tip, after the test. Three layers are observed on these micrographs, the light gray top layer, the dark gray intermediate layer and the substrate where a large number of bright spots are present. The thickness of the top layer is generally not greater than 3 µm while the intermediate layer is of the thickness of 0~7 µm. It is hard to distinguish the substrate and the intermediate layer in some areas since both the Si3N4 body and the intermediate layer appear to be dark gray under SEM. Based on the EDS results from the analysis of spot C (Fig. 8.4g), the bright areas within the substrate are MoSi2, which is used as an additive for this type of Le-Mark glow plug. With 41wt% Si and 51wt% O, spot B suggests the gray intermediate layer to be SiO2. Thus, the dark areas in Figure 8.4b, 8.4d and 8.4f should be the silica underneath the applied yttrium silicate coating. The spot analysis of the top layer (spot A) indicates its chemical composition to be 31wt% Y, 20wt% Si, 43wt% O, 2.3wt% Mo and 3.4wt% Al. It is speculated that the coating was turned into a mixture 89  of Y2SiO5, Y2Si2O7 and SiO2 after the HCWV test, which remains to be verified by further investigation. At 3 mm from the GP tip (Figure 8.4d), the material loss seemed to be more severe in the areas where the applied coating was not well-sintered. Micrograph 8.4e was captured at exactly the same spot as 8.4f at 5 mm, which can be distinguished from the surface morphology of the coating. The comparison between the two micrographs clearly suggests that the well-sintered areas remained compact and much less degraded than those not-well-sintered areas after the exposure to high concentration water vapor. 8.4 Discussion and conclusion Both the natural gas burner rig test and the high concentration water vapor test suggest that the Y2Si2O7 coating is capable of providing a certain yet limited level of corrosion protection. The limited protection capability of the coating is both due to the inherent limitation of the corrosion resistance of Y2Si2O7 and the undesirable microstructures of the fabricated coatings, as well as their overall low thickness (less than 3 µm) on most parts of the ceramic pin. The specimens discussed in this chapter were not able to achieve satisfactory thickness and compactness since both of the tests were carried out in the relatively early stage of this research project, when the processing parameters of the coating was still under investigation. Given the better performance of the thicker and well-sintered areas of the coatings, whether the coating with an adequate thickness and enhanced adherence, like the 6-layer coating discussed in Chapter 7 (not tested due to limited time frame of this project), will be able to provide improved corrosion resistance remains promising and worthy of further research.90  Chapter 9: Summary and Conclusions  A simple and cost-effective sol-gel dip coating route has been established to develop multi-layer crack-free Y2Si2O7 environmental barrier coatings on the exposed Si3N4-based ceramic pins of the Le-Mark glow plugs (GPs). As a result of the migration of the sintering additive Y3+ cations and their reaction with the surface silica scale, acicular yttrium silicate crystals were formed on the surface of the ceramic pin. This self-grown yttrium silicate, however, formed at the expense of the oxidation of the ceramic pin and the depletion of Y2O3 within the Si3N4 body. As such, externally applied yttrium disilicate coating was employed for the corrosion and oxidation protection of the Le-Mark GPs.  