THE TUNIVERSITY OF MICHIGArN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING SOME RELATIONSHIPS IN THE SYSTEM ZIRCONIA-FERROUS OXIDE SILICA Ralph G. Wells A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The University of Michigan 1960 Julry, 1960 IP-441

ACKNOWLEDGEMENTS The author wishes to express his gratitude to the several persons who aided him in various ways during the course of this investigation. He wishes especially to thank Professor L. H. Van Vlack for his helpful discussions and constructive criticisms, also for making equipment available Professor L. S. Ramsdell and Professor Lo Thomassen for helpful suggestions in the X-ray diffraction phase of the work. Professor R, M. Denning for helpful discussion on the D.T.Ae phase, Professor E. W.m. Heinrich for interesting and helpful discussions of phase relationships in systems involving zirconia and hafniao He also wishres to thank Mr. Robert Ho Insley of Champion Spark Plug Company Research. Laboratory in Detroit for making available his high temrrpera-'iure X-ray diffraction equipment for my ulse, and to thank Dr. So F: Urban of Titanium Alloys Manufacturing Corporation for making availabl.e enough, high purity zirconia and stabilized zirconia cruci, bles for this investigation. The author is indebted to the Rackhanm (raduate School for a grant which covered the cost of gases used, and chemical analyses necessary in this work. ii

.AB.LE OF CONTIENTS Page ACKNOWLEDGEMENTSo o o o o o o o o o o o o o a o o o o o o o o o o o o o a o o o o o o ii LISjT OF TABLES 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0IV LIST OF FITGRES 0 o o o o 0 o 0 0 0 0 0 0 o 0 0 0 0 0 0 0 0 0 a o 0 o. a 0 0 0 0 a 0 0 0 o o 00 0 V INTRODUCTION 0 0 a 0 0 0 0 0 0 0 0 0 0 0a 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a 0 0 0 0 0 0 0a 0 0 0 0 0 0 0 a 1 MATERIALS AND EXPERIMENTAL WORK000o < o o o o o 000 o o o o o o 6 RESULTS AND DISCUJSSION o o o o o o o... o o o o o o o o o o o.o o 14 Compatibility Relationships o.......oooooooooooOOo 14 Binary Systems and Joins o.o oooo0000OOOO o.000000oo00000OOOo.o 18 Ternary Invariant Points oooooo.o.oooooooooooo ooooooooo000000032 Primary Crystallization Fields 00o0000000000000o000000o00ooO33 Effect of HfO2 and Other Impurities on ZrOJ2 Transformation.., 36 Dissociation of Zircon oooooooo000000 00000000oo 000o000o 00000 41 Immiscible Liquid Regions... o.o00oooooo...o0 00.00004o.0000.o 48 Effects of Ferrous Oxide on Zircon and Zirconia ReFfractories....OOOOOO.OOOOOOOOOOOOOO.aOOOOOOo OoOOOOOO 49 REFERENCES dL o o o o a 0 0 0 0 0 0 0 0 0 0 0 a a a 0 0 O a 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0o o o 5 a 4 a O 0 a 4 iii

LIST OF TABLES Page I Semi-Quantitative Spectrographic Analdyses of Raw Materials Usedo a.oaoooooooooooooooooooooooo.oooooooaooaooooooo 43 II Analysis of Ferrovac "E" 3/8-in. Rod Used for Crucibles and Lids a o a *a a 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a.....0 a0 0 0 0o.... 0 00 a 0 44 III Results of Quenching Experiments9 1020 Steel Specimen 3/8' Square by 3/4" Long, with Thermocouple Sealed in Center~ Temperatures in Degrees Centigrade ooooooooeoo ooooaoaooooo 45 IV Comparison of Ferrous and. Ferric Oxide Content of So.e Specimens Fired in an Atmosphere of 50/50 CO/CO2 with ThQse Fired in Iron Crucibles.oo. oaooooooo.oooo.oooo.ooooooo 46 V DTA Data on Zirconia Monoclinic-Tetragonal Transform.a;ionsoos..o 47 VI Effect of Heating Rate upon ZrO2 Transformation....... o........oo 47 iv

LIST OF FIGURES Page lo Compositions of Mixes Investigatedo..00o,,o...OO.....,,.ooeo.... 7 2. Iron Crucible and Plug from 3/8" Ferrovac'E"' Rodo,,, oo.o.oo O, 9 3. Crucibles and Apparatus for Quenching Experiments..O.,,...oo.. 10 4o Large Iron Crucible and Plugo...o..o,.,,,.oooooooo,,,,oooooooo 11 5o Large Iron Crucibles Loaded, Plugged, and the Tops Welded, Ready to be Firedo,ooo......oo..o.,oo, o oooo...oo.ooo,,.,,, ooo,. 12 6, Furnace Temperature Measuring, Quenching. and Atmosphere Apparatus O,.. a.......o4 a a * o.4o o 0 0o a..... a.0 0 o 0. 13 7- Compatibility in the System FeO-,ZrO2-SiO o o o o.... 16 80 The System FeO-ZrO2-SiO20...o...0.0000000............OOO.ooooo ooo 17 9. Photomicrographs Illustrating the Three Compatibility Regions Present at 1200~Coooooo.......ooooo..ooooooooooooo..ooooo 19 10o Equilibrium Diagram of the System FeO-SiO-SiO from No L, Bowen, Jo Fo Schairer........o.o.o.,.o.oo.oooooaoooooo.oo.o..o 20 11 The System FeO-ZrO2 according to W. A, Fischer and A, Hoffman,,o 21 12, Mix 26 (Zr02 + 5% FeO) o,.0 o o.o, o. oo..o...0o........o.. 24 130 ZrO2-SiO2 from Ro Fo Geller, S. Mo Lang, Nato Buro of Standards o.o........,Oo o. o..o o 0 0 0 a....... 0 o 0 0 26 14, Constitution Diagram of the ZrO2-SiO2 System from Co Eo Curtis2 Ho Go Sowmanooo,.o..,0,o..oooooooo 4 ooooooooooo 27 15o The System Zr02-Si02 according to No Ao Toropov and Fo Jo Galakov, Izvestiao........ o..o.....,..ooo0000000000o00000, 28 16o The System Fayalite-Zirconia...o..o..ooooo..... oo...oo ooo...o.. 30 17. Photomicrographs of Compositions in Fayalite-Baddeleyite System, o.000400004.000.0000000 o a 0..000000000ao0.00a.0a0a0.0 31 v

LIST OF FIGURES (CONT D) Page 18o Iron Content of Samples Fired at 1200~C...................0.... 35 19o Mix 17, Effects of Loading Technique 00.................,......0 37 20O Effects of Impurities on Zirconia Phase Transformations, DTA Curves, Endothermic on Heating, Exothermic on Cooling.O,, O..O, 40 21o Schematic Free-Energy versus Temperature of Beginning of Transformation for ZrO2po Oo oo oooooooooooooo0ooooo, OO o o 42 vi

SOME RELATIONSHIPS IN THE SYSTEM ZrO2-FeO-SiO2 INTRODUCTION Although there is little, if any, geological interest in the system zircbnia-ferrous oxide-silica, this system is becoming of increasing importance in metallurgical and ceramic fields, With the increasing use of zircon and zirconia refractories at high temperatures, under conditions in which contact with slags rich in ferrous oxide might be expected, it is important to learn more about this system. One application in which the above conditions are met is in the foundry. In certain types of castings in which surface condition and dimensional tolerance are very important, shell molds made of zircon sand are used. Although in general little slag as such is encountered in this operation, ferrous oxide may be formed at the interface between the molten metal and the mold, This ferrous oxide may be in contact with the refractory for periods of time up to several minutes or more at temperatures in excess of 1200~Co Another application in which zircon refractories have been used is in the front and back walls of open hearth furnaces. Here the refractory may be in contact with slags of varying ferrous iron content for long periods of time also at temperatures above 1200~C. Zirconia refractories, as well as zircon refractories, are used in induction furnace linings, laboratory furnace muffles, crucibles, and other kiln furniture, all at times may be under conditions in which ferrous oxide is encountered at high temperatures.

