Effect of yttria on interfacial reactions between titanium melt and hot-pressed yttria/zirconia composites at 1700 degrees C

Effect of Yttria on Interfacial Reactions Between Titanium Melt and

Hot-Pressed Yttria/Zirconia Composites at 17001C

Chien-Cheng Lin,*

,w

Yao-Wen Chang, and Kun-Lin Lin

Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu 30050, Taiwan

Kun-Fung Lin

Materials and Electro-Optics Research Division, Chung-Shan Institute of Science Technology, Taoyuan 32546, Taiwan Various Y2O3/ZrO2 samples were fabricated by hot pressing,

whereby Y2O3was mutually dissolved or reacted with ZrO2as a solid solution or Zr3Y4O12. Hot-pressed samples were allowed to react with Ti melt at 17001C for 10 min in argon. Micro-structural characterization was conducted using X-ray diffrac-tion and analytical electron microscopy. The Y2O3/ZrO2 samples became more stable with increasing Y2O3 because Y2O3 was hardly reacted and dissolved with Ti melt. The in-corporation of more than 30 vol% Y2O3could effectively sup-press the reactions in the Ti side, where only a very small amount ofa-Ti and b0_{-Ti was found. When ZrO}

2was dissolved into Ti on the zirconia side near the original interfaces, Y2O3 reprecipitated in the samples containing 30%–70 vol% Y2O3, because the solubility of Y2O3in Ti was very low. In the region far from the original interface, a-Zr, Y2O3, and/or residual Zr3Y4O12were found in the samples containing more than 50 vol% Y2O3and the amount ofa-Zr decreased with increasing Y2O3.

I. Introduction

T

HEmajority of ceramic materials seriously react with titani-um and titanititani-um alloys during casting, resulting in an a-casing and the deterioration of mechanical properties. Because no suitable ceramic crucible is available, titanium and titanium alloys are usually melted in a water-cooled copper crucible by consumable electrode vacuum arc melting (VAR) instead of vacuum induction melting (VIM). However, there are some drawbacks to the VAR process, including the high cost of the equipment, scrape recycle, and long cycle time. If a suitable cru-cible material were available, the VIM process would be feasible in industry. Furthermore, titanium castings need to be chemi-cally milled in order to remove the reaction products on their surface. If the interfacial reactions between titanium and ceram-ic mold were well controlled, the chemceram-ical milling of the so-called a-casing would not be required. Therefore, determin-ing how to control the interfacial reactions between titanium melt and some ceramic materials is of great interest.

Extensive studies have been carried out on the interfacial reactions between titanium melt and zirconia molds and/or cru-cibles in the last few decades.1–8Saha and Jacob4indicated that a brittle a-case was formed at the surface of titanium parts and thus adversely affected their mechanical properties. Weber and his coworkers5,7found a feather-like eutectic phase and

black-ened zirconia after titanium alloys reacted with a zirconia cru-cible. Zhu et al.6claimed that there were two distinct chemical reaction layers at the interface between titanium melt and zir-conia after reaction at 17001C. Holcombe and Serandos9stated that the composite crucible of W and Y2O3revealed little

con-tamination after reaction with titanium melt.

Recently, Lin and colleagues10–16_{have thoroughly}

investigat-ed the interfacial reaction mechanisms between titanium (or titanium alloys) and 3Y–ZrO2 using analytical transmission

electron microscopy. The lamellar orthorhombic Ti2ZrO and

the ordered titanium suboxide (Ti3O) were formed in a-Ti(Zr,

O) at the interface between Ti melt and 3Y–ZrO2after reaction

at 17501C.10For the Ti/3Y–ZrO2diffusion couple, both lamellar

orthorhombic Ti2ZrO and spherical hexagonal Ti2ZrO were

found in a-Ti(Zr, O) after annealing at 15501C.11,13,14In addi-tion, the orientation relations of acicular a-Ti and b0_{-Ti were} determined to be [2110]a
Ti//[001]b0
_Ti and (0001)_a
Ti//

(100)b0
_Ti in combination of [2110]_a
Ti//[021]_b0
_Ti and

(0001)a
Ti//(112)b0
_Ti, respectively. Lin and Lin12also found

in-tergranular a-Zr, twinned t0-ZrO2
x, lenticular t-ZrO2
x, and

ordered c-ZrO2
xon the zirconia side far from the interface

be-tween Ti and 3Y–ZrO2after annealing at 15501C.

Various ratios of Y2O3/ZrO2samples were attempted in this

study to achieve better control over the reactions on the titanium side as well as the ceramic side. The powder mixtures of Y2O3/

ZrO2were hot pressed and then allowed to react with titanium

melt at 17001C for 10 min in argon. Various reaction layers at the interface between titanium melt and Y2O3/ZrO2 samples

were characterized using analytical scanning electron microsco-py and analytical transmission electron microscomicrosco-py. The effect of Y2O3on the interfacial reactions between Ti melt and Y2O3/

ZrO2samples will be elucidated in this study.