Aging of the precursor solution (sol) is critical to the microstructural tailoring of the yttrium disilicate coatings. Coatings prepared from four sols with different aging conditions were investigated. Both aging at 75 ℃ for a relatively short time (8 h and 16 h) and at room temperature for a longer time (2 weeks) contributed to the improvement of the coatings’ microstructures, including the uniformity, compactness, size and distribution of the pores and cracks of the coatings. XRD analysis of the Y75C16h powder sample revealed that the bulk material, whose sol was aged at 75 ℃ for 16 h, consists of y-Y2Si2O7 as the major phase and α-Y2Si2O7 as the minor phase after heat treatment at 1200 ℃ for 1.5 h in furnace. The additional heat treatment at 1300 ℃ for 1.5 h in furnace didn’t trigger the phase transformation of y-Y2Si2O7 to α-Y2Si2O7 or β-Y2Si2O7 of the bulk material. 91  Besides the aging conditions, another processing parameter that has been studied in detail is the withdrawal rate of the glow plug substrate during the dip coating procedure. It was observed that the effects of the withdrawal speed on the microstructures and thickness of the single-layer Y2Si2O7 coating were dependent on the locations of the coating areas on the ceramic pin of the GP. In general, two opposite coating thickness evolution trends were present in the range of the withdrawal speeds explored in this study (5-500 mm/min). Based on the semi-quantitative indicators (yttrium contents and molybdenum contents acquired by EDS analysis of the coated GPs), the coating thickness decreases with the decrease of the withdrawal speed when the speed is in the range of 50-500 mm/min. In contrast, the coating thickness increases when the withdrawal speed decreases from 10 mm/min to 5 mm/min. When the speed is in the range of 10-50 mm/min, the coating thickness evolution trend varies at 2 mm, 4 mm and 6 mm from the GP tip. The attempt on explaining the mechanisms behind such phenomena was made by referring to the two opposite evolution regimes, the “draining regime” described by the classic Landau-Levich model [56] and the “capillary regime” proposed by Grosso [49]. The optimal withdrawal speed out of the 6 investigated speeds was determined at 20 mm/min.  A better control of the coating thickness can be achieved through adjusting the sol aging conditions and the withdrawal speed. However, it remained challenging to simultaneously achieve sufficient thickness and avoid the formation of cracks in the processing of the single-layer coating. The solution employed in this study was to successively apply and process thin layers of Y2Si2O7 coating, each in the thickness range of ~1 µm. The evolution of the coating thickness and surface morphology with the increase of the number of layers has been investigated. The 6-layer crack-92  free coating was able to achieve an average thickness of ~5.5 µm, as well as satisfactory compactness and enhanced