In general, three types of information about the ZrO2-FeO-SiO2 system will be useful for the judicious use of zircon and zirconia refractories: (1) phase equilibrium relationships in the binary and ternary systems; (2) data on the kinetics of reaction and phase changes in the system; (3) thermodynamic information such as activity coefficients, heats of formation, free energy relationships. The present work is concerned with the first of these three general areas. A survey of the literature has shown that the three component oxides and the two normally stable binary compounds have been well investigated. There are still a few regions in which agreement is not complete) but except for these few the agreement is good. The three limiting binary systems have been investigated, but some questions still remain~ The silica phases in the one component system SiO2 have been under investigation for many years by many investigators and there are still a number of unresolved questions about this intriguing oxideo However, except to note that even in this well, investigated one component system work remains to be done in order that a complete understanding may be gained, further discussion will not be made, Wustite (FeO) has also been studied by many investigators from many viewpoints (1-5)*o It was early discovered that the stoichiometric (1) compound FeO does not exist o There is always a deficiency of iron positions in the lattice with electrical neutrality being maintained by * References at end of papero

the presence of ferric as well as ferrous ionso This, of course, complicates any system in which wustite is a component, for it means that ferric iron as well as ferrous iron can exist in equilibrium with metallic iron, It also means that wustite can be considered a single component only if the ferric iron is ignored. The iron-oxygen system as reported by Darken and Gurry(3) is that which is most universally acceptedo They show that the composition of wustite can vary from 23,2 percent oxygen to 25o5 percent oxygen at 1375~Co The composition at 560~C is 2353 percent oxygen, below which wustite can exist only in. a metastable condition0 Consequently, the composition of wustite in any particular specimen will depend upon a number of factors such as the temperature from which it was quenched and the atmosphere in which it was heated and quenched0 Zirconia (ZrO2), which occurs in nature as the monoclinic mineral. baddeleyite, has also been investigated by a number of workers, some of which are listed in the bibliography 62) Some questions still remain in certain aspects of the monoclinic-tetragonal transformation, and on (21) the existence of a cubic phase above 1900~Co Blumenthal' gives the names "ruffite" to the tetragonal phase, and "arkelite" to the cubic phase. However, since these phases have not been found in nature, their use as mineral names should be discouragedo From the work of Curtis, Donney, and Johnson(22), it appears that the high temperature cubic phase can be present only if there is some impurity present which can cause this phase to prevail. In the high purity zirconia they found the

tetragonal phase to be stable from just above 1000~C all the way to the melting point at 2700~Co The three limiting binary systems FeO-Si02, ZrO2-Si02, and FeO-ZrO2 have all been investigated to some extent, The work of Bowen and Schairer () in the system FeO-SiO2 has remained essentially unchallenged since it was published in 1932o Part or all of it has been rechecked by a number of workers, most recently by Schuhmann and Ensio(23) and by Allen and Snow (24) who showed a flexure in the wustite liquidus but made no other changes. The system contains one stable binary phase fayalite (Fe2Si04) and a region of liquid immiscibility on the silica-rich endo Another binary phase "ferrosilite" (FeSiO3) is reported in this system by Bowen(25) but no region of stability has been positively shown for this phase, Bowen was not able to cause this phase to precipitate out of a (26) glass of the stoichiometric composition. Schairer and Yagi(2 also Roedder7) had results similar to Bowen's. Rigby(2), however, reported ferrosilite may be encountered when a 1:2 ferrous oxide-silica mixture is heated to 1350~Co Ferrosilite was not encountered under the conditions of the present investigation~ In the Zr02OSi02 system there is somewhat less agreement among the various investigators, However9 since this will be discussed further in a later section, an attempt will not be made to pursue it in detail at this point0 In this system there is one binary oxide compound zircon (ZrSiO4) The effects of ferrous oxide, either as such2 or, combined with silica

-5~ upon this compound forms one of the major parts of this investigation, The system FeO-ZrO2 has been studied to a lesser extent than any of (20) the other limiting systems. Recently, Fischer and Hoffmann", in Germany, have presented a diagram of the system in the temperature range of from 1300~C to 18000C which includes the eutectic region. Since the present results did not exactly match their results, further discussion will be postponed until a later section. No compounds have been found in this system. Therefore, in the three limiting systems, only two binary compounds, fayalite and zircon, are found which have proven regions of stability. This investigation is chiefly concerned with relationships in the ternary system FeO-Zr02-Si02o The following general areas have been considered: (1) compatibility relationships in the ternary system, (2) binary joins, (3) ternary compounds, (4) ternary invariant points, (5) fields of primary crystallization, (6) effect of HfO2 and other impurities on the monoclinic-tetragonal transformation of ZrO2, (7) dissociation of zircon and (8) region of liquid immiscibilityo

-6MATERIALS AND EXPERIMENTAL WORK The sources and compositions of the materials used in this investigation are summarized in Table Io Figure 1 shows the mix compositions investigated. Most of the experimental work was done in one of two furnaceso Both are horizontal tube furnaces heated with silicon carbide elementso One furnace was controlled by a total off-on type controller, and the other furance was controlled by a saturable reactor unito The compatability studies and the preliminary field boundaries were done in the former furnaceo The work demanding more precise temperature control was done in the furnace controlled by the saturable reactoro Because wustite is not stable in an air atmosphere, and becomes oxidized to magnetite it air at elevated temperature, it was necessary to fire the samples in an environment in which wustite is stableo This was accomplished in one Of two ways. (1) In the compatibility studies, discs one centimeter in diameter by about 5 millimeters thick were fired in an atmosphere of 50 percent carbon momoxide and 50 percent carbon dioxide. It has been established that wustite is stable in this atmosphere in the temperature ranges which were used (2) (2) Wustite is compatible with iron, therefore iron crucibles could be used for temperatures up to 1500~Co Since all of the invariant

SiO2 _ E 38 40x x41 to\4 x43 x 21 H C x 44 x 34 x 52 19 18 x ZrSiO4 FeSiO4 17 5 4 x 3 x 50 22=0.01% FeO 23=0.10% FeO x 2 FeO 1 A 12 11 10 26 Fig-Lre 1. Compositions of Mixes Investigated.

points investigated in this ternary system are well below 1500~C, iron crucibles with tightly fitting plugs were used for determining these points and for locating phase field boundaries (Figures 2 through 6). Figure 6 shows the furance set-up used in the quenching experiments and for the differential thermal analysis worko Table I gives the chemical analysis of the Ferrovac "E" rod used for the crucibles and plugs. In the runs where the charge became mostly molten, it was found to be necessary to weld around the top of the crucible and plug to keep the liquid in the crucibleo It was also found necessary to press the plugs into the crucible in an inert atmosphere to preclude the possibility of increasing the FeO content of the charge because of reaction of entrapped oxygen with the iron crucibleo This was accomplished by enclosing the loaded crucible, with the plug to one side, in a pliable polyethylene bago A piece of vacuum tubing and a short length of steel rod were inserted into the opening of the bag and the opening sealed shut around them. The bag was evacuated and flushed repeatedly (at least five times) with purified argon, or heliumo The gas was purified by passing it over calcium chips at 1000~C and copper gauze at 600~Co The plug was placed on the crucible and pressed into place with the steel rod in a Carver press. The crucible was then removed from the bago In order to determine the most effective quenching medium, a

-9 Figure 2. Iron Crucible and Plug from 3/8" Ferrovac "E" Rod.

-10 Figure 3. Crucibles and Apparatus for Quenching Experiments. Empty crucibles and plugs above. Platinum vessel, thermocouple, and loaded crucible ready to be placed in the furnace, below.

-11-..:.::::::......::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~............... E::::-::: i::: iE:::i E i tiL::::::::;:::: D: 0 f:::::::0000:::i; i;00;iC:; f:?;::::::;:;:;::;;f;:................. i:: E: y::::::::: iR E f i::E:::::E:::::: i:::y:EEE::::;::.............!:E:::: iR E i f:::::::: *:: E::!iE;:::E::~~~~~~~~~~~~~~~~~...........:i::::;:::: E:: i::::::::::::::::::E: i::::;::i:::::; i: i:::::::::::!0::;i.............:E:::::: E::: E::::::::::::: E:::::: E i:::; i E i::::::::E::........:.: i!:i~~fE:::::::::::: E: i: iE:::::::::::::: i:::E:::E:::::::::::::................... ifi:::;f f:::;:;::::;;::::::::::::;;;00::::;:::::::::0::::y:;:;::::::::::::;:::;:: i::;;0:::!:::L:;:::E::;;:;::0:....::.........0;i;i~ i:i:i::i:::i i::i iiii:iii::I:::i::i::: i::::I:::::ilii:: i iii:i:ii::iiiii i:ii Figure 4. Large Iron Crucible and Plug. Crucibles of this type were used to make samples for chemical, X-ray, and microscopic analysis.

-12-..... Figure 5. Large Iron Crucibles Loaded, Plugged, and the Tops Welded Ready to be Fired.

-13-::::::: i~. Figure 6. Furnace, Temperature Measuring, Quenching, and Atmosphere Apparatus for the Quenching Experiments.