II. Experimental Procedure

The starting powders used were zirconia (499.95 wt% ZrO21HfO2with HfO2accounting for approximately 2%–3%

of this total,o0.002 wt% SiO2,o0.002 wt% Al2O3,o0.005

wt% Fe2O3,o0.005 wt% TiO2,o0.002 wt% CaO; 0.5 mm in

average; Toyo Soda Mfg. Co., Tokyo, Japan), yttria (499.9 wt% Y2O3 with trace rare earth oxides o0.001 wt% CeO2,

o0.001 wt% Pr6O11,o0.01 wt% Nd2O3,o0.003 wt% Sm2O3,

o0.005 wt% Tb4O7,o0.005 wt% Dy2O3,o0.001 wt% CaO,

o0.001 wt% Fe2O3 and other traces in balance; 0.5 mm in

average; NYC Ltd., Fukuoka, Japan), and commercially pure titanium (with a nominal composition of 99.31 wt% Ti, 0.3 wt% Fe, 0.25 wt% O, 0.1 wt% C, 0.03 wt% N, 0.01 wt% H, 60–70 mm average in diameter, Alfa Aesar, Ward Hill, MA).

Powder mixtures with various Y2O3/ZrO2 ratios were

dis-persed in the ethanol solvent. The pH of the suspension was adjusted to 11 by adding NH4OH, and the suspension was

ultrasonically vibrated for about 10 min (Model XL-2020, Sonic-ator, Heat Systems Inc., Farmingdale, NY). The suspension was M. Rigaud—contributing editor

*Member, The American Ceramic Society.

Research supported by Chung-Shan Institute of Science Technology under Contract No. BV96D08P.

w

Author to whom correspondence should be addressed. e-mail: [email protected]. edu.tw

Manuscript No. 24082. Received December 10, 2007; approved February 17, 2008.

Journal

DOI: 10.1111/j.1551-2916.2008.02428.x r2008 The American Ceramic Society

dried in an oven at 1501C, ground with an agate mortar and pestle, and then the mixtures were passed through 80 mesh. Samples with various Y2O3/ZrO2ratios were prepared by hot

pressing in a graphite furnace at 15501–16001C for 30 min in an Ar atmosphere (Model HP50-MTG-7010, Thermal Technology Inc., Santa Rosa, CA). Oxygen-deficient zirconia was formed in the as hot-pressed samples. To avoid the inaccuracy, samples were annealed in air at 13001C for 1 h so that stoichiometric zirconia was obtained.

The hot press conditions, compositions, relative densities, and designations of Y2O3/ZrO2samples are listed in Table I. The

Y2O3/ZrO2samples contained 10, 30, 50, 70, 90, and 100 vol%

Y2O3, respectively, and were balanced with ZrO2. The sample

consisting of 10 vol% Y2O3and 90 vol% ZrO2was designated

as 10Y90Z, and so on. The apparent densities of all Y2O3/ZrO2

powder mixtures were measured using gas pycnometry with 99.99% pure helium (Model MultiVolume Pycnometer 1305, Micromeritics, Norcross, GA). The bulk densities of the hot-pressed Y2O3/ZrO2samples were determined by the

Archi-medes method using water as an immersion medium. For a nonporous powder, the apparent density approximates the true density and can be used as the reference point in calculating the percentage of the theoretic density of a hot-pressed body. The relative densities of the hot-pressed samples were calculated as follows: relative density 5 (bulk density/apparent density)  100%.

Hot-pressed samples were cut and machined to dimensions of about 10 mm 10 mm  5 mm. A bulk specimen was vertically placed into the zirconia crucible, and then tightly packed with commercially pure titanium powder. The crucible was put in an electric resistance furnace (Model No. 4156, Centorr Inc., Nashua, New Hampshire, UK) with tungsten mesh heating elements. The chamber was evacuated to 10
4torr and then refilled with argon. The cycle of evacuation and purging with argon was repeated three times. The temperature was increased to 17001C at a heating rate of 101C/min, held at 17001C for 10 min and then cooled to room temperature in the furnace. The bulk spec-imen was fully immersed in the titanium melt above the melting point of titanium (16681C), resulting in an interfacial reaction between Ti and Y2O3/ZrO2composites.

The phase identification of Ti and Y2O3/ZrO2samples was

performed using an X-ray diffractometer (XRD, Model MXP18, Mac Science, Yokohama, Japan). The operating con-ditions of X-ray diffraction were CuKa radition at 50 kV and 150 mA, and a scanning rate of 21 /min.

A scanning electron microscope (SEM, Model JSM 6500F, JEOL Ltd., Tokyo, Japan) was used for microstructural obser-vation on the interfaces between Ti and various Y2O3/ZrO2

samples. Cross-sectional SEM specimens were cut and ground using standard procedures and finally polished using diamond pastes of 6, 3, and 1 mm in sequence.

