Reaction between titanium and zirconia powders during sintering at 1500 degrees C

(1)

Reaction Between Titanium and Zirconia Powders During

Sintering at 15001C

Kun-Lin Lin and Chien-Cheng Lin*

,w

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

Zirconia–titanium (ZrO2–Ti) composites have been considered potential thermal barrier graded materials for applications in the aerospace industry. Powder mixtures of Ti and 3 mol% Y2O3partially stabilized ZrO2in various ratios were sintered at 15001C for 1 h in argon. The microstructures of the as-sintered composites were characterized by X-ray diffraction and trans-mission electron microscopy/energy-dispersive spectroscopy. Ti reacted with and was mutually soluble in ZrO2, resulting in the formation ofa-Ti(O, Zr), Ti2ZrO, and/or TiO. These oxygen-containing phases extracted oxygen ions from ZrO2, whereby oxygen-deficient ZrO2was generated. For relatively small Ti/ ZrO2ratios, specimens withr30 mol% Ti, TiO were formed as oxygen could be sufficiently supplied by excess ZrO2. For the specimens with 50 mol% Ti, lamellar Ti2ZrO was precipitated ina-Ti(Zr, O), with no TiO being found. Both m-ZrO2xand t-ZrO2xwere found in specimens withr50 mol% Ti; however, only c-ZrO2xwas formed in the specimen with 70 mol% Ti. As ZrO2 was gradually dissolved into Ti, yttria was retained in ZrO2because of the very limited solubility of yttria ina-Ti(O, Zr) or TiO. The concentration of retained yttria and the degree of oxygen deficiency in ZrO2increased with the Ti content. The complete dissolution of ZrO2 into Ti was followed by the precipitation of Y2Ti2O7in the specimen with 90 mol% Ti.

I. Introduction

I

T is well known that partially stabilized zirconia (PSZ) has good toughness for wide structural applications,1,2 and its mechanical properties can be enhanced by incorporating titani-um (Ti).3–8 Weber et al.3 found that sintered zirconia (ZrO2)

crucibles containing 15 at.% Ti showed good strength and ther-mal shock resistance. Arias6,7 investigated the thermal shock resistance of ZrO2with 15 mol% Ti after sintering at 18701C for

1 h in vacuum, claiming that the grain growth of ZrO2 was

inhibited due to the addition of Ti so that both thermal shock resistance and strength were increased. For ZrO2 with 5–15

mol% Ti sintered at 17401C for 1 h under a vacuum of 101– 102torr, Lin et al.8_{reported that the ZrO}

2was partially

sta-bilized as cubic and tetragonal phases, and its thermal shock resistance was significantly improved because of the dissolution of TiO. However, without taking into account possible reactions between Ti and ZrO2, they hypothesized that the formation of

TiO was caused by residual oxygen in the vacuum furnace. In the past decade, ZrO2–Ti functionally graded materials

(FGMs) have been considered potential thermal barrier graded materials for applications in the aerospace industry. For the ZrO2–Ti FGMs, Teng and colleagues9,10showed that only

a-Ti(O), tetragonal (t)-ZrO2, and m-ZrO2were found in various

Ti/ZrO2composites after annealing from 14001 to 16501C. They

reported that the volume fraction of m-ZrO2increased with the

Ti content in the Ti/ZrO2composites, while the interfacial

stress-es, arising from the plastic deformation of Ti and the thermal expansion mismatch of Ti and ZrO2, were the driving forces for

the phase transformation from t-ZrO2to m-ZrO2.