-14 series of tests was performed by embedding a thermocouple in the center of a piece of iron of about the same mass as the crucible. This piece of iron was heated to 1200~C then quenched in the mediumo Various media were checked including water, mercury(specimen submerged), and mercury (specimen floating), various oils and gases. It was found that the most severe quenching was obtained with water. The tests were repeated by embedding the thermocouple in a piece of silica bricko Again it was found that water gave the most severe quench. This information is summarized in Table III, The study of the immiscible liquid regions was accomplished by pressing small cylinders of the appropriate compositions (1/8 inch in diameter by approximately 1/8 inch high) in a steel die. The top surface of the cylinders was melted in a natural gas-oxygen flame and quenched with a stream of water. Phase identification in this investigation was accomplished by means of transmitted and reflected light microscopy() and by X-ray diffraction. RESULTS AND DISCUSSION Compatibility Relationships In the ternary system FeO-ZrO2-Si02, if we assume first that there are no ternary compounds, there are at most three possible compatibility

regions Because of the dissociation of zircon in the solid state, it seems likely that there will. be only two regions at the liquidus surface. These two will then most probably be FeO-Fe2SiO4-ZrO2 and ZrO2-Fe2SiO4-SiO2, The present compatibility study has verified these expectations by showing that there are no ternary compounds in the system and that zirconia and fayalite are compatible forming a binary system, Below about 1550~C fayalite and zircon are also compatible and three compatibility regions are found, These three regions are FeO-Fe2SiO4-ZrO2, Fe2SiO4-ZrSiO4-ZrO2, and Fe2SiO4-ZrSiO4-SiO2. It may be noted that fayalite is at one apex of each compatibility triangle. The two compatibility systems are shown in Figure 7. The overall results are summarized in Figure 8, The arrangement of the three compatibility triangles in the ternary system has a direct effect upon the behavior of zircon sand in the presence of ferrous oxide, It may be seen from Figure 8 that an addition of approximately 15 percent of wustite to zircon would cause a marked dissociation of the zircon at, for instance, 1200~C with the formation of fayaliteo It may also be seen that if about 44 percent of wustite is added to zircon at 1200~C the zircon would be completely dissociated with about 62 percent fayalite and 38 percent of baddeleyite resulting on coolingo Figures 9-A, B, and C illustrate the phase compositions which are found at three points along the line between zircon and wustiteo These represent two of the three compatability triangles. The third is

-16Si02 FeO - Fed FeO ZrO2 (a.) Below 1550~C ZrSiO4 (b.) Above 1550~C Figure 7. Compatibility in the System FeO-ZrO2-SiO2.

Cohpo sition Phase Co ositMelting FeO ZrO2 Si02 Point ~C FeO 100 - D, 1580 SiO - - 100 1715 ~ 5 ZrO2 - 100 - 2700 Fe2SiO4 71 - 29 1205 + 2 ZrSi04 - 67 33 D 1530 SiO2 E F /Y A^A~~~~ Other Invariant Points Desig- Composition nation FeO ZrC Si02 Temperature A 97 5 - 1520 ~ 5 B 76 - 24 1177 + 5 C 62 - 38 1178 + 2 D 42 - 58 1690 + 10 E 3 - 97 1690 ~ 10 F - 5 97,1675 G - 38 62 ~2250 H - 59 41 42250 J 71 7 22 1157 4 K 67 4 29 1182 + 4 L 66 4 30 1180 4 M 58 4 38 1166 4 See Figure 1 for mix compositions. / X \/ v \G DA A A 7\ /\~~~~~~~ 1 H -s~~~~~~~~~~~ ~~~H 1500 C / /1400 A 1300 / \ -/ /*-y/^n -1200/ _\ --— \ - ~ "7 " ZrSiO4 Fe2SiO4 f B \ I'l // BA A V \/ \/ V\/ /rconia ~2 zirconia I I _-~~~~= FeO A Zr02 Figure 8. The System FeO.ZrO2-Si02

represented by Figure 9-Do The compositions in Figure 9 have all been heated to 1200C in an atmosphere of 50/50 CO/C02 and quenched in water, None was completely melted, although of course some contained more liquid at this temperature than others. For instance, the diagram for the system, Figure 8, shows that the composition of mix 2 (Figure 9-A) would be approximately 1/3 liquid at 12000C while mix 4 would contain about half that amount of liquid at the same temperature. The photomicrographs show the phases present, but do not necessarily show the proper relative phase quantities for the mix composition as a whole. Binary Systems and Joins In the ternary system there are four binary systems, the three limiting binary systems, and the join ZrO2-Fe2SiO4o The join Fe2SiO4-ZrSiO4 is not a binary system at all temperatures because of the dissociation of zircon in the solid state to ZrO2 and Si02o The limiting systems have all, been studied previously although there is not universal agreement on all points, The FeO-SiO2 system was worked out by Bowen and Schairer( ) (shown in Figure 10)o The system FeO-ZrO2 (29) as reported recently by Fischer and Hoffman is shown in. Figure 11. These authors report a binary system with a single eutectic at about 3 percent of ZrO2 and at 1330 + 15~C. They also report a solubility of about 4 percent of FeO in the tetragonal form of ZrO2 at 14500~C