The compositions of various phases in the reaction layers were measured by an electron probe microanalyzer (EPMA, JXA-8800M, JEOL Ltd., Tokyo, Japan) with an atomic num-ber, absorption, and fluorescence correction (ZAF) program.17 The operating conditions for EPMA were as follows: the accel-erating voltage was 15 kV, the probe current was 1.5 10
8A, and the beam diameter was 1 mm.

The interfacial microstructures were also characterized using a transmission electron microscope (TEM, Model Tecnai 20, Philips, Eindhoven, Holland) equipped with an energy-disper-sive X-ray spectrometer (EDS, Model ISIS 300, Oxford Instru-ment Inc., London, UK). Cross-sectional TEM specimens were cut and then ground down toB80 mm with a diamond matted disk, polished with diamond pastes of 6, 3, and 1 mm in se-quence, and dimpled to a thickness of 50 mm. Finally the TEM specimens were ion milled by a precision ion miller (Model 691, Gatan Inc., Pleasanton, CA). Quantitative analyses of individ-ual phases in the reaction layers were conducted by the Cliff– Lorimer standardless technique.18

III. Results and Discussion

(1) XRD Analyses

Figure 1 shows the X-ray diffraction spectra of various Y2O3/

ZrO2 samples as well as pure Y2O3 after hot pressing. These

spectra were arranged for Y2O3, 90Y10Z, 70Y30Z, 50Y50Z,

30Y70Z, and 10Y90Z, respectively, in a sequence from top to bottom. X-ray phases of these hot-pressed Y2O3/ZrO2samples

are summarized in Table I. In the hot-pressed 90Y10Z, all of the zirconia was dissolved in yttria such that only cubic Y2O3was

detected. The hot-pressed 70Y30Z consisted of rhombohedral Zr3Y4O12and cubic Y2O3, while the XRD spectrum of the

hot-pressed 50Y50Z showed only rhombohedral Zr3Y4O12peaks.19

As for the hot-pressed 30Y70Z, only the cubic ZrO2phase was

detected because all of the Y2O3went into solid solution in

zir-conia. In other words, the cubic ZrO2was fully stabilized in

30Y70Z. However, c-ZrO2, t-ZrO2, and m-ZrO2were found in

10Y90Z or in partially stabilized ZrO2. In general, Y2O3 was

mutually dissolved or reacted with ZrO2as a solid solution or

Zr3Y4O12 in hot-pressed Y2O3/ZrO2 samples. Pascual and

Dura´n19found Y6ZrO11after sintering at 20001C for 3 h and

subsequent prolonged annealing. However, Y6ZrO11 was not

Table I. Designations, Compositions, Hot-Pressing Conditions, Relative Densities, and XRD Phases of Hot-Pressed Y2O3/ZrO2 Samples

Specimens Composition (vol%) Composition (mol%) Hot-pressing conditions Relative densities XRD phases

10Y90Z 10% Y2O3190% ZrO2 5% Y2O3195% ZrO2 15501C/30 min/1 atm Ar 98.0% c-ZrO2, t-ZrO2, m-ZrO2

30Y70Z 30% Y2O3170% ZrO2 17% Y2O3183% ZrO2 15501C/30 min/1 atm Ar 98.6% c-ZrO2

50Y50Z 50% Y2O3150% ZrO2 32% Y2O3168% ZrO2 15501C/30 min/1 atm Ar 98.8% Zr3Y4O12

70Y30Z 70% Y2O3130% ZrO2 52% Y2O3148% ZrO2 15501C/30 min/1 atm Ar 98.1% Y2O3, Zr3Y4O12

90Y10Z 90% Y2O3110% ZrO2 81% Y2O3119% ZrO2 16001C/30 min/1 atm Ar 98.1% Y2O3

100Y0Z 100% Y2O3 100% Y2O3 16001C/30 min/1 atm Ar 98.3% Y2O3

Fig. 1. X-ray diffraction spectra of as hot-pressed Y2O3/ZrO2samples.

found in this study because the equilibrium was probably not established.

(2) SEM and TEM Analyses

Figures 2(a)–(f) display backscattered electron images of cross-sections normal to the interfaces of Ti and various Y2O3/ZrO2

samples after reaction at 17001C for 10 min. Titanium is shown to the left of the micrograph, while zirconia is on the right-hand side. Arrows indicate the original interfaces of Ti and individual Y2O3/ZrO2 samples. The original interfaces were deliberately

located according to the characteristic Ka X-ray maps of

yttri-um (not shown), which was relatively immobile with respect to Zr, O, Ti, etc. The large pores in the ceramic side, as shown in Figs. 2(a) and (b), were attributed to the Kirkendall effect be-cause Zr and O diffused to the titanium side more rapidly than Ti diffused to the zirconia side.