The interfacial reaction between Ti and ZrO2has been

stud-ied in the past few decades.3–5,11–13These studies indicated that oxygen was dissolved into Ti, causing the ZrO2to become

ox-ygen-deﬁcient and blackened. Distinct reaction layers were formed at the interface between Ti and ZrO2.5,14–17 Lin and

Lin14,16–21 have conducted an investigation on the diffusional reaction between Ti and ZrO2 at temperatures ranging from

11001 to 15501C using transmission electron microscopy/energy-dispersive spectroscopy (TEM/EDS), electron probe microanal-ysis (EPMA), and scanning electron microscopy (SEM)/EDS. They found that a-Ti, b0_{-Ti, the ordered Ti suboxide (Ti}

3O), and

lamellar and spherical Ti2ZrO were formed after annealing from

11001 to 17001C.14,20At the ZrO2side far away from the

inter-face of Ti/ZrO2, Lin and Lin 21

observed twinned t0_-ZrO

2x,

lenticular t-ZrO2x, ordered c-ZrO2x, and intergranular a-Zr

after reaction at 15501C.

The review of the literature implies that some contradictions exist with respect to the reaction products and the origin of ox-ygen in Ti oxides and/or solid solutions. In the present study, powder mixtures of Ti and ZrO2in various ratios were sintered

at 15001C for 1 h in argon. The microstructures of the as-sin-tered composites were characterized using X-ray diffraction (XRD) and TEM/EDS. The reaction mechanisms between Ti and ZrO2were elucidated.

II. Experimental Procedure

The starting powders used in this study were 3 mol%Y2O3

PSZ (494 wt% ZrO21HfO2, 5.4 wt% Y2O3,o0.001 wt% Fe2O3,

o0.01 wt% SiO2,o0.005 wt% Na2O,o0.005 wt% TiO2,o

0.02 wt% Cl,o0.005 wt% SO42, 0.3 mm in average, Toyo Soda

Mfg. Co., Tokyo, Japan), and commercially pure Ti (with a nominal composition of 99.31 wt% Ti, 0.25 wt% O, 0.01 wt% H, 0.03 wt% N, 0.10 wt% C, 0.30 wt% Fe., 60–70 mm average diameter, Alfa Aesar, Ward Hill, MA). Several composites in various Ti/ZrO2ratios were prepared by pressureless sintering.

The compositions and designations are summarized in Table I. The Ti/ZrO2composites contained 10, 30, 50, 70, and 90 mol%

Ti, respectively, with ZrO2in residue. The composite containing

10 mol% Ti and 90 mol% ZrO2was designated as 10T90Z, and

so on.

The mixtures of ZrO2and Ti powders were ball milled (with

ZrO2milling media) in ethanol (containing 1 wt% polyvinyl

al-cohol, or PVA, as a binder for forming green bodies) for 24 h. The slurries were dried in an oven at 1501C, ground with an agate mortar and pestle, and then shaken vigorously in a plastic bottle. The powder mixtures were pressed into disks (10 mm in diameter 3 mm thick) at a pressure of 150 MPa and then sin-tered in argon at 15001C for 1 h at 101 and 51C/min heating and cooling rates, respectively.

T. Besmann—contributing editor

Research supported by National Science Council of Taiwan under Contract No. NSC 94-2216-E-009-011.

*Member, American Ceramic Society. w

Author to whom correspondence should be addressed. e-mail: chienlin@faculty. nctu.edu.edu.tw

Manuscript No. 22492. Received November 14, 2006; approved February 15, 2007.

Journal

r2007 The American Ceramic Society

(2)

After sintering, the linear shrinkage (a) was calculated by the following equation: a 5 [(d0d)/d0] 100%, where d0and d are

the measured diameters of the green body and the sintered Ti/ ZrO2composite, respectively. The apparent densities (ra) of all

the Ti/ZrO2powder mixtures were measured by gas pycnometry

using 99.99% pure helium gas (Model MultiVolume Pycno-meter 1305, Micromeritics, Norcross, GA). The bulk densities (rb) of the sintered Ti/ZrO2composites were determined by the

Archimedes method using water as an immersed medium. For a nonporous powder, its apparent density approximates to the true density and can be used as the reference point in calculating the percentage of the theoretic density of a sintered body. Thus, the relative densities (rrb) of the sintered specimens were given as

follows: rrb5(rb/ra) 100%, assuming that the change in

den-sity owing to the reaction was negligible. This assumption is reasonable as most of the reaction products are ZrO2x and

a-Ti(O, ZrO2)- derived compounds (e.g. a-Ti(O, Zr), Ti2ZrO,

and TiO). The only concern is Y2Ti2O7; however, its amount is

insignificant with respect to other phases. Then, the relative densities (rrg) of the green bodies were calculated by the

follow-ing relationship: rrg5rrb (1a)3.