CD tJ lzYi -' ~'~~~~~~~~~~~~~~~~~~~~~~~~~~-:~:~:~.... 0 c-F cl) Hc-F 0........., "d -. p-:-q::::::::p::'' I ",-ili I~' t9 -—'O::-::iV 0- m d) HY N ~.::::..:...1,.:~,s..,.:..,..,..i.,. (D::i::::'0'ii iiiiilli:i.i~;j —......"'.1. I Iz?. I ~ 1 c~ H iJ~ c-I-.....V:::-::.... ~~~.::::::::::::::::::::::j::i:i::::::l::::::::::::::::: I: -: Iii:::i::l ii::.i-.I ~ ~ ~ ~ ~ ~ ~ - - X I-m-sX..;~~~~~~~~lil::illi~iji:......-...... - il::i::i~ - "Ill.- -,.,~,:::ll~,:,::: ceF F ~~~~~~~~~~~~~~~~~~~~~~'..- IjI rm. I-~.::........................... -i I i-. CD ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ O iiiiiiiiii~~~~~~~i~i~~ili::.:,.. - a -..i,. i1:11 -:::::::~::::::::-:-,' ~~~_-.. ('D CD DO!'i1..,.,.,.,......................... ~...... - --........ 11,.......'' - - - -,....,".1.1..... —.:i:l::::lili:::::i;;i.::. ~:~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~.......... I:-,....-.1.... - i...:..:i~:''"''i~ l I::i.::ii iii:: i:ii ii~i'~:l''1:::.I::::::::::::::i:i::::::::::::::::::::::::::::::::::::::::::::........::i:: i:1:i:1ii.i ili::i~:~~~: iiil::i:..:l: CD-.. r.4...- -..4:!l. i'~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~ii-'iiiiiiiiliiill; i 0iiijj~;;;: - C...M..... l. I~ c-tD ii::ii::~:.~~;:~:~~:~:~:~:::~-:~::~::~:~:.~:i:~:~:~::-.-.1'......::i:'~i-i:i~i:ii::iiii::l::'::::: ~. Iiii',:::: iii::.. ii,::':::::~1::::: i ":ii i ii?,iii:::::::.:.:,:.':_. ~ ":1":::..... I~i I:.~::iiiiiii ~!::ii~'~ t I ~......... I —] ~' " ~ ~ ~ ~ ~ ~~ ~ ~ ~~~~~~~~~iiiilllll'~li:i:ji::!i~'i:l::::i:::!i:iii......._::: cD ~~~~~~~~~~~g lii~~~~~~~~~~~~::::: i::::::::ii i:.:....:;::-::-::::,::.1..-........ - I,::::::::::::i::iii: -:::: I....I.........:::::::::::::::,...".......,..,.,..,...:,.:''..':l.:.".'." I.. 11 -- ~ ~ ~ ~ ~ ~ ~ k: 1,....ll, —,,3...::::i L ii~~~~ieiiiiiiiiiiiiiiiiiiiili~~~~~~~~~~~~~~~~iiiiiiiii ii~:::::::::::'lii''::l" ":'::':::::I'~~~~~~~~~~~~..~:i:8:Z~~::~iii? (1) 0 CI::ii.,)~:::::::::::),!i::": i~::::;~~'i::::::'-. -.'"l:::::::::.i::!:i~,i;'.C......::::'""!':11i:i...........................~ili:ii?,i?:':~:":'!i::i:i:'::if-iiiii',ii~:~i:f!!'::' ~ ]~jii:iii::~:,{::...................;:,:,.:.::,:'.::::":::'::~.::..:::;,!~ii::................::.......". -.....'.....:.......... I............ D -.::.,::.::.i;::.::.;:::.:::I::..:.......... oh..... I'.... -........'..: X.' -..... I.. 11-:I',.I - i:,i:: i.:: - -..:: ":::':~':::: 1.1-...: ii:iii:iil:l:: i:;i:::: 1:::::::::::. "::~.:::.:::::: 1 1. 1. 1 I.... 1.1 —.2,......:~.::::?!:i':iI I~~i'l::i::~i...::::s:~:.....''' I'l l. - - 11 -:'..'','',.:::::::::::::::::,:::::::::::::,:.::.:::::1............'... -::.:m... "....:::: ft"i.):: i:-:::;:::: - I I;:::":::: -1. I I I':: I. -. I - X --.:1.11....~...............~...:::::::::::':::"::::::/:::::;:,:::::::::!...::-.......................'.-.. -,:::i::::::i:::i:~:'::....:: i -...:',, I..... - -..... -.. 1',::::: -::::.,: i::il:: iiii......'::::i1.'l'l:'::i':i~i i'::: ~:: 11-11...::.... 1:-:..:....:;:,,i1i:::ll1 -.I~::::::::::i:..:.;::::.::~::.....::::::::l:..o........,. I........ Ii:.~ ~ ~ ~ ~ ~ ~~~~~~:;.........1 -~... ":::~:~:~i~iif:!~i ~: i~:~.~::"::::::::::::::: i::j j::::j::l:: I::':............... -......"::-A.011::::::: 14,:::::':::i::::::'i:::...:::......''. - - I iiii -::::::::::i' i!::i::::' c t ~i l~liiii~ii;:i~ir1'':i.:il: - - ii.. "::I':':'.. - X....'.:"ili':'i': I........::. i i:. iril::q: i~ lii:::. -. i I.l 1.11,,:: I'll.- I'll''.1 - 1 I- -,....~:~::::: i l::;:i':'::, - ------ I... - -,........,~~~~~~~~~~~ "',... 11 - ~ ~ ~ ~ ~ ~ ~ ~ I ~ ~ ~':': -::::............1 -1i.: lil: iiil:1: i:1.::.:ii L:::::::::::::::::.:..,:.11 - 1 1 1... - l'..'' 1 I...111 11 1 I.~ -' I'. -I l — d.:.....:.............:.:''.:.,......,,:,:!:::..:..'''I - 11:::::::::'::::::.::,11 -1Il~ iii:E- ~ i: ii:.... i. *iiiiii::::..... —.:::::::::::-::::::',' -..::::::::::: ~::::: 1:: 1:::-:::::::: H 1:,... C~~~~~~~~~~~~~~~~~~~~~~~'.I...- I.. 1..I, ~ ~ ~ ~~~~J ~..1I', ~'''.-:.~I..1.11111: Y ~ ~ ~ -..11,....0 c-3 ~'....1I - 11. ~.11.. I I I... - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.....,. ~ ~ ~ ~ ~ ~ ~ ~:iiiiiii: li.!!i~~~~~~~~~~~~~~~~~~~~~!~~~~~~~~~~~!! -':'..:-':':: - -11 11 - -... 1.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~::::i~~~: ~:~-~il C+:.... 11.. -.1-1.....-........ -,,::-..............~ ~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:i~~ bJ~~~~~~~~~~..1 -W -,14-1 111 Q3 d~~~~~~~m,..1:::::-... —I-,...I-111'.` ~~:~~' 1.1.111 1,... 1 1 -''.:.:::::v:....I 1-:':'':':::'':: I tJ~ ~ ~ jir~lilili' iiiii::.I..',.. I.-.... 11.1'. 11::::::::.:.::- 1..' C C,::::::::. - I I-::::::::'. -1 - ~::::;::::'.'."'.'..l~~~~~~~~~~~~~~~~l'v....:::;::::::::::::::::'::: - -A ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~':i::",-' I-i'l- -1 - Iiiili l:P.::::::;:.:::.:.",.:,.~~~~,:.:,::::,.,:,..:::"''~~~...... 11"'. -.,,,.::: -......':::-l... 1.1- - l -..........-1:::.::::::::::::::..:.........,...,.........'.'',.,'', I, 11.,:::::: -::::.:'::: _ - -.- 11.. ii..,~...,",..,....,..:::::::::::..':.','. " I l..::::::........11,11, ~~~~~~~~~~~~..'',,:: I I..... 11:::.. -:::.:::::':':''~~~~ii~i i;,::i::: / 1 0 9 I~~~~~ii Ij:~~~~j'''.'... - i1...I.......:.'... 111....~....,:::::::::::::::::::::I' —::::'.'.-:-:::':,:,:':':.~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:i. I _.......''i::l:I':::''I::''iiiiX~ji-:iiiii. I'll: ~ i~i~~l:iiI'll..;~l:~i~i;:::~ ~::::::::i::::::::::::;::::::::::::::::::::,:::::::::........ I llll-~~~~~~~~~~~~~~~~~~~~~~~~~~~'~~'-.''I'll........ ~~~~~~~~~~~~'........I..... 11..............~~~~~~~~~~~~~~~~~~~...... ~i::i II'll, - i::::....."...'........'.2%...-.... 1 _. -...'i':: V.......''''.::a I c CD ct i;:i?.':::.~~~~~~~~~~~~~~^:::::i::: i:~~~~~~~~~....:iki~~~~~~ll L ~:::~~~~~~~~.::::..1-I.'. - -::;:::::::: iii::::::::~~~~~~.1::~.~':~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~s.. -... -I. 1i:.1 1 - -:::.::: -.:::I-...'.:::::::::..''....."'..,::::::....:....... 1 -.. I' - "::::::::-,. o ii:::i:: i::::.::::::::,:::::::. 1,..::::: r o ~ ~ ~ ~ ~ -.11.::'~:,i::;'''''''..'.....:"',,a'.,. I..., 11, 111'.,.. 1.::...... - -.1..... 11'..l:. -..............I........:. $ - - -:.:::::: j:. I'll... Ii..''.. l u o:::i~~i~llii lil::-ii::., - -,::::::::::,::. i......': _ _:::::::-: -:::: 11..... - -... I -1.1.1.1.1 11 - II. 11 - -. - -., 11 I.-::::::::::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~: - - I.- - - -1-1 11~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~::::I:::::I -6d

-20 18oo 16oo U 0 or E 40oo 1200 SiO2 Figure 10. Equilibrium Diagram of the System FeO-SiO from N. L. Bowen, J. F. Schairer. Am. Jour. Sci., 4, 177-213 (1932).

2700 2500 2300 I H U N? 2100 1900 1700 1500 I \D 1-' 1300 FeO 0 20 30 40 50 60 70 80 WEIGHT PERCENT Zr02 Figure 11. The System FeO-ZrO2 according to W. A. Fischer and A. Hoffman. Arch. Eisenhuttenw., 28, 739-743 (1957).

-22 Quenching experiments to check this work have shown the eutectic to be less than 5 percent ZrO2 and probably about 3 percent as they reported. The eutectic temperature, however, was found to be 1323 + 3~Co Careful measurements in the back reflection region of mixes of ZrO2 with concentrations of 0Ol, olO, Oo50, 100O and 5O00 percent of FeO, quenched from 1500~C after 1 hour showed only the monoclinic phase with no detectible shift in lattice lines. No evidence of exsolution was observed in the microsections examined by optical, and electron microscopes. Fischer and Hoffman based their conclusion of the solubility of FeO in ZrO2 at 1800~C upon the observation that after heating a sample of ZrO2 containing 3 percent of FeO at 18000C in argon and quenching in water, no wustite was detectable microscopically or by means of X-ray diffraction. They state further that this "solid solution" does not decompose upon heating at 1350~C to 1450~C for several hours. Their observations do not necessarily mean that the FeO has gone into solid solution, however, for at temperatures of 1800~C in an atmosphere of argon it is entirely possible that the FeO could have been lost by vaporization, or by the draining away of liquid oxide. On the other hand, if the FeO had gone into solution in the tetragonal phase as they suggest; and there is very little or no solubility of FeO in the monoclinic phase, as they also suggest, then there should be evidence of an exsolution phase in the zirconia grainso

In order to check this possibility, a mixture of ZrO2 with 5 percent by weight of wustite was pressed into a cylinder 1/8 inch in diameter by 1/4 inch high at 5000 psi, placed in a stabilized zirconia crucible in a graphite susceptor and heated in a nitrogen atmosphere by induction to a little over 1800~C for 1 hour. The sample was allowed to cool in the furnace. Although the sample cooled to a black heat in a little less than 10 minutes any exsolution phase formed would be large enough to be visible under the microscope. No evidence of exsolution was observed at a magnification of 1000o The sample, when it was removed from the crucible, was filled with cracks parallel to the long axis of the cylinder, but considerable sintering had taken placeo Several pieces of the specimen were mounted in bakelite and polished for microscopic examination, and powder specimens were prepared for X-ray diffraction analysiso Microscopic examination of the polished specimen showed that complete recrystallization of the zirconia had occurred. In samples fired at temperatures up to 1.600~C the zirconia grains appeared as rounded to irregular particles in a matrix which appeared smooth, or occasionally contained a eutectic structure. In the sample fired to 18000~C the zirconia grains appear as small cigar shaped particles in a smooth matrix, the reflectivity of which appears to be nearer to that of zirconia than that of the samples heated at lower temperatureso (See Figure 12.) This can be explained either as an.increase of reflectivity of the zirconia because of iron in

Figure 12. Mix 26 (ZrO2 + 5% FeO). 1500~C for 1 hour. 1000X Magnification

-25 solution in the zirconia or as in increase of zirconia in solution in the matrix phase which would lower the reflectivity of the matrix making it nearer that of the zirconia, Electron microscope examination of polished surfaces of the material fired at 1800~C for 1 hour and furnace cooled revealed no evidence of a precipitate within the zirconia grains, confirming light-microscope examinations. X-ray diffraction patterns showed only the monoclinic zirconia lines with no detectable shift in lattice lines. Thus, it seems doubtful that any appreciable solubility of FeO in ZrO2 exists up to 18000~C A number of studies have been made of the system ZrO2-SiO2 of which three seem the most deserving of mention, Gellar and Lang(30) Curtis and Sowman (31) and Toropov and Galakov (3)Figures 13, 14 and 15)o Gellar and Lang reported a eutectic at about 3 percent ZrO2 and at 1675~Co They show zircon to dissociate to ZrO2 plus a liquid at about 1775~C and a solubility of SiO2 in ZrO2 of approximately 10 percent by weight. Earlier work by Washburn and Libman(33) and by Zirnova(34) had reported a true melting temperature for zircon at 25500C and 2430~C respectively, Curtis and Sowman show the eutectic at 3 percent ZrO2 and 1675~C and also the SiO2 solubility in ZrO2 at 10 percent, but they show zircon to dissociate to ZrO2 and silica in the solid state at about 1530~C. Toropov and Galakov agree with Curtis and Sowman on all of these points, but they report a region of two immiscible liquids in the region of 62 percent to 41 percent of Si02 at 22500C. Barlett(34) also