Figure 2 indicates that extensive reactions took place at the interface between Ti and 10Y90Z. It was previously reported that needle-like a-Ti and some lamellar phases were usually found in the titanium side because of the interfacial reactions between Ti and ZrO2.

10,11,13,14

However, only a limited reaction took place on the titanium side at the interface between Ti and those samples containing more than 30 vol% Y2O3, while

Fig. 2. (a)–(f) Backscattered electron images of the cross section between Ti and Y2O3/ZrO2samples after reaction at 17001C for 10 min. Arrows

indicate the original interfaces between Ti and Y2O3/ZrO2samples.

90Y10Z and pure Y2O3reacted minimally with Ti melt. This

indicated that interfacial reactions were effectively suppressed in those samples containing more than 30 vol% Y2O3. This fact

plays an important role in the engineering aspect of Ti castings such that a controlled interfacial reaction results in a lower amount of a-casing and thus better mechanical properties. Even though the system became more stable with increasing Y2O3,

several reaction layers were found on the zirconia side after in-terfacial reactions between Ti and various Y2O3/ZrO2samples.

Microstructures of the reaction layers at the interface between Ti and various Y2O3/ZrO2samples were characterized using SEM/

EDS and TEM/EDS and the results are listed in Table II. The details will be described below.

(A) Reaction Layer ‘‘I’’ on the Metal Side: The reaction layer ‘‘I’’ was observed on the metal side of the interfaces

between Ti and Y2O3/ZrO2 samples containing less than 70

vol% Y2O3as shown in Figs. 2(a)–(d). An a-Ti phase with a

small amount of oxygen in solid solution was to the left of re-action layer ‘‘I.’’ The rere-action layer ‘‘I’’ consisted of b0-Ti(Zr, O) and/or acicular a-Ti(Zr, O) in the a-Ti matrix. A clear compar-ison could be made with Fig. 3, displaying the backscattered electron images of the reaction layer ‘‘I’’ in 10Y90Z, 30Y70Z, and 50Y50Z, respectively, at higher magnification. As shown in Fig. 3(a) together with Fig. 2(a), a large amount of b0_{-Ti(Zr, O)} and acicular a-Ti(Zr, O) ws precipitated in the a-Ti matrix at the interface between Ti and 10Y90Z. However, a relatively small amount of b0_{-Ti and acicular a-Ti existed at the interface} be-tween Ti and 30Y70Z [Fig. 3(b)] and only very few b0_{-Ti at the} interface between Ti and 50Y50Z [Fig. 3(c)]. For comparison, several reaction layers consisting of a-Ti(Zr, O), b0_{-Ti(Zr, O),} and Ti2ZrO were found in the diffusion couple of Ti and 3Y–

ZrO2after reactions ranging from 14001 to 15501C.11,13,14

(B) Reaction Layer ‘‘II’’ on the Ceramic Side: Figure 4 shows the backscattered electron images of reaction layers ‘‘II’’ on the outermost ceramic side of Ti/10Y90Z after reaction at 17001C for 10 min. b0-Ti (bright), acicular a-Ti (dark), and spherical c-ZrO2
xcoexisted in reaction layer ‘‘II’’ at the

inter-face between Ti and 10Y90Z, where zirconia was extensively dissolved in titanium. Based on EPMA results, b0-Ti in reaction layer ‘‘II’’ consisted of 43.98 at.% Ti, 25.97 at.% Zr, and 30.05 at.% O, indicating that b0_{-Ti was stabilized by dissolving a} sig-nificant amount of Zr. Concurrently, zirconia was retained as a cubic phase because it dissolved as high as 10.7 at.% Y as a sta-bilizer. This was consistent with the results found by Zhu et al.,6 indicating that Y was retained in cubic zirconia, when Y2O3

-sta-bilized zirconia was reacted with molten titanium. Lin and Lin13,14also indicated a two-phase region of b0_{-Ti and c-ZrO}

2

in the Ti/3Y–ZrO2diffusion couple after reaction at 15501C. It

was believed that reaction layer ‘‘II’’ was formed because titani-um melt infiltrated along the grain boundaries of 10Y90Z.