adhesion to the substrate. The performance of Y2Si2O7 as EBC for Le-Mark GP has been tested at ~1200 ℃ on a natural gas burner rig and in high concentration water vapor atmosphere. The applied Y2Si2O7 coatings were capable of providing a certain level of corrosion protection in both tests. The possible degradation mechanisms have been discussed. The thicker and well-sintered areas of the coatings demonstrated enhanced corrosion resistance, suggesting that the coating with an adequate thickness and enhanced adherence (like the aforementioned 6-layer coating) remains promising and worthy of further investigation. 93  Bibliography  [1] C. Oprea, F. Wong, H. Karimi Sharif, T. Troczynski, C. Blair, and A. Welch, “Degradation of Silicon Nitride Glow Plugs in Various Environments Part 1: dc Electric Field in Ambient Air,” Int. J. Appl. Ceram. Technol., vol. 8, no. 3, pp. 669–683, May 2011. [2] D. S. Fox, E. J. Opila, Q. N. Nguyen, D. L. Humphrey, and S. M. Lewton, “Paralinear Oxidation of Silicon Nitride in a Water‐Vapor/Oxygen Environment,” J. Am. Ceram. Soc., vol. 86, no. 8, pp. 1256–1261, 2003. [3] H. T. Lin and M. K. Ferber, “Mechanical Reliability Evaluation of Silicon Nitride Ceramic Components after Exposure in Industrial Gas Turbines,” J. Eur. Ceram. Soc., vol. 22, no. 14–15, pp. 2789–2797, Jan. 2002. [4] T. Mitsuoka, “Advanced Ceramic Glow Plugs,” in Ceramics Science and Technology, Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KGaA, 2014, pp. 191–206. [5] BERU, “All about Glow Plugs-Technical Information.” [Online]. Available: http://beru.federalmogul.com/sites/default/files/ti_04_gb_2014_fm.pdf. [Accessed: 09-Nov-2017]. [6] A. Okada, “Automotive and Industrial Applications of Structural Ceramics in Japan,” J. Eur. Ceram. Soc., vol. 28, no. 5, pp. 1097–1104, Jan. 2008. [7] L. Le-Mark Ceramic Technology Co., “Glow Plug for Diesel Engine.” [Online]. Available: http://www.le-mark.cn/E-pages/Product/glow plug.asp. [Accessed: 06-Nov-2017]. [8] M. H. Bocanegra-Bernal and B. Matovic, “Mechanical Properties of Silicon Nitride-based 94  Ceramics and Its Use in Structural Applications at High Temperatures,” Mater. Sci. Eng. A, vol. 527, no. 6, pp. 1314–1338, Mar. 2010. [9] E. J. Opila, D. S. Fox, and N. S. Jacobson, “Mass Spectrometric Identification of Si-O-H(g) Species from the Reaction of Silica with Water Vapor at Atmospheric Pressure,” J. Am. Ceram. Soc., vol. 80, no. 4, pp. 1009–1012, Jan. 1997. [10] S. K. Lee and M. J. Readey, “Development of a Self-Forming Ytterbium Silicate Skin on Silicon Nitride by Controlled Oxidation,” vol. 40, pp. 1435–1440, 2002. [11] C. S. Tedmon, “The Effect of Oxide Volatilization on the Oxidation Kinetics of Cr and Fe-Cr Alloys,” J. Electrochem. Soc., vol. 113, no. 8, p. 766, Aug. 1966. [12] E. J. Opila and R. E. Hann, “Paralinear Oxidation of CVD SiC in Water Vapor,” J. Am. Ceram. Soc., vol. 80, no. 1, pp. 197–205, Jan. 1997. [13] W. Han, Y. Li, G. Chen, and Q. Yang, “Effect of Sintering Additive Composition on Microstructure and Mechanical Properties of Silicon Nitride,” Mater. Sci. Eng. A, vol. 700, pp. 19–24, Jul. 2017. [14] G. Ling and H. Yang, “Pressureless Sintering of Silicon Nitride with Magnesia and Yttria,” Mater. Chem. Phys., vol. 90, no. 1, pp. 31–34, Mar. 