-26 2800 2600 2400 0 Co Eq 2200 2000 I ^ ~-_2700~ I I _ I I I + _ I I I Liquid (L) - 1:1 I I I I I I I I I I I I I N i8oo 16oo 1775~ + 10~ I - I I 0 - I I I ZrO2 SS + ZrSiO4 I I I ZrSiO4 + Liquid 1675~ + 5~ Si02 + - \\~ L \ \ \.0_ 0 \_ -e - ZrSiO4 + SiO2 88% I I I i I I IL 1713~ ZrO2 10 20 30 40 50 60 70 80 90 siO2 Figure 13. ZrO2-SiO2 from R. F. Geller, S. M. Lang, Nat. Bur. of Standards. Revised and corrected phase diagram, Phase Diagrams for Ceramists (1956), figure 132, p. 67.

-27 2000 1900 1800 0 o EA 1700 I I I I I I I I I I I I I I I I L^ ZrO2 + Liquid 1600 1500 I I I l Zr02 - I -d + I I r -- z I ZrSiOq ZrO2 + Sio2 SiO2 + Liquid ZrSiO4 + SiO2 I I I I I I Zr02 10 20 ZrSiO4 40 50 60 70 80 90 SiO2 Figure 14. Constitution Diagram of the C. E. Curtis, H. G. Sowman. Zr20-Si2 System from Jour. Am. Ceram. Soc., 36, 6198 (1953).

2700 2500 2300 - \ Zr02 S.S. - \ liquid \%^<~~ //^^ ~~~two \ immiscible 2\50\ / liquids 2250~\/ \ 2100 1900 1700 1500 1 -nn I I - mI I I - I I I I I I, 1662~ ZrO2 S.S. + liquid Liquid Liquid I r3 oo I ZrO2 S.S' ZrO I.ZrO2 S.S. + Si02 1540~ IZrO2 S.S. + ZrSiO4 ZrSiO4 + SiO2 I I I z-: rO2 10 20 30 40 50 60 70 80 90 WEIGHT PERCENT SiO2 Figure 15. The System ZrO2-SiO2 according to N. A. Toropov and F. J. Galakov, Izvestia. Akad. Nauk. SSSR, Otdel. Khim. Nauk. No. 2, 160 (1956). Si02

-29reported a two liquid region but did not attempt to define the limitso The present study confirms the dissociation of zircon to ZrO2 and silica at temperatures above 1550~CO An attempt to find the two phase region by melting the proper compositions in an oxy-gas flame and quenching in a stream of water was unsuccessful although the two liquid regions in the FeO-SiO2 and the MgO-SiO2 systems could be observed by this method. In the ternary system, the join between zirconia and fayalite is a binary system, This system has not been previously reported in the literature. In this study it was found to have a eutectic at about 4 percent ZrO2 and at 1182 + 4~Co Because the melting temperature of the iron crucibles is about 1530~C, the liquidus surface above 1500~C was not determined. This system was investigated by quenching experiments as described aboveo The compositions which were studied are shown in Figure 16o An illustration of the appearance of the microstructure in samples quenched from below the eutectic and from above the eutectic is given in Figure 17-A and 17-B. In 17-A the phases present are essentially those of the mixed Lnd pressed specimen with some solid state sintering possibleo In 17-B, the primary crystals of zirconia and the eutectic of fayalite and zirconia are evident, By this method the eutectic temperature was determined to be 1182 + 4~Co Figures 17-C and 17-D also show the primary crystals of zirconia at 5 percent and 10 percent of ZrO2a At 2o5 percent of ZrO2 only the

2700 2500.- - 2500 - ~ 2100 - H F]@X~~~~ l/ ZrO2 + liquid u) / c 1900 / ~' / / QX I~/ / 1700 / <. / l1500 Fe2SiO4 + liquid 1300 1182 + 4~ 1100 Fe2Si04 + ZrO2 FeSiO 10 20 I 40 5 60 70 80 I9 ZrI FepSiO4 10 20 30 40 50 60 70 80 90 ZrO2 WEIGHT PERCENT ZrO2 Figurel6. The System Fayalite - Zirconia (x's indicate compositions studied).

A. Mix 31, 11750C, 15 min, water quenched. B. Mix 31, 12500C, 3 min, water quenched. I C. Mix 31, 1250~C. 5 min, then cooled to 11500C, 15 min, water quench. Figure 17. Phot FayC D. Mix 30, 12500C, 5 min, then cooled to 1150~C, 15 min, water quench. tomicrographs of Compositions in alite-Baddeleyite System. 500X

eutectic structure and a glassy phase were observed0 It was, therefore, concluded that the eutectic is between 2.5 percent and 5 percent of ZrO 2 Ternary Invariant Points In the compatibility triangle FeO-ZrO-Fe2SiO4, the ternary eutectic lies at approximately: 71% FeO, 7% ZrO2, and 22% SiO2o The eutectic temperature is 1157 +4~Co In the other compatibility triangle, ZrO2-SiO2-Fe2SiO4 the ternary eutectic lies at approximately: 58% FeO, 4% ZrO2, and 38% SiO2, and at a temperature of 1166 + 4~Co The latter triangle also contains a ternary peritectic as the result of the dissociation of zircon, That is, zircon has a field of primary crystallization which does not include the zircon composition, The peritectic temperature is reported by Curtis and Sowman to be about 1530~C. The present work has shown nothing that would be contrary to thiso Figure 18 shows the invariant points in the system. The method for determining the ternary eutectic points was essentially the same as that described for the binary system. In the ternary system the primary field of crystallization may be established. The approximate distance from the boundary line where the second phase begins the crystallization may be determined in the microsection by observing the distribution and amounts of the phase present. For example in establishing the position Of the eutectic in the -cp0.oeition -triglte -eO-ZrG2-Fe2SiO04 microsections of the mixes 32, 33, and 49 were examined after quenching

from 1225~Co In mix 32 the primary phase was observed to be zirconia with wustite as the second phase to precipitate. All of the fayalite is in the eutectic precipitate. The situation in mix 33 is somewhat different for here wustite is the primary phase with fayalite as the second phase and baddeleyite only in the'utectic, Mix 49 again shows a different sequence, with fayalite as the pr.-lary phase, wustite as the second phase, and baddeleyite only in the eutectic precipitate. By constructing lines through the mix compositions along the paths of crystallization, as indicated by the microstructure, the eutectic can be located to within a small area. The same procedure was followed in the system ZrO2-Fe2SiO4-SiO2 with mixes 35, 535, nd 58, with the results as reported in the preceding paragraph. Primary Crystallization Fields In the quasi-ternary FeO-ZrO2-SiO2 system there are six fields of primary crystalli-zation, namely: wustite, fayalite, tridymite, cristobalite, zircon, tetragonal zirconia, The field of metallic iron also extends into this sytem. The phase having by far the largest region of primary crystallization is zirconia, which covers well over threequarters of the area, This is, of course, the result of having all the eutectics in the system at the extreme opposite side from ZrO2o The liquidus temperatures in this area are all well above the termperature range in which monoclinic zirconia is stable, and therefore, the zirconia which crystallizes out of the melt is tetragonal, This accounts for the

difference in appearance of the precipitated zirconia from that zirconia which has formed or grown in. the solid state. Because of the defect nature of ferrous oxide, some ferric iron is present. However, it is not present as a separate phase but is present within the wustite structure, and maintains electrical balance offsetting the missing ferrous iron positions, The relative concentration of ferric iron in different compositions according to atmosphere and quenching medium is shown in Table IJ and Figure 18o Because of the possibility of oxygen pickup during quenching in all samples except those in the sealed iron crucibles, it is probable that the latter more nearly represents the true equilibrium relationship between free iron and ferric oxide in the various compositions at 1200~Co Iron also occurs as a primary phase over an appreciable region around the FeO apexo The FeO-ZrO2-Si0 2 system can be considered as ternary only if this iron phase is neglected0 The field of primary crystallization of wustite is relatively small, being bounded by lines joining the points at FeO, A, J, and B in Figure 8. The method for determining the eutectic points and the boundary lines was discussed in the previous section, The exact shape of the wustite-zirconia and the wustite-fayalite boundary lines is not known, but they must be close to those shown in order to explain the phase relationships found in the microsectionso The fayalite field of primary crystallization, which also contains some free iron, is even smaller than the wustite field0 It is depressed somewhat toward the fayalite composition at the fayalitezirconia eutectic point0 In the binary system ZrO2-SiO2 the phase zircon

Si02 GG - Gas* atmosphere, gas quench GW - Gas* atmosphere, water quench 1W - Iron crucible, water quench co/CO2 ZrO2 FeO 13.64, GG 18.57, GW Figure 18. Iron Content of Samples Fired at 1200~C.