(C) Reaction Layers ‘‘III’’ and ‘‘IV’’ on the Ceramic Side: Microstructures of reaction layers ‘‘III’’ and ‘‘IV’’ at the interface between Ti and 30Y70Z were very different from those of the corresponding reaction layers previously found in the Ti/ZrO2 diffusion couples.13,14Figure 5 demonstrates

mi-crostructural variations in reaction layers ‘‘III’’ and ‘‘IV,’’ as well as the observation that reaction layer ‘‘III’’ consisted of a-Ti, acicular a-Ti, b0_{-Ti, and Y}

2O3at the interface between Ti

and 30Y70Z after reaction at 17001C for 10 min. Acicular a-Ti was found in b0_{-Ti as indicated by the arrow in Fig. 5(a). Based} on EPMA analyses, b0_{-Ti consisted of 57.92 at.% Ti, 20.31 at.%} Zr, and 21.77 at.% O, while the composition of acicular a-Ti was measured as 52.63 at.% Ti, 13.86 at.% Zr, and 33.51 at.% O. Because b0_{-Ti had a larger Zr/O ratio than a-Ti, b}0_{-Ti looked} brighter than a-Ti in backscattered electron images. The

com-Fig. 3. Backscattered electron images of reaction layer ‘‘I’’ in the tita-nium side at the interface between (a) Ti and 10Y90Z, (b) Ti and 30Y70Z, and (c) Ti and 50Y50Z after reaction at 17001C for 10 min.

Fig. 4. Backscattered electron image of reaction layer ‘‘II’’ at the in-terface between Ti and 10Y90Z after reaction at 17001C for 10 min.

position of Y2O3was measured as 5.93 at.% Ti, 34.51 at.% Y,

4.50 at.% Zr, and 55.06 at.% O, indicating a significant amount of Ti and Zr in solid solution.

Figure 5(b) displays the backscattered electron image of re-action layer ‘‘IV’’ at the interface between Ti and 30Y70Z after reaction at 17001C for 10 min. Reaction layer ‘‘IV’’ in 30Y70Z consisted of acicular a-Ti, b0_{-Ti, and Y}

2O3. b0-Ti was Ti with

zirconium and oxygen in solid solution, consisting of 52.9 at.% Ti, 27.6 at.% Zr, and 19.5 at.% O, while Y2O3was composed

of 3.0 at.% Ti, 37.1 at.% Y, 1.6 at.% Zr, and 58.3 at.% O, respectively. All compositions were measured using EPMA. Reaction layer ‘‘IV’’ dissolved a larger amount of Zr in Ti than reaction layer ‘‘III.’’ It was thus inferred that reaction layer ‘‘IV’’ was likely to be stabilized as the b phase at 17001C because Zr is an effective b-stabilizer. During cooling, b-Ti was transformed into b0-Ti accompanied by the precipitation of acicular a-Ti in the b0-Ti matrix.20

As 30Y70Z reacted with Ti melt, a large amount of Zr and O from 30Y70Z was dissolved in titanium, giving rise to the for-mation of Y2O3due to the very limited solubility of yttrium in

titanium. The precipitation of Y2O3was increased with

increas-ing Y2O3/ZrO2ratio. It was noted that no interfacial reactions

were found at the interface between Ti and 90Y10Z after reac-tion at 17001C for 10 min. It was concluded that increasing Y2O3 content was useful for better controlling the interfacial

reactions.

In the case of the Ti/ZrO2diffusion couple, Lin and Lin 13,14

found previously that acicular a-Ti precipitated in the b0_-Ti matrix of the titanium side . However, acicular a-Ti and the b0_{-Ti matrix existed in reaction layers ‘‘III’’ and ‘‘IV’’ of the} ceramic side in this study. This was attributed to the infiltration of titanium melt into the ceramic side, whereby reactions be-tween Ti and ZrO2took place. In contrast, the faster diffusion of

O and Zr into the titanium side led to the formation of the (a-Ti1b0-Ti) layers in the Ti–ZrO2diffusion couple.

Figure 6(a) shows the bright field image of the cross-section normal to the interface of Ti and 30Y70Z after reaction at 17001C for 10 min. The orientation relationship of acicular a-Ti and b0_{-Ti was identified to be [2110]}

a
Ti//[001]b0
_Ti and

(0001)a
Ti//(100)b0
_Ti as indicated in Fig. 6(b), in agreement

with the results presented by Lin and Lin.13,14_{The selected-area}

diffraction pattern depicted in Fig. 6(c) indicated the existence of cubic Y2O3in reaction layer ‘‘IV.’’

Microstructural evolution of reaction layer ‘‘III’’ at the interface between Ti and 30Y70Z at 17001C is schematically displayed in Fig. 7(a). Upon heating to 17001C, titanium melt infiltrated and dissolved a large amount of Zr and O, resulting in the formation of a two-phase (a-Ti1b-Ti) layer. Because the solubility of yttrium in titanium is quite limited, it remained as Y2O3[the center of Fig. 7(a)]. During cooling, cubic b-Ti was

transformed into orthorhombic b0-Ti, where a small amount of acicular a-Ti was precipitated [the right of Fig. 7(a)]. Figure 7(b) displays a proposed model of microstructural evolution in re-action layer ‘‘IV’’ at the interface between Ti and 30Y70Z at 17001C. Because less titanium melt infiltrated into reaction layer ‘‘IV,’’ Ti dissolved more concentrated Zr and O and existed as b-Ti during heating at 17001C. As the solubility of yttrium in titanium was quite limited, Y2O3was retained in the matrix of

b-Ti at high temperatures [the center of Fig. 7(b)]. During cool-ing, the cubic b-Ti was transformed into orthorhombic b0_-Ti where acicular a-Ti was precipitated in this reaction layer [Fig. 7(b), right side].