The microstructural characterization of sintered Ti/ZrO2

composites was carried out using an x-ray diffractometer (Mod-el MXP18, Mac Science, Yokohama, Japan) and an analytical transmission electron microscope (Model Tecnai 20, Philips, Eindhoven, the Netherlands) equipped with an energy-disper-sive x-ray spectrometer (Model ISIS 300, Oxford Instruments Inc., London, U.K). The measurement conditions of XRD were CuKa radiation at 50 kV, 150 mA, and a scanning rate of 21/min. The TEM specimens were cut, ground, polished, dimpled, and then ion milled by standard procedures. The Cliff–Lorimer22 _{standardless technique was used to analyze}

the compositions of the various phases. However, the composi-tions of oxygen-deﬁcient ZrO2in various Ti/ZrO2composites

were measured using stoichiometric ZrO2as a standard.

III. Results and Discussion

(1) Appearance, Shrinkage, and Densities of Sintered Bodies Figures 1(a) and (b) display the photographs of the green bodies and the sintered Ti/ZrO2 composites, respectively. For green

bodies, the 90T10Z appeared dark gray, while the 10T90Z was light gray. After sintering, the 90T10Z had a metallic luster sim-ilar to Ti. However, the sintered 10T90Z became blackened as it was mainly composed of oxygen-deﬁcient ZrO2after sintering at

15001C/1 h in argon. Obviously, the color of the sintered 10T90Z was similar to that of oxygen-deﬁcient ZrO2.

The linear shrinkage (a) of the sintered Ti/ZrO2composites

with respect to their corresponding green bodies is shown in Table II. It was found that the linear shrinkage of the sintered Ti/ZrO2 composite increased with increasing ZrO2 content.

Table II also shows the measured apparent densities (ra) of

powder mixtures and the measured bulk densities (rb) of

sin-tered Ti/ZrO2composites. The relative density (rrb) of the

sin-tered body, calculated by rrb5rb/ra 100%, increased with

increasing ZrO2content due to a decrease in the porosity. It was

noticeable that the sintered 10T90Z had a high relative density of about 98%. The densiﬁcation of Ti/ZrO2composites by

solid-state sintering could have been enhanced by the highly active major component, i.e., submicrometer ZrO2. In contrast, the

relative density of the green body increased with increasing Ti content as Ti particles underwent plastic deformation during cold pressing. The calculated relative density (rrg) of the 90T10Z

green body was about 83%, while that of the 10T90Z green body was about 58%. However, the densiﬁcation of the 90T10Z was insignificant (about 2%) during sintering probably due to the very large particle size of Ti powder, which was 60–70 mm in diameter.

(2) XRD Analyses

Figure 2 shows the XRD spectra of various sintered Ti/ZrO2

composites. The sintered 10T90Z consisted of m-ZrO2, t-ZrO2,

and TiO; however, no peaks of a-Ti were observed. The a-Ti appeared in all specimens except in the sintered 10T90Z. Ti2ZrO

instead of TiO was found in sintered composites containing more than 50 mol% Ti.