-36 on heating dissociates in the solid state below the eutectic and therefore does not crystallize out of a melto However, in this ternary system a field of primary crystallization of zircon does exist (Figure 19-A). This field appears to be associated with the 1500~C isotherm, It is the field which was least well defined in the present study. Figure 19-B illustrates the necessity of sealing the crucible in an inert atmosphere. In 19-A the iron plug was inserted in a purified argon atmosphere then the plug was welded in. In 19-B the plug was inserted in air and was not welded. It may be observed that the system has moved down into the ternary region containing baddeleyite as the result of the formation of ferrous oxide from the small amount of entrapped air. Effect of HfO2 and other Impurities on ZrO2 Transformation A number of authors have pointed out that the monoclinictetragonal transformation on heating ZrO2 takes place over a range of temperatures rather than at a single temperature, It has also been noted that the reverse transformation on cooling takes place at lower temperatures than that on heating0 This information as well, as that of the present study is shown in Table V, Since all of the ZrO2 available until recently has contained from 1.5 to 4 percent by weight, or more, of hafnium, and since hafnia has been found to transform at temperatures about 500~C higher than zirconia, it seemed reasonable to suppose that

A. 625X B. 500X Figure 19. Mix 17, Effects of Loading Technique. A, Sealed in purified Argon, top welded, heated at 1500~C for 30 minutes, air quenched, shows primary Zircon crystals in matrix of Zircon-Fayalite eutectic and glass; B, sealed in air, top not welded, heated at 1200~C for 30 hours, quenched in water, shows Baddeleyite, Wustite, Fayalite, and Glass.

538 the presence of hafnium may have some effect upon the transformation of zirconiao Curtis, Donney and Johnson(22) have investigated the effects of hafnia upon the lattice parameter of zirconia by means of X-rays. The same authors checked on heating the monoclinic-tetragonal transformation of zirconia containing only 80 ppm of hafnia by means of DTAo Mumpton(36) also checked the transformation by DTA on heating and on cooling, The zirconia used by Mumpton was reported to contain less than 300 ppm of hafnia but other impurities such as TiO2 were not listed. Curtis, et al, found an endothermic peak starting at about 1100~C reaching a maximum at 1190~C and ending at 1250~Co They did not report the results on cooling. Mumpton found an endothermic peak beginning at about 1170~C, reaching a peak. at 11.90~C and ending at 12300C on heating. On cooling the exothermic peak began at 1040~C, reached a maximum at 1015~C and ended at 980cCo In the present investigation, with zirconia containing less than 500 ppm of hafnia and total impurities less than 025 percent, on, heating an endothermic peak was found to begin at 1170~C, reach a maximum at 1199~C and reach completion at 12300Co These temperatures are averages of four runs~ On. cooling the peak was found to begin at 9080~C reach a maximum at 880~C and end at 850~Co This is a difference of about 135~C lower than Mumptonis data, However, it is interesting to note that when 10 percent of HfO2 is added to the "pure" zirconia, fired for 4 hours at 1650~C and a DTA is run, the peak on cooling is very near that

-39reported by Mumptono The same is true when 5 percent of silica is added to the pure zirconia, fired for 4 hours at 1500~C and a DTA determination is made, Figure 20 shows schematically these DTA results as compared with that reported by Mumptono Qualitatively the above results may be explained either from the standpoint of free energy relationships or that of solid state reaction kineticso If we assume first that Mumpton's material contained sufficient impurities other than hafnia to have an effect upon the transformation temperature the data begin to fall into a certain patterno It is, of course, very probable that the "true" transformation temperature lies someplace between. the transformation which takes place on heating and that which takes place on cooling. The difference in free energy between the "true" transformation temperature and the actual temperature would correspond to the driving force necessary for the nucleation and growth of the new phaseo -In the "pure" material the temperature difference would be about 1i0~C which would correspond an activation energy of Xo In the impure materials the "true" transformation temperature would probably be from 50 to 75 degrees C higher than that of the "pure" material on the basis of the DTA data, The free energy difference representing the driving force would in turn be considerably less, This would correspond. to a different activation energy Y perhaps one-half that of X. This could be interpreted On the basis that the impurities act as nucleation agents thereby lowering the

Mumpton's "Pure" () -c WC I F0 X Z LLJ LL Commercial (Reagent Grade) T.A.M. "Pure" + 10% HfO2 T.A.M. "Pure" + 5% SiO2 T.A.M. "Pure" + 5% FeO T.A.M. "Pure" \.. -- 200~/hr. --- 500 0/hr........ 900 /hr. 0 o I I 700 8oo 900 1000 TEMPERATURE - DEGREES CENTIGRADE 1100 1200 Figure 20. Effects of Impurities on Zirconia Phase Transformations, DTA Curves, Endothermic on Heating, Exothermic on Cooling.

-41, energy required for the formation of the new phase (Figure 21)o This, of course, does not rule out the possibility of kinetic effects, However, two other bits of information make the kinetic concept less attractiveo First, Table VI shows essentially no effect of heating rate upon the transformation of monoclinic to tetragonal zirconia as determined by differential thermal analysis. Second, Mumpton found that upon heating ZrO2 to above the transformation temperature then cooling to below that temperature and just above the lower "cooling" transformation temperature baddeleyite lines were not found even after 12 hours holding time, only the tetragonal lines were found. The differential thermal analysis work of the present investigation was supplemented by high-temperature X-ray diffraction and excellent correlation was obtained on the phase transformation temperatures by the two methodso In the high-temperature X-ray work the (111) peak of the tetragonal phase and the (111) and (11i) peaks of the monoclinic (baddeleyite) phase were used for identification. The high-temperature X-ray camera used (36) was similar to that described by Van Valkenburg and H. F. McMurdie ) Dissociation of Zircon It has been realized for some time that zircon dissociates into zirconia and silica at elevated-temperatures. Matignon(38) observed that zircon dissociates into zirconia and silica at 1800~C and above. Barlett(35) found that a quenched melt.of zircon had decomposed into ZrO2 and a high

i I95e, c ZrO2 + o10% HfO2 -m-' ZrO2 + 5% SiO2 I I I I I I I I I I I 600 700 800 900 1000 1100 1200 1300 TEMPERATURE DEGREES CENTIGRADE Figure 21. Schematic Free Energy versus Temperature of Beginning of Transformation for ZrO2.

-43 Table I Semi-Quantitative Spectrographic Analyses of Raw Materials Used (They are reported as oxides of the elements indicated.) TAM FISHERi ZIRCON AUSTRALIAN SILICA ZrO2 Zr2 FLOUR ZIRCON FLOUR -._ - - -. --— __ __ __ __ Mg.008%.04%.003%.001%.01% Al.001.02.2.003.01 Si.02.5 high high high Ca.003.03.005.003.003 Ti.015.2.4.3.005 Cr.008.003.003.003.003 Fe.05.02.04.01.01 Ni.003.01.006.005.007 Cu.001.001.001.001.003 Sr.001.008.001.001.004 Zr high high high high.005 Ba.004.05.003.003.0005 Pb.008.01.008.008.008 Sc.002.002.005.003 Y.005.005.3.15 Yb.002.002.05.02 Hf. 05 4. 2. 1.5

-44 rI I, Table II Analysis of Ferrovac "E" 3/8-in. Rod used for Crucibles and Lids Element Wt. Percent C 0.025 Mn 0.005 or less Si 0.007 Ni 0.005 0 0.0023 N 0.0004 P <0.01 S <0.01 Mo (0.01 Cr <0.01 V <0.01 Sn (0.01 Al (0.01 Co (0.01 Cu (0. 01 Pb (0.01 I I i

Table III Results of Quenching Experiments, 1020 Steel Specimen 3/8" Square by 3/4" Long, with Thermocouple Sealed in Center: Temperatures in Degrees Centigrade Quenching Medium 0 sec. l sec. 2 sec. 3 secs 4 sec.J 5 sec. 10 sec.l 15 sec. 20 sec. __W,,,, L,":' 7',' f "J r.... - -,,~,,..._____.. L_; 7s:t I I I Water 1065 11000 760 38 - - - - - Water 1095 1060 815 590 65 ~ - Water 1110 1040 870 540 260 65- - - Dibutyl 1095 1040 970 855 730 605 290 205 160 Phthalate Dibutyl PDbutyl 1095 1010 900 815 705 605 290 205 155 Phthalate Dibutyl 1065 980 900 815 705 620 290 205 120 Phthalate Mercury 1095 1010 905 800 755 680 480 280 205 Spec. Floating Mercury 1040 980 900 820 775 730 565 410 315 Spec. Floating Mercury 1120 1065 955 875 815 745 540 300 175 Spec. Submerged Mercury 1095 1045 1915 800 720 650 455 250 150 Spec. Submerged -4, -t 4-.0