(D) Reaction Layer ‘‘V’’ on the Ceramic Side: Figures 8(a)–(c) show the bright field images of reaction layer ‘‘V’’ on the

Fig. 5. Backscattered electron images of (a) reaction layer ‘‘III’’ and (b) reaction layer ‘‘IV’’ at the interface between Ti and 30Y70Z after reaction at 17001C for 10 min.

Fig. 6. (a) Bright-field image of reaction layer ‘‘IV’’ at the interface between Ti and 30Y70Z after reaction at 17001C for 10 min; (b) selected-area diffraction patterns of acicular a-Ti and the matrix b0_{-Ti; (c) a}

selected-area diffraction pattern of Y2O3with the zone axis [111].

ceramic side far from the original interface of Ti and 10Y90Z, 30Y70Z, and 50Y50Z, respectively, after reaction at 17001C for 10 min. Figure 8(a) demonstrates that t-ZrO2
xwith two variants

was precipitated in c-ZrO2
x at the interface between Ti and

10Y90Z after reaction at 17001C for 10 min. Because Y2O3was

completely retained in ZrO2, no free Y2O3was found in 10Y90Z.

Figure 8(b) shows that several a-Zr particles were embedded in c-ZrO2
xin reaction layer ‘‘V’’ of 30Y70Z after reaction at 17001C

for 10 min. The composition of a-Zr was measured using EPMA, indicating that it contained 2.43 at.% Y, 69.14 at.% Zr, and 28.43 at.% O. Figure 8(c) shows the bright field image of reaction layer ‘‘V’’ in 50Y50Z after reaction at 17001C for 10 min. Residual Zr3Y4O12was found and its crystal structure was identified to be

rhombohedral based on the inset selected-area diffraction pat-tern, as shown in the upper right corner of Fig. 8(c). Figure 8(d) shows an EDS spectrum of this ternary compound. The average composition was calculated from six measurements to be 15.88 at.% Zr, 22.08 at.% Y, and 62.04 at.% O in correspondence withthe composition of Zr3Y4O12.

It was believed that the oxidation–reduction reaction rather than dissolution was the predominant reaction mechanism in reaction layer ‘‘V.’’ Dissolution did not play a significant role, as the titanium was not detected by EDS in reaction layer ‘‘V.’’ The oxidation–reduction reaction between Ti and ZrO2resulted in the

formation of metastable oxygen-deficient zirconia (ZrO_2
x). It was inferred that a-Zr precipitated from the supersaturated c-ZrO2
x, accompanied with an increasing O/Zr ratio in ZrO2
x21.

Furthermore, the amount of a-Zr decreased with increasing amounts of Y2O3 because Y2O3 could effectively suppress the

reaction between Ti and ZrO2.

(3) A General Discription of Interfacial Reaction Layers According to the above discussion, the reaction layers were formed at the interface between titanium and yttria/zirconia samples after reaction at 17001C for 10 min (summarized in Table II). Briefly speaking, extensive reactions occurred at the interface between Ti and 10Y90Z as previous studies indicat-ed.11-14 However, interfacial reactions were effectively sup-pressed by incorporating more than 30 vol% Y2O3. The

interfaces were increasingly stable with the amount of Y2O3

be-cause Y2O3functioned as a reaction barrier phase. On the metal

Fig. 8. Bright-field images of reaction layer ‘‘V’’ in the zirconia side far away from the interface between (a) Ti and 10Y90Z, (b) Ti and 30Y70Z, and (c) Ti and 50Y50Z after reaction at 17001C for 10 min. The inset in the upper right-hand corner of Fig. 8(c) is the selected-area diffraction pattern of Zr3Y4O12with the zone axis [112]; (d) an energy-dispersive spectrum of Zr3Y4O12.

Fig. 7. Schematic diagrams showing the microstructural evolution of (a) reaction layer ‘‘III’’ and (b) reaction layer ‘‘IV’’ at the interface be-tween Ti and 30Y70Z at various stages.

side, b0_{-Ti and acicular a-Ti were observed in the a-Ti matrix of} reaction layer ‘‘I’’ after Ti reacted with 10Y90Z or 30Y70Z, although only a small amount of b0-Ti was found in the a-Ti matrix of reaction layer ‘‘I’’ for the cases of 50Y50Z and 70Y30Z. Furthermore, no reaction layer ‘‘I’’ was formed for the cases of 90Y10Z and pure Y2O3. In the outermost ceramic

region, b0_{-Ti and a-Ti were found along with c-ZrO}

2
xin

reac-tion layer ‘‘II’’ of 10Y90Z, while b0_{-Ti and a-Ti were found} along with Y2O3in reaction layers ‘‘III’’ and ‘‘IV’’ of 30Y70Z,