As for ZrO2, cubic (c)-ZrO2 and t-ZrO2 could be

distin-guished by the splitting {002} and {200} reﬂections of the t-ZrO2. By comparing the XRD spectra of the sintered 10T90Z,

30T70Z, and 50T50Z, it was seen that the (002) reﬂection peak of t-ZrO2was gradually diminished with decreasing ZrO2

con-tent. The (002) and (200) reﬂections of t-ZrO2were eventually

replaced by the (200) reﬂection of c-ZrO2 in the sintered

70T30Z. c-ZrO2was fully stabilized in the sintered 70T30Z as

excessive Ti gave rise to a large concentration of oxygen vacan-cies in ZrO2.22 The stabilization of c-ZrO2was unlikely to be

caused by the grain size, because the ZrO2grain grew

apprecia-bly up to about 2.5–4 mm in the sintered 70T30Z [see Fig. 5(a)]. Neither m-ZrO2 nor t-ZrO2 peaks were found in the sintered

70T30Z and 90T10Z. Furthermore, a trace of c-ZrO2remained

in the sintered 90T10Z because ZrO2was almost dissolved in Ti.

The present results were inconsistent with those reported by Teng et al.,9,10who found only t-ZrO2 and m-ZrO2, but no

c-ZrO2, Ti2ZrO, or Y2Ti2O7 (shown in the TEM analyses),

in various Ti/ZrO2 composites. They also indicated that the

Fig. 1. (a) Ti/ZrO2green bodies; (b) sintered composites after sintering

at 15001C/1 h in argon.

Table I. Compositions of Ti/ZrO2Composites

Specimensw Composition (mol%)

90T10Z 90 mol % Ti110 mol% ZrO2

w

The powder mixtures were pressed into disks at a pressure of 150 MPa and then sintered at 15001C/1 h in argon.

Table II. Linear Shrinkages (a), Apparent Densities of Powder Mixtures (qa), Calculated Relative Densities of Green Bodies (qrg), and Bulk Densities (qb) and Relative Densities (qrb)

of Sintered Bodies Specimen Aw(%) ra(g/cm 3 ) rrgz(%) rb(g/cm 3 ) rrb(%) 90T10Z 0.8 4.07 83 3.48 85 70T30Z 2.6 4.27 80 3.69 86 50T50Z 8.4 4.57 68 4.05 88 30T70Z 13.1 4.64 60 4.26 92 10T90Z 16.2 5.02 58 4.94 98 w

Calculated by a 5 [(d0d)/d0] 100%, where d0and d are the measured

di-ameters of the green and the sintered bodies, respectively. z_{Calculated by}

rrg5rrb (1a) 3

.

(3)

volume fraction of m-ZrO2in Ti/ZrO2composites increased with

the Ti content, and the interfacial stress arising from the thermal expansion mismatch and the plastic deformation of Ti enhanced the t-ZrO2-m-ZrO2transformation.10However, in the present

study, the amount of m-ZrO2decreased with increasing Ti

con-tent, as the reaction between Ti and ZrO2gave rise to retained

yttria and oxygen vacancies, of which the concentrations in ZrO2increased with the Ti content. It was also noted that the

c-ZrO2was fully stabilized due to the high concentrations of

re-tained yttria and oxygen vacancies in ZrO2 of the sintered

70T30Z.

(3) Microstructures of Various Ti/ZrO2Composites Figure 3(a) shows a bright-ﬁeld image of TiO, m-ZrO2x, and

(t1m)-ZrO2x in the 10T90Z after sintering at 15001C/1 h.

Based on the Ti distribution, it appeared that hard ZrO2

particles were embedded in soft Ti particles, which were subject-ed to plastic deformation during cold pressing. Figures 3(b) and (c) show the selected area diffraction patterns (SADPs) of TiO with [001] and [111] zone axes, respectively. TiO has the B1(NaCl) structure and its lattice parameter was calculated as 4.43 A˚. From the EDS spectrum, shown in Fig. 3(d), TiO dissolved a significant amount of ZrO2and was composed of 45.7 at.% Ti, 50.2 at.% O,

and 4.2 at.% Zr. It indicated that the stabilizer of ZrO2(i.e. yttria)

was not dissolved into TiO within the limits of detection of EDS. Figure 3(e) shows the EDS spectrum of m-ZrO2x, which

con-sisted of 57.6 at.% Zr, 40.5 at.% O, and 1.8 at.% Y, correspond-ing to ZrO1.39with Y2O3in solid solution. Figure 3(f) shows the