-46 Table IV Comparison of Ferrous and Ferric Oxide Content of some Specimens Fired in an Atmosphere of 50/50 CO/CO2 with those Fired in Iron Crucibles Quenching~ ie++ Fe+++ Specimen No. i Environment QMedium Total Fe ~ F~~~~~ 1 1 2 3 4 4 6 7 10 11 11 12 2 11 21 32 i. i i.tI i Gas Gas Gas Gas Gas Gas Gas Gas Gas Gas Gas Iron Iron Iron Iron Iron Iron i I I j i I I i Gas Water Gas Water Gas Water Gas Gas Gas Gas Gas Water Water Water Water Water Water Wfater I 62.09 56.10 48.66 19.37 8.20 9.28 50.84 9.48 6.52 28.51 45.38 58.21 19.26 54.09 36.67 31.65 54.64'i i i I 13.64 18.57 9.14 2.86 i 2.46 1.84 1.46 1.97 1.37 9.58 14.59 3.00 0.29 0.42 0.75 0.07 0.09 1 I 7 I ~, t ii i I I 75.73 74.67 57.80 22.23 10.66 11.12 52.30 11.45 7.89 38.09 59.97 61.21 19.55 54.51 37.42 31.72 54.73 i I J -4

Table V Monoclinic-Tetragonal Transformations DTA Data on Zirconia Heating Cooling Source Begin Peak End Begin Peak End Curtis et al. 1100 | 1190 1250 - - - Mumpton 1170 3 1190 1230 1040 1015 980 Allison & Taylor 987 Bunting 1170 1034 Present "Pure" 1162 1187 1213 908 880 850 Present (10% HfO2) 1136 1185 1207 1046 1002 967 Present (,5% SiO2) 1176 1215 1244 1012 992 973 Present ( 5% FeO) 1125 1187 - 959 941 915 Present ( 4% HfO2 1141 1177 1215 1014 990 955 Commercial) __________________________________________ S ________~~~! Table VI Rate upon ZrO2 Transformation Effect of Heating Heating Rate Peak Begin End 2000/hour 1192 1165 1217 500~/hour 1185 1170 1212 9000/hour 1184 1150 1209.........~~~~~~~.

-48 silica glass, Geller and Lang(3) in a phase diagram of the system Si.02-ZrO2 show a dissociation of zircon into zirconia and a liquid at 17750~C Curtis and Sowman3) found zircon to dissociate in the solid state into zirconia and silica at about 1540~Co Mumpton(36) also found the dissociation to take place in the solid state someplace between. 1500 C and 1600~Co The present investigation has shown that zirconia and silica intimately mixed will combine to form zircon at temperatures up to 1500~C while zircon fired for 4 hours at 1600~C shows zirconia lines when X-rayed, indicating that dissociation has occurred. No evidence of a melt could be found microscopically in the zircon fired at 16000~C Immiscible Liquid Regions In the system FeO-Si02 a region. of two immiscible liquids extending from 96,9 percent Si02 to 55o3 percent SiO2 at a temperature of 16900C was (539,40) reported by Greig(39 Further work in ferrous oxide-silica systems has been in general agreement with Greig's work. In. the system ZrO2-SiO2 Barlett35) reported that when melted, a mix of 50 percent ZrO2 and 50 percent Si02 forms two immiscible liquids, one of which devitrifies upon cooling. Barlett did not attempt to define the composition nor temperature (32) of the ipnmiscible liquids region, Toropov and Galakov ) investigated the system Zr02-SiO2 at temperatures above 1800~C supplementing the work of Curtis and Sowman(31) who investigated the system below 1800~Co The Russian authors report a rqgion of liquid immiscibility in the system from

49 41 to 62 weight percent of silica at 2250~C, extending to a maximum at 53 percent silica and 2430~Co In melts in the two liquid region they find a clear highly siliceous liquid containing spheres of a second liquid which upon cooling becomes turbid due to the very fine precipitate of zirconiao Barlett35 melted the mixes in a carbon arc and allowed the liquid (32) droplets to fall into a trough of running water. Toropov and Galakov 3 melted small samples of the mixes in a 5 mm spiral of tungsten wire in an oxygen-free argon atmosphere. The present work checking compositions in the binary regions (FeO-SiO2 and ZrO2-SiO2) and in the ternary region ZrO2-FeO-SiO2 was accomplished by melting small. plugs 1/8" in diameter by 1./8" high in an oxy-acetylene flame and quenching the melt with a water spray. Some compositions were also mixed with a small amount of Duco cement and formed into fibers 1 mm or less in diameter and two or three centimeters long. These fibers were melted in an oxy-acetylene flame and the molten ends allowed to drop into a beaker of water. Melts of FeO and SiO2 and of MgO and Si02 in. the proper proportions made by the former process showed clearly the two liquid characters. However, because of the rapid devitrification of one liquid in the ZrO-SiO2 system, the results were not quite so clear cuto Effects of Ferrous Oxide on Zircon and Zirconia Refractories Probably the most important result of this study, as far as

-50 refractories is concerned, is that wustite and zircon are not compatible. As a result of this incompatibility zircon will dissociate to form an iron-rich silicate liquid or fayalite and baddeleyite when in contact with a slag containing free ferrous oxideO This may occur even at temperatures below 1200~Co The amount of liquid formed at any particular ferrous oxide level will not be greatly different from that which forms when ferrous oxide comes in contact with a silica brick. However, the effect of the liquid will not necessarily be the same for the two different refractorieso The surface tension of the liquid in contact with the refractory grains and the degree of solid-to-solid contact between the grains play an. (4o 4l) important role in the behavior of the refractory in service( 0 4 In the references cited, the authors compare the role of the liquid phase in periclase refractories as compared with silica refractorieso It is shown that with the large degree of solid-to-solid contact in the silica refractory that there is considerable strength under load even with 25 percent of liquid present. With less than 5 percent of liquid in periclase refractories failure can occur through spalling and peeling. This results chiefly from the small degree of surface contact between the nearly spherical periclase grains. The lower viscosity of the magnesiairon silicate liquid and a probably lower surface tension of liquid in contact with the periclase grains no doubt have an additional important effect on this high temperature weaknesso The present investigation would indicate that zircon too would be expected to be relatively weak at

51-, high temperatures in the presence of ferrous oxide because here, too, a fluid slag is formed which penetrates easily between the grains causing partial or total dissociation of the zircono The zirconia resulting from this reaction is in the form of rounded "rice grain" particles which have little solid to solid contact. This may be seen in Figure.19-Bo In zirconia refractories the effect of ferrous oxide would not be expected to be so severe as with zircon. Ferrous oxide and zirconia are compatible, and the temperature at which the first liquid is formed is over 150 degrees centigrade above that at which it is formed in the zircon refractory. The microstructure of zirconia would indicate that it should be a more refractory material at high temperatures than at intermediate temperatures. The reason for this is that at intermediate temperatures zirconia tends to be rounded, similar to periclase) and the ferrous oxide liquid penetrates easily between the rounded grains (Figure 12)o At higher temperatures this liquid could aid in the sintering of these grains to form a network somewhat similar to that in a silica bricko Of course, zirconia has one drawback which overcomes all of its other desirable characterizationso The transformation from monoclinic to tetragonal phase is accompanied by a change of volume of several percent0 In practice this is overcome by adding calcium oxide to the zirconia to form a cubic phase which does not transform. and which has more desirable thermal expansion characteristics0 The effect of ferrous oxide upon this cubic phase is outside the scope of the present investigation and was therefore not studiedo

SUMMARY It has been shown that in the quasi.-ternary system. FeO-ZrO2-SiO2 there are no ternary compoundso Above about 1550~C there are two compatibility regions FeO-Fe2SiO4-ZrO2 and FeO-Fe2Si04-Si02. Below 1550~C there are three compatibility regions FeO-Fe2SiO4-ZrO2, Fe2SiO4-ZrSiO4-ZrO2, and Fe2SiO4-ZrSi04-Si02o Fayalite and zirconia are compatible and form a binary systemo Zircon dissaciates at about 1550~C to ZrO2 and SiO2, and therefore the zircon-fayalite system is not a binary system at all. temperatures. The binary system FeO-ZrO2 has been re-examined and the eutectic temperature located at 1.323~ 3C, The reported solubility of four percent of FeO in ZrO2 at1450~C was disproved. A reexamination of the SiO2-ZrO2 system uphel.d the work of Curtis and Sowman. In the system fayalite-zirconia a eutectic was found at four percent ZrO2 and 1182 + 4C, Two ternary invariant points were locatedo In. the FeO-ZrO2-Fe2SiO4 region a ternary eutectic lies at approximately 71 per cent FeO, seven per cent ZrO2, and 27 percent SiO2 and at 11.57 ~ 4Co In the ZrO2-Si02-Fe2SiO4 region a ternary eutectic lies at approximately 58 percent FeO, four percent ZrO2, and 38 percent Si02 at a temperature of 1166 + 4Co There is a small field of primary crystallization of zircon and a ternary peritectic in the system also, Six fields of primary crystallization were observed: wustite, fayalite, tridymite, cristobalite, zircon. and tetragonal zirconiao -52

-53 The field of metallic iron also extends into this system making it a quasi-ternary system, The presence of Hf02 and other impurities were shown to have a decided effect upon the monoclinic-tetragonal transformation of zirconiao The chief effect was that of closing the hysteresis between the transformation temperature on heating and that on coolingo The region of liquid immiscibility reported by Toropov and Galakov in the system SiO2-ZrO2 could not be reproduced, but was not definitely disproved by this work. The most important result of this study to a refractories application is that wustite and zircon are not compatibleo As a result of this incompatibility relatively small amounts of wustite will cause a dissociation of zircon to form an iron rich silicate liquid or fayalite and baddeleyite at temperatures as low as 1200C.