2005. [15] J. H. Eom, Y. K. Seo, Y. W. Kim, and S. J. Lee, “Effect of Additive Composition on Mechanical Properties of Pressureless Sintered Silicon Carbide Ceramics Sintered with Alumina, Aluminum Nitride and Yttria,” Met. Mater. Int., vol. 21, no. 3, pp. 525–530, May 2015. [16] N. S. Jacobson, “Corrosion of Silicon-Based Ceramics in Combustion Environments,” J. Am. Ceram. Soc., vol. 76, no. 1, pp. 3–28, Jan. 1993. [17] P. Tatarko, M. Kašiarová, J. Dusza, and P. Šajgalík, “Influence of Rare-earth Oxide 95  Additives on the Oxidation Resistance of Si3N4–SiC Nanocomposites,” J. Eur. Ceram. Soc., vol. 33, no. 12, pp. 2259–2268, 2013. [18] H. Karimi Sharif, “Degradation of Silicon Nitride Glow Plugs in Electric Field-Experiments and Modeling,” The University of British Columbia, 2011. [19] C. Oprea, F. Wong, H. Karimi Sharif, C. Blair, A. Welch, and T. Troczynski, “Degradation of Silicon Nitride Glow Plugs in Various Environments. Part 2: Gas Burner Rig,” Int. J. Appl. Ceram. Technol., vol. 9, no. 2, pp. 259–271, Mar. 2012. [20] C. Oprea, F. Wong, H. Karimi Sharif, C. Blair, A. Welch, and T. Troczynski, “Degradation of Silicon Nitride Glow Plugs in Various Environments Part 3: NGDI Engine,” Int. J. Appl. Ceram. Technol., vol. 9, no. 2, pp. 272–279, Mar. 2012. [21] S. Ramasamy, S. N. Tewari, K. N. Lee, R. T. Bhatt, and D. S. Fox, “Environmental Durability of Slurry Based Mullite-Gadolinium Silicate EBCs on Silicon Carbide,” J. Eur. Ceram. Soc., vol. 31, no. 6, pp. 1123–1130, Jun. 2011. [22] K. N. Lee, D. S. Fox, and N. P. Bansal, “Rare Earth Silicate Environmental Barrier Coatings for SiC/SiC Composites and Si3N4 Ceramics,” J. Eur. Ceram. Soc., vol. 25, no. 10, pp. 1705–1715, 2005. [23] D. Zhu, “Environmental Barrier Coatings for Turbine Engines : Current Status and Future Directions,” pp. 1–49, 2013. [24] Z. Sun, M. Li, and Y. Zhou, “Thermal Properties of Single-phase Y2SiO5,” J. Eur. Ceram. Soc., vol. 29, no. 4, pp. 551–557, Mar. 2009. [25] D. D. Jayaseelan, S. Ueno, T. Ohji, and S. Kanzaki, “Sol-gel Synthesis and Coating of Nanocrystalline Lu2Si2O7 on Si3N4 Substrate,” Mater. Chem. Phys., vol. 84, no. 1, pp. 192–195, Mar. 2004. 96  [26] K. N. Lee, “Key Durability Issues With Mullite-Based Environmental Barrier Coatings for Si-Based Ceramics,” Trans. ASME, vol. 122, no. October 2000, pp. 632–636, Oct. 2000. [27] Y. Xu, X. Hu, F. Xu, and K. Li, “Rare Earth Silicate Environmental Barrier Coatings: Present Status and Prospective,” Ceram. Int., vol. 43, no. 8, pp. 5847–5855, Jun. 2017. [28] K. N. Lee, J. I. Eldridge, and R. C. Robinson, “Residual Stresses and Their Effects on the Durability of Environmental Barrier Coatings for SiC Ceramics,” J. Am. Ceram. Soc., vol. 88, no. 12, pp. 3483–3488, Dec. 2005. [29] N. Maier, K. G. Nickel, and G. Rixecker, “High Temperature Water Vapour Corrosion of Rare Earth Disilicates (Y,Yb,Lu)2Si2O7 in the Presence of Al(OH)3 Impurities,” J. Eur. Ceram. Soc., vol. 27, no. 7, pp. 2705–2713, Jan. 2007. [30] N. S. Jacobson, “Silica Activity Measurements in the Y2O3-SiO2 System and Applications to Modeling of Coating Volatility,” J. Am. Ceram. Soc., vol. 97, no. 6, pp. 1959–1965, Jun. 2014. [31] Z. Sun, Y. Zhou, and M. Li, “Low-temperature Synthesis and Sintering of γ-Y2Si2O7,” J. Mater. Res., vol. 21, no. 6, pp. 1443–1450, Jun. 2006. [32] M. Díaz, I. Garcia-Cano, S. Mello-Castanho, J. S. Moya, and M. A. Rodríguez, “Synthesis of Nanocrystalline Yttrium Disilicate Powder by a Sol-gel Method,” J. Non. Cryst. Solids, vol. 289, no. 1–3, pp. 151–154, Aug. 2001. [33] E. M. Levin, C. R. Robbins, and H. F. McMurdie, Phase Diagrams for Ceramists—1969 Supplement. The American Ceramic Society Inc, 1969. [34] W. M. Haynes, CRC Handbook of Chemistry and Physics. 