SADPs, as well as the redrawn diagram, of (t1m)-ZrO2xwith

[011]t//[011]mand (200)t// (100)m. The small lathe of m-ZrO2x

originated at the grain boundaries of t-ZrO2xwith a stress-ﬁeld

contrast (labeled as ‘‘t1m’’ in Fig. 3(a)), owing to the stress con-centrations at such a region.23,24For the sintered 10T90Z and 30T70Z, the interfacial stress ﬁeld arose from the thermal expan-sion mismatch and plastic deformation of Ti or TiO, resulting in an enhanced driving force for the t-ZrO2x-m-ZrO2x

trans-formation.10 Several martensite lathes have grown completely across the grain and changed to twinned m-ZrO2x(labeled as

‘‘m’’ in Fig. 3(a)). Figure 3(g) shows the SADP, as well as the re-drawn diagram, of m-ZrO_2xwith the [111] zone axis. The twin-ning plane of m-ZrO2xwas identiﬁed to be (0 1 1). It was noted

that part of Ti was also oxidized and became TiO in the sintered 30T70Z as well.

Figure 4(a) displays the bright-ﬁeld image of a-Ti(Zr, O), m-ZrO2x, and (t1m)-ZrO2xin the 50T50Z composite after

sinte-ring at 15001C/1 h. For comparison, the oxide TiO was found in the sintered 10T90Z instead of a-Ti(Zr, O). As the Ti content increased above 50 mol%, a-Ti(Zr, O) rather than TiO was formed because the supply of oxygen from ZrO2 was limited

compared with the ZrO2-rich composites. Based upon the Ti–O

phase diagram25 and EDS results (Table III), a-Ti dissolved o33 at.% O and the O/Zr ratio was o1.14 in composites con-taining 50 mol% Ti. As shown in Fig. 4(a), the a-Ti(Zr, O) contained many dislocations, which were likely caused by the

Fig. 2. X-ray diffraction spectra of Ti/ZrO2composites.

Fig. 3. (a) Transmission electron micrograph (bright-ﬁeld image) of TiO, m-ZrO2x, and (t1m)-ZrO2xin the 10T90Z composites after sintering at

15001C/1 h; (b) and (c) selected area diffraction patterns (SADPs) of TiO with the [001] and [111] zone axes, respectively; (d) and (e) energy-dispersive spectra of TiO and m-ZrO2x, respectively; (f) SADPs of the (t1m)-ZrO2xalong the [011]tor [011]mzone axis; (g) an SADP of the m-ZrO2xalong the

(4)

extended plastic deformation during fabrication. Dislocations disappeared when Ti was oxidized as TiO as illustrated in Fig. 3(a). Figure 4(b) displays the SADP of a-Ti(Zr, O) with the [2 1 1 0] zone axis. The lattice constants of hexagonal a-Ti(Zr, O) were calculated to be a 3.27 A˚ and c 4.92 A˚. Significant amounts of Zr and O were dissolved into the a-Ti. Figure 4(c) shows the EDS spectrum of a-Ti(Zr, O), revealing that it comprised 73.5 at.% Ti, 7.9 at.% Zr, and 18.6 at.% O. The solubility limit of ZrO2 in Ti was reported to be up to

approximately 10 mol% in an early study conducted by Ruh.4

Figure 4(c) also shows that the solubility of yttria was very lim-ited in a-Ti(O, Zr).

In previous studies,8–10_Ti/ZrO

2composites with more than

50 mol% Ti showed no Ti-containing phases except for a-Ti (Zr, O). However, Ti2ZrO was precipitated in a-Ti(Zr, O)

when the Ti content increased above 50 mol% in the present study. Figure 5(a) shows the bright-ﬁeld image of c-ZrO2x

and the lamellae of a-Ti1Ti2ZrO in the sintered 70T30Z.