REFERENCES.o Bowen, NoLo and Schairer, JoFo "The System FeO-SiO2, " Am. Jlouro Sci,9 5th Series, 24, (1932), 1.41 1.77-21.o 20 Mathewson, C oHo, Spire, Eo and Milligan, W, "Evolution of the IronOxygen Constitutional Diagram," Trans. Am.. sco Steel Treati.ng, _, (1931), 66-88~ 3o Darken, LoSo and Gurry, RoWo "The System Iron-Oxygen-ZI, The Wustite Field and Related Equilibrium, " Jour. Am, Chem. Soc., 6 (1.945)9 1598-14122. 4, Darken9 L,So and Gurry, RoWo "The System Iron-Oxygen-II, Equilibrium and Thermodynamics of Liquid Oxide and Other Phases " Jour, Am.r Chemo Soco, 68, (1946), 798-816. 5. Wells, RoG. "Microscopic Identification of Wu.stite in Presence of Other Oxides of Iron." Analyto Chem., 26, (.1954). 71l5-717o 60 Yardley9 Ko "The Structure of Baddeleyite and of Prepared Zirconia," Mineral. Mag., 219 (1926) 1.69-175o 7. Ruff, 0. and Ebert, F, "Die Formen des Zirkondioxyds," Zeito anorg. und allgem. Chem., 180, (1929), 19-41. 80 v, Naray-Szaboc Lto "Zur Struktur des Baddeleyits, Zr02.^ Zeito fur Krysto., 4 (.19536) 414-416. 9, Davey, WoPo "Crystal Structure of Zirconium Dioxide," Phys. Revo,:7 (1926), 798-801, 10o Cohn, WoMo and Tolksdcrf, So "Die Formen des Zirkondioxd.s in Abhangigkeit von der Vorbehandlung, " Zeit. Physik. Chem., 18, I19.30) 331-3560 1, Blak e, GoSo and Smith, GoF.Ho "Baddeleyite from Ceylon," Mineral Ma, 14, (1907) 5378-3584o 12o Fletcher, L. "On Baddeleyite a New Mineral from. Rakwa.na, Ceylon," Mineral, Mag.,. 0, (1893), l48-160o 13 Hussak,, E Ueber Brazi.lit, ein neues Tantal- (Niob.-) Mi.neral. von der Eisenmine Jacu.piranga, S'ud- Sao Paulo o Ne-es Jahrb.uh fur Mineralogie; Briefliche Mittheilungen, Part II, (1892), l4l 1-6o 14o Hussak, E Ueber den Baddelyit, (syn, Brazilit) von der Eisenm.ine Jacupiranga in. Sao Pauloo Tschermak s Mineralog.sche und petrographische Mitteilungen; 1,49 (31895) 4o6

-55-.15 Hussak, E, and Reitinger, Jo "Ueber Monazit, Xenotim, Senait und naturaliches Zirkonoxyd aus Brasilien," Zeito fur Kryst,, 37T, (1.903)) 550-579~ 16i Hiemstra, So Ao "Baddeleyite from Phalabouva, Eastern Transvaal," Am, Mineral,. 40, (1955), 275-282o 170 von Hevesy, Go and Jantzen, Vo To "The Hafn.ilm Content of Zirconium Ores," Zeit, anorg. und allgem. Chemo, 123 (1924), 113-118, and Journ. Chem. Soco, London9 (19253) 3218-32235 18o Frondel. Co Zirconiurm. Mineralogy and Geochemistry, Paper No. 32, 19o Franco, Ro Ro and Loewenstein, "Zirconium. from the Region of Pocos de Caldas " Am, Mineral., 33, (1948) 147-151o 20, Allison, Eo Bo and Taylor, J, "The Gamma to Alpha Transition in Zirconia Solid Solutions with Lime and Magnesia, " Transo Brit Ceramic Soc,, 549 (1955), 11, 677-692o 21. Blumenthal, W, B o The Chemical, Behavior of Zirconiumo Princeton, New Jersey: Do Varn Nostrand COo, Inc,, 1958, Chapter 4, 151-200, 22o Curtis, Co Eo, Doney, Lo M, and Johnson, Jo Ro "Some Properties of Hafnium Oxide, Hafnium Silicate, Calcium Hafnate, and Hafnium Carbide, Journo Am, Ceramic Soco, 35, (1954), 10, 458-465, 235 Schuhmann, Ro and Ensi.o P, J, "Thermodynamics of Iron Silicate Slags, Slags Saturated with Gamma Iron," Trans, AIME, 191, (1951), 401-4llo 24. Allen, Wo Co and Snow9 Ro Bo "The Orthosilicate-Iron Oxide Portion of the System CaO:-"FeOs-..SiO,2 2 Journ, Am. Ceramic Soc o, 38 (.1955), 264380, 25a Bowen9 No Lo "Ferrosilite, as a Natural. Mineral," Am, JournO Sci,, 5th Series, 30, (1935), 482-4940 26. Schairer, J. Fo and Yagi, Ko "The System FeO-*A1203-Si0 2t" Am, Journ, Sci, Bowen Vol.,, (1952). 471-512, 27, Roedder, Eo "A Reconnaissance of Liquidus Relations in the System K20O2SiO2-FeO-SiO2,' Am, Journ. Sci,, Bowen Volo (1952), 435-456, 28o Rigby, Go Ro Mineralogy of Ceramic Materials, Stoke-on Trent, England: Brit, Ceramic Res, Assno, 1.948, 1.21 and 1,335

29~ Fischer, Wo A, and Hoffman, A, "'Gleichgewichtsuntersuchungen im System Eisen (II)-oxyd-Zirkonoxyd, " Arch. fur das Eisenhuttenw,, 28, (1957), 739-7430 30. Geller, So and Lang, S. Mo "System Si02-Zr02," Journ. Am Ceramic Soc, 32, (1949), 11, Part II, 157o 31o Curtis, C, Eo and Sowman, Ho Go "Investigation of the Thermal Dissociation, Reassociation, and Synthesis of Zircon, " Journo Am, Ceramic Soc,, 36,, (1953), 6, 1904-198, 32, Toropov, No Ao and Galakov, Fo J "Liquid Formation in the System ZrO-SiO2," IzvestO Akad, Nauk SSSR, Otdel. Khimo Nauk, No, 2, (1956) 158-1.61 33 Washburn, Eo Wo and Libman, E, Eo "Approximate Determination of Melting-Point Diagram of System Zirconia-Silica,," Journo Am. Ceramic SOC, 2., (1920), 8, 634-640o 34, Zirnowa, No "Schmelzdiagram des Systems ZrO2-Si02," Zeit, anog und allgemo Chem,, 218, (1934), 2,.193-"200o 55o Barlett, Ho Bo "X-Ray and Microscopic Studies of Silicate Melts Containing ZrO2," Journ Amo Ceramic SCo,, 14, (1931)5, 837-8430 36, Mumpton, F, Ao Stability Studies of the Zircon-Thorite Group and the Effect of Related Oxides, PhoDo Thesis, Penno State Univ,, College Park, Pao, 1958, 37. Van Valkenburg, Ao and McMurdie, HH. Fo High Temperature X-Ray Diffraction Apparatus, Amn. Ceramic Soco Bulletin,9 26, (1,947), 5, 1.39-142 380 Matignon, Mo Co Actionr des Temperatures Elevees sur Queue Substances Refractaires Paris~ Compt, Rendus, Acad, Sci., (192\.3 1,290-1.293, 39o Greig, Jo W, "Immiscibility in Silicate Melts," Am. Journ, Sci,, 1, (1927), 73, 1-44, and 1335-154, 40, Greig, J, Wo "On. Liquid Immiscibility in the System, FeO-Fe203-A1203 Si02a " Am. Journo Sci_, 14, (1.927), 473-484o 41, Wells, Ro Go and Van Vlack, Lo Ho "Mineral and Chemical Changes in Periclase Brick under Conditions of Steel. Plant Operation," Journo Amo Ceramic SocO,.4, (1951), 2, 64-70o 42, Van Vlack, Lo Ho "Microstructure of Silica in the Presence of Iron Oxide," Journ, Am, Ceramic Soco, 43, (1960), 3, 140-145, o __