92nd Edition. CRC Press, 2011. [35] H. J. Seifert et al., “Yttrium Silicate Coatings on Chemical Vapor Deposition-SiC-97  precoated C/C-SiC: Thermodynamic Assessment and High-temperature Investigation,” J. Am. Ceram. Soc., vol. 88, no. 2, pp. 424–430, Feb. 2005. [36] N. Al Nasiri, N. Patra, D. Horlait, D. D. Jayaseelan, and W. E. Lee, “Thermal Properties of Rare-Earth Monosilicates for EBC on Si-based Ceramic Composites,” J. Am. Ceram. Soc., vol. 596, no. 36851, pp. 589–596, 2016. [37] Z. Sun, M. Li, and Y. Zhou, “Recent Progress on Synthesis, Multi-scale Structure, and Properties of Y–Si–O Oxides,” Int. Mater. Rev., vol. 59, no. 7, pp. 357–383, 2014. [38] H. Park, H.-E. Kim, and K. Niihara, “Microstructural Evolution and Mechanical Properties of Si3N4 with Yb2O3 as a Sintering Additive,” J. Am. Ceram. Soc., vol. 80, no. 3, pp. 750–756, Mar. 1997. [39] M. K. Cinibulk, G. Thomas, and S. M. Johnson, “Fabrication and Secondary-Phase Crystallization of Rare-Earth Disilicate-Silicon Nitride Ceramics,” J. Am. Ceram. Soc., vol. 75, no. 8, pp. 2037–2043, Aug. 1992. [40] S. K. Lee, K. A. Koskarian, and M. J. Readey, “Rare Earth Silicate Coating on A Silicon-based Ceramic Component Formed by Controlled Oxidation,” GB2355005A, 2000. [41] E. García, P. Miranzo, and M. I. Osendi, “The Prospect of Y2SiO5-based Materials as Protective Layer in Environmental Barrier Coatings,” J. Therm. Spray Technol., vol. 22, no. 5, pp. 680–689, Jun. 2013. [42] M. Aparicio and A. Durán, “Yttrium Silicate Coatings for Oxidation Protection of Carbon-Silicon Carbide Composites,” J. Am. Ceram. Soc., vol. 83, no. 6, pp. 1351–1355, Dec. 2004. [43] H. J. Seifert et al., “Yttrium Silicate Coatings on Chemical Vapor Deposition-SiC-Precoated C/C-SiC: Thermodynamic Assessment and High-Temperature Investigation,” J. 98  Am. Ceram. Soc., vol. 88, no. 2, pp. 424–430, Feb. 2005. [44] J. Ito and H. Johnson, “Synthesis and Study of Yttrialite,” Am. Mineral., vol. 53, pp. 1940–1952, 1968. [45] J. Felsche, “The Crystal Chemistry of the Rare-earth Silicates,” in Rare Earths. Structure and Bonding, Springer, Berlin, Heidelberg, 1973, pp. 99–197. [46] W. J. Heward, R. Sarrafi-Nour, and Y. Gao, “The Structure of Yttrialite and Its Identification Using Laboratory and Synchrotron-Based Powder X-Ray Diffraction,” J. Am. Ceram. Soc., vol. 92, no. 2, pp. 511–516, Feb. 2009. [47] V. Kahlenberg, W. Wertl, D. M. Többens, R. Kaindl, P. Schuster, and H. Schottenberger, “Rietveld Analysis and Raman Spectroscopic Investigations on α-Y2Si2O7,” Zeitschrift für Anorg. und Allg. Chemie, vol. 634, no. 6–7, pp. 1166–1172, Jun. 2008. [48] K. Liddell and D. P. Thompson, “X-ray Diffraction Data for Yttrium Silicates,” Trans. J. Br. Ceram. Soc., vol. 85, no. 1, pp. 17–22, 1986. [49] H. I. Leonyuk et al., “High-Temperature Crystallization and X-Ray Characterization of Y2SiO5, Y2Si2O7 and LaBSiO5,” J. Cryst. Growth, vol. 205, no. 3, pp. 361–367, 1999. [50] W. Dias, F. P. Glasser, R. P. Gunwardane, and R. A. Howie, “The Crystal Structure of δ-Yttrium Pyrosilicate, δ-Y2Si2O7,” Zeitschrift für Krist., vol. 191, no. 1–2, pp. 117–123, 1990. [51] E. Galunin, M. Vidal, and M. D. Alba, “The Effect of Polymorphic Structure on the Structural and Chemical Stability of Yttrium Disilicates,” Am. Mineral., vol. 96, no. 10, pp. 1512–1520, Oct. 2011. [52] C. Wang, M. Chen, H. Wang, X. Fan, and H. Xia, “Fabrication and Thermal Shock Resistance of Multilayer γ-Y2Si2O7 Environmental Barrier Coating on Porous Si3N4 99  Ceramic,” J. Eur. Ceram. Soc., vol. 36, no. 3, pp. 689–695, 2016. [53] L. L. Hench and J. K. West, “The Sol-Gel Process,” Chem. Rev., vol. 90, no. 1, pp. 33–72, Jan. 1990. [54] C. J. Brinker, “Dip Coating,” in Chemical Solution Deposition of Functional Oxide Thin Films, T. Schneller, R. Waser, M. Kosec, and D. Payne, Eds. Vienna: Springer Vienna, 2013, pp. 233–261. [55] M. Faustini, B. Louis, P. A. Albouy, M. Kuemmel, and D. Grosso, “Preparation of Sol-gel Films by Dip-coating in Extreme Conditions,” J. Phys. Chem. C, vol. 114, no. 17, pp. 7637–7645, May 2010. [56] L. Landau and B. Levich, “Dragging of A Liquid by A Moving Plate,” Acta Phys Chim URSS, vol. 17, pp. 42–54, 1942. [57] Y. Sun, C. Zhao, Y. Zhou, and F. Wang, “A Sol-Gel Synthesis Method for Yb2Si2O7 Powder,” CN102964128A, 2013. [58] N. W. M. Ritchie, D. E. Newbury, and J. M. Davis, “EDS Measurements of X-Ray Intensity at WDS Precision and Accuracy Using a Silicon Drift Detector,” Microsc. Microanal., vol. 18, no. 4, pp. 892–904, Aug. 2012. [59] R. Castaing, “Application of Electron Probes to Local Chemical and Crystallographic Analysis,” University of Paris, 1951. [60] D. E. Newbury and N. W. M. Ritchie, “Is Scanning Electron Microscopy/Energy Dispersive X-ray Spectrometry (SEM/EDS) Quantitative?,” Scanning, vol. 35, no. 3, pp. 141–168, May 2013. [61] D. E. Newbury and N. W. M. Ritchie, “Performing Elemental Microanalysis with High Accuracy and High Precision by Scanning Electron Microscopy/Silicon Drift Detector 100  Energy-Dispersive X-ray Spectrometry (SEM/SDD-EDS).,” J. Mater. Sci., vol. 50, no. 2, pp. 493–518, 2015. [62] C. J. Brinker, “Hydrolysis and Condensation of Silicates: Effects on Structure,” J. Non. Cryst. Solids, vol. 100, no. 1–3, pp. 31–50, Mar. 1988. [63] T. Peltola et al., “Effect of Aging Time of Sol on Structure and in vitro Calcium Phosphate Formation of Sol-gel-derived Titania Films,” J. Biomed. Mater. Res., vol. 51, no. 2, pp. 200–208, Aug. 2000. [64] D. M. Liu, T. Troczynski, and W. J. Tseng, “Aging Effect on the Phase Evolution of Water-based Sol-gel Hydroxyapatite,” Biomaterials, vol. 23, no. 4, pp. 1227–1236, Feb. 2002. [65] K. Wongcharee, M. Brungs, R. Chaplin, Y. J. Hong, R. Pillar, and E. Sizgek, “Sol-Gel Processing by Aging and Pore Creator Addition for Porous Silica Antireflective Coatings,” J. Sol-Gel Sci. Technol., vol. 25, no. 3, pp. 215–221, 2002. [66] H. Wen, S. Dong, P. He, Z. Wang, H. Zhou, and X. Zhang, “Sol-gel Synthesis and Characterization of Ytterbium Silicate Powders,” J. Am. Ceram. Soc., vol. 90, no. 12, pp. 4043–4046, Sep. 2007. [67] P. Melnikov, V. A. Nascimento, L. Z. Z. Consolo, and A. F. Silva, “Mechanism of Thermal Decomposition of Yttrium Nitrate Hexahydrate, Y(NO3)3·6H2O and Modeling of Intermediate Oxynitrates,” J. Therm. Anal. Calorim., vol. 111, no. 1, pp. 115–119, Jan. 2013. [68] D. Nurkowski, P. Buerger, J. Akroyd, and M. Kraft, “A Detailed Kinetic Study of the Thermal Decomposition of Tetraethoxysilane,” Proc. Combust. Inst., vol. 35, no. 2, pp. 2291–2298, Jan. 2015. 101  [69] H. Kozuka, S. Takenaka, H. Tokita, T. Hirano, Y. Higashi, and T. Hamatani, “Stress and Cracks in Gel-Derived Ceramic Coatings and Thick Film Formation,” J. Sol-Gel Sci. Technol., vol. 26, no. 1/3, pp. 681–686, 2003.  102  Appendices  Unrounded original data acquired by EDS analysis, which have been presented in the tables of this thesis after being rounded to 1-2 significant figures  Table 4.1 Content (wt%) Location O Na Mg Al Si K Ca Fe Y Mo as-received GP 37.99   3.02 34.16    2.56 22.28 IS10min-2 mm 46.22   4.14 29.29    3.06 17.29 IS10min-2 mm-A 42.58   2.39 25.18    1.01 28.84 IS10min-2 mm-B 48.90 0.74 0.62 7.01 26.81 0.45 2.14 3.71 4.72 4.89 IS10min-4 mm 45.65 0.36  3.78 29.57    2.76 17.87  Table 4.2 Content (wt%) Location O Al Si Y Mo As-received GP 37.99 3.02 34.16 2.56 22.28 IS24h-1 mm 48.49 3.63 30.08 6.42 11.38 IS24h-2 mm 50.00 3.70 30.77 6.22 9.30 IS24h-4 mm 46.39 4.24 32.31 9.11 7.95 IS24h-6 mm 46.01 3.70 34.42 6.67 9.21 IS24h-8 mm 48.93 2.89 32.32 3.14 12.72  Table 4.3 Content (wt%) Location O Al Si Y Mo IS100h-1 mm 47.30 4.57 32.98 9.45 5.71 IS100h-2 mm 49.75 4.29 32.44 8.71 4.80 IS100h-4 mm 47.27 5.32 33.11 12.11 2.19 IS100h-5 mm 51.02 4.50 31.59 9.63 3.26 IS100h-8 mm 51.23 3.84 34.19 5.71 5.03 103  Table 6.1 Location O  Al Si Y  Mo #500-2mm 40.38 2.55 20.94 35.99 0.14 #100-2mm 42.03 3.82 19.99 33.27 0.89 #50-2mm 38.68 3.20 22.31 32.61 3.20 #20-2mm 43.34 2.64 21.76 28.98 3.29 #10-2mm 45.31 2.84 24.36 18.88 8.61 #5-2mm 41.89 3.63 23.10 27.46 3.93  Table 6.2 Location O  Al Si Y  Mo #500-4mm 39.94 2.67 21.75 35.47 0.17 #100-4mm 41.69 4.25 21.95 30.11 2.00 #50-4mm 41.81 3.25 27.91 16.86 10.17 #20-4mm 40.05 3.57 28.14 20.01 8.23 #10-4mm 43.91 2.98 26.29 15.69 11.13 #5-4mm 38.45 2.19 23.91 30.83 4.62  Table 6.3 Location O  Al Si Y  Mo #500-6mm 38.22 2.11 21.92 37.38 0.38 #100-6mm 41.98 3.89 24.28 25.78 4.07 #50-6mm 40.85 2.66 29.88 14.22 12.40 #20-6mm 42.90 3.04 28.41 14.89 10.77 #10-6mm 43.45 2.31 26.70 14.22 13.32 #5-6mm 38.12 2.17 27.22 24.48 8.01  Table 7.1  O  Al Si Y  Mo 1L-2mm 37.88 1.11 25.01 30.36 5.63 2L-2mm 33.61 0.50 18.79 47.11 0.00 3L-2mm 36.83 1.18 18.16 43.82 0.00 5L-2mm 36.98 1.25 18.16 43.62 0.00 6L-2mm 33.76 1.05 19.25 45.94 0.00   104  Table 7.2  O  Al Si Y  Mo 1L-4mm 39.48 2.15 31.28 15.19 11.89 2L-4mm 36.88 2.07 25.01 32.62 3.42 3L-4mm 41.10 3.05 18.58 36.96 0.31 5L-4mm 37.76 1.84 18.59 41.80 0.00 6L-4mm 36.84 1.51 18.43 43.21 0.00  Table 7.3  O  Al Si Y  Mo 1L-6mm 38.73 2.15 31.18 14.09 13.85 2L-6mm 39.20 1.85 25.01 28.51 5.43 3L-6mm 39.10 2.61 22.09 33.59 2.62 5L-6mm 35.78 1.47 19.20 43.54 0.00 6L-6mm 32.93 1.22 20.08 45.76 0.00   

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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0362389/manifest

Comment

Related Items