50Y50Z, and 70Y30Z. Free Y2O3existed in reaction layers ‘‘III’’

and ‘‘IV’’ due to a very limited solubility of Y2O3in Ti when

ZrO2was completely dissolved in Ti. On the ceramic side far

from the original interface, dense a-Zr was formed in addition to residual ZrO2 in reaction layer ‘‘V’’ of 10Y90Z and 30Y70Z,

where a-Zr was excluded from metastable c-ZrO2
x. However,

a-Zr was formed in reaction layer ‘‘V’’ of 70Y30Z and 90Y10Z, because oxygen in ZrO2was extracted by Ti through the

oxida-tion–reduction mechanism. The amount of a-Zr decreased with increasing amounts of Y2O3. Although Y2O3was dissolved into

ZrO2as a solid solution or reacted with ZrO2as Zr3Y4O12during

hot pressing, Y2O3 was reprecipitated, due to the oxidation–

reduction mechanism and strong affinity of O and Zr to Ti, in 50Y50Z and 70Y30Z after reaction at 17001C for 10 min. It was also noted that some Zr3Y4O12grains were retained in 50Y50Z

after reaction at 17001C for 10 min..

IV. Conclusions

(1) The incorporation of more than 30 vol% Y2O3

signifi-cantly suppressed reactions at the interfaces between Ti and various Y2O3/ZrO2samples. In contrast, an extensive reaction

occurred at the interface between Ti and 10Y90Z (or a partially stabilized ZrO2) as mentioned previously.

(2) On the metal side, b0_{-Ti and acicular a-Ti were observed} in the a-Ti matrix after Ti melt reacted with 10Y90Z or 30Y70Z at 17001C for 10 min. However, only a very small amount of b0_{-Ti was found in the a-Ti matrix after Ti reacted with 50Y50Z} or 70Y30Z. Ti was almost kept intact after reaction with 90Y10Z or Y2O3at 17001C for 10 min.

(3) After reaction at 17001C for 10 min, b0_{-Ti, a-Ti and} c-ZrO2
xwere found in the outermost region of 10Y90Z, while

b0_{-Ti, a-Ti and Y}

2O3existed in the outermost region of 30Y70Z,

50Y50Z, and 70Y30Z. The formation of Y2O3in this region was

caused by the extensive dissolution of ZrO2in Ti together with a

very limited solubility of Y2O3.

(4) Dense a-Zr was formed along with residual ZrO2(cubic

and/or tetragonal) or Y2O3far from the original interface after

reaction with Ti at 17001C for 10 min. The amount of a-Zr de-creased with increasing amounts of Y2O3.

(5) Y2O3, existing as a solid solution or Zr3Y4O12after hot

pressing, was reprecipitated in 50Y50Z and 70Y30Z far from the original interface after reaction at 17001C for 10 min. This was due to the oxidation–reduction reaction and the strong Ti affin-ity of O and Zr.

(6) Some Zr3Y4O12grains were retained in the sample

con-taining 50 vol% Y2O3after reaction at 17001C for 10 min.

Acknowledgments

The authors would like to thank Mr. Wen-Shao Liao at the Department of Material Sciences and Engineering in National Chiao Tung University, Hsinchu, Taiwan, for preparing the SEM specimens.

References

1

A. I. Kahveci and G. E. Welsch, ‘‘Effect of Oxygen on the Hardness and Alpha/Beta Phase Ration of Ti–6Al–4V Alloy,’’ Scr. Metall., 20 [9] 1287–90 (1986).

2

K. I. Suzuki, S. Watakabe, and K. Nishikawa, ‘‘Stability of Refractory Oxides for Mold Material of Ti–6Al–4V Alloy Precision Casting,’’ J. Jpn. Inst. Met., 60 [8] 734–43 (1996).

3

G. Welsch and W. Bunk, ‘‘Deformation Modes of the Alpha-Phase of Ti– 6Ai–4V as a Function of Oxygen Concentration and Aging Temperature,’’ Metall. Trans. A, 13A [5] 889–99 (1982).

4

R. L. Saha and K. T. Jacob, ‘‘Casting of Titanium and its Alloy,’’ Def. Sci., 36 [2] 121–41 (1986).

5

B. C. Weber, H. J. Garrett, F. A. Mauer, and M. A. Schwartz, ‘‘Observations on the Stabilization of Zirconia,’’ J. Am. Ceram. Soc., 39 [6] 197–07 (1956).

6

J. Zhu, A. Kamiya, T. Yamada, W. Shi, K. Naganuma, and K. Mukai, ‘‘Sur-face Tension, Wettability and Reactivity of Molten Titanium in Ti/Yttria-Stabi-lized Zirconia System,’’ Mater. Sci. Eng. A, 327, 117–27 (2002).

7

B. C. Weber, W. M. Thompson, H. O. Bielstein, and M. A. Schwartz, ‘‘Ceramic Crucible for Melting Titanium,’’ J. Am. Ceram. Soc., 40 [11] 363–73 (1957).

8

G. Economos and W. D. Kingery, ‘‘Metal–Ceramic Interactions: II, Metal Oxide Interfacial Reactions at Elevated Temperatures,’’ J. Am. Ceram. Soc., 36 [12] 403–9 (1953).

9

C. E. Holcombe and T. R. Serandos, ‘‘Consideration of Yttria for Vacuum Induction Melting of Titanium,’’ Metall. Trans., 14B [9] 497–8 (1983).

10

K. F. Lin and C. C. Lin, ‘‘Transmission Electron Microscope Investigation of the Interface between Titanium and Zirconia,’’ J. Am. Ceram. Soc., 82 [11] 3179– 85 (1999).

11

K. L. Lin and C. C. Lin, ‘‘Ti2ZrO Phases Formed in the Titanium and

Zir-conia Interface after Reaction at 15501C,’’ J. Am. Ceram. Soc., 88 [5] 1268–72 (2005).

12

K. L. Lin and C. C. Lin, ‘‘Zirconia-Related Phases in the Zirconia/Titanium Diffusion Couple after Annealing at 11001 to 15501C,’’ J. Am. Ceram. Soc., 88 [10] 2928–34 (2005).

13

K. L. Lin and C. C. Lin, ‘‘Microstructural Evolution and Formation Mech-anism of the Interface Between Titanium and Zirconia Annealed at 15501C,’’ J. Am. Ceram. Soc., 89 [4] 1400–8 (2006).

14

K. L. Lin and C. C. Lin, ‘‘Effects of Annealing Temperature on Microstruc-tural Development at the Interface Between Zirconia and Titanium,’’ J. Am. Ceram. Soc., 90 [3] 893–9 (2007).

15

K. F. Lin and C. C. Lin, ‘‘Interfacial Reactions between Zirconia and Tita-nium,’’ Scr. Metall., 39 [10] 1333–8 (1998).

16

K. F. Lin and C. C. Lin, ‘‘Interfacial Reactions between Ti–6Al–4V Alloy and Zirconia Mold During Casting,’’ J. Mater. Sci., 34, 5899–906 (1999).

17

J. I. Goldstein, Scanning Electron Microscopy and X-ray Microanalysis, 2nd edition, Plenum Press, New York, 1992.

18

G. Cliff and G. W. Lorimer, ‘‘The Quantitative Analysis of Thin Specimens,’’ J. Microsc., 130 [3] 203–7 (1975).

19

C. Pascual and P. Dura´n, ‘‘Subsolidus Phase Equilibria and Ordering in the System ZrO2–Y2O3,’’ J. Am. Ceram. Soc., 66 [1] 23–7 (1983).

20

R. F. Domagala, S. R. Lyon, and R. Ruh, ‘‘The Pseudobinary Ti–ZrO2,’’

J. Am. Ceram. Soc., 56 [11] 584–7 (1973).

21

R. J. Ackermann, S. P. Garg, and E. G. Rauh, ‘‘High-Temperature Phase Diagram for the System Zr–O,’’ J. Am. Ceram. Soc., 60 [7–8] 341–5 (1977). & Table II. Reaction Layers Formed at the Interfaces of Ti and Y2O3/ZrO2Samples After Reaction at 17001C/10 min

Specimens

Mol% Y2O3

Interface reaction layers and phases

Reaction layer ‘‘V’’ in the ceramic side

Ti side Ceramic side

10Y90Z 5 I a-Ti, b0_{-Ti1acicular a-Ti II b}0_{-Ti1acicular a-Ti, c-ZrO}

2
x c-ZrO2
x, t-ZrO2
x, a-Zr

30Y70Z 17 I a-Ti, b0_{-Ti1acicular a-Ti III Y}

2O3, a-Ti, b0-Ti1acicular a-Ti c-ZrO2
x, a-Zr IV Y2O3, b0-Ti1acicular a-Ti

50Y50Z 32 I a-Ti, b0_{-Ti III Y}

2O3,a-Ti, b0-Ti1acicular a-Ti Y2O3, Zr3Y4O12, a-Zr IV Y2O3, b0-Ti1acicular a-Ti

70Y30Z 52 I a-Ti, b0-Ti III Y2O3, a-Ti, b0-Ti1acicular a-Ti Y2O3, a-Zr

90Y10Z 81 Insignificant interfacial reaction Y2O3, a-Zr

100Y0Z 100 Insignificant interfacial reaction Y2O3