From the SADPs shown in Fig. 5(b), the lamellar phases (a-Ti1Ti2ZrO) were identiﬁed as orthorhombic Ti2ZrO and

hexagonal a-Ti, respectively, with the orientation relation-ship of 111½ _Ti₂_ZrO== 1213_aTi and 110_Ti

2ZrO== 1010

a-Ti. The lattice constants of the Ti2ZrO orthorhombic unit cell

were calculated to be ao 4.91 A˚, bo 8.21 A˚, and co 3.19 A˚,

and those of the a-Ti hexagonal unit cell were ah 3.11 A˚,

ch 4.72 A˚. The lamellae of a-Ti1Ti2ZrO had been found in the

Ti/ZrO2 diffusion couple 16,20

; however, they were not found previously in Ti/ZrO2 composite systems. The c-ZrO2x was

found in 70T30Z and identiﬁed by the SADP shown in Fig. 5(c), which was taken along the zone axis of [111]. Based on its struc-ture type, the reﬂections of the type odd, odd, even were not allowed for c-ZrO2x. This fact was applied to distinguish

between c-ZrO2xand t-ZrO2x. However, the {112}-type

re-ﬂections appeared in the SADPs of c-ZrO2xbecause a large

concentration of oxygen vacancies caused the change in the structure factor of c-ZrO2x. A similar result also indicated the

existence of c-ZrO2xfrom the XRD analyses (Fig. 2).

In sintered 90T10Z, ZrO2, together with its stabilizer (yttria),

was completely dissolved in Ti. Because of the very low solu-bility of yttrium in Ti (Fig. 4(c)), Y2Ti2O7was precipitated from

the a-Ti(Zr, O). Figure 6(a) shows the bright-ﬁeld image of a-Ti(Zr, O) and Y2Ti2O7in the 90T10Z composite after

sinte-ring at 15001C/1 h. It indicated that both intergranular and intragranular Y2Ti2O7existed in the a-Ti (Zr, O) matrix. The

precipitates of Y2Ti2O7 were coarsened through the Ostwald

ripening effect so that they were spherical in shape. Figures 6(b) and (c) show the SADPs of Y2Ti2O7along the zone axes of [111]

and [001], respectively. The calculated lattice parameter of the Y2Ti2O7unit cell was about 11.08 A˚. The Y2Ti2O7possesses the

pyrochlore structure in which one out of every eight oxygen ions is missing in the stoichiometric ﬂuorite. The ideal Y2Ti2O7

struc-ture has a cubic unit cell consisting of eight ﬂuorite-type cells with ordered oxygen vacancies. The calculated lattice parameter of Y2Ti2O7(a 5 11.08 A˚) is approximately twice that of ﬂuorite

Fig. 4. (a) Transmission electron micrograph (bright-ﬁeld image) of a-Ti, m-ZrO2x, and (t1m)-ZrO2xin the 50T50Z composite after

sinte-ring at 15001C/1 h; (b) a selected area diffraction pattern of a-Ti(Zr, O) along the [2 1 1 0] zone axis; and (c) an energy-dispersive spectrum of a-Ti(Zr, O).

Table III. Compositions of ZrO2xin Various Ti/ZrO2 Composites After Sintering

Specimens Composition of ZrO2x(at.%) w O/Zr O/Zr K-factorz ZrO2x (x value) Zr O Y Ti 10T90Z 57.6 40.5 1.8 0.1 0.71 1.39 0.6170.07 30T70Z 59.4 37.7 1.9 1.1 0.64 1.26 0.7470.04 50T50Z 61.9 35.7 2.0 0.4 0.58 1.14 0.8670.05 70T30Z 61.8 33.8 3.8 0.6 0.55 1.08 0.9270.03 w

Used stoichiometric ZrO2as a standard.zk-factor is 1.97.

Fig. 5. (a) Transmission electron micrograph (bright-ﬁeld image) of c-ZrO2x and lamellae a-Ti1Ti2ZrO in the 70T30Z composite after

sintering at 15001C/1 h; (b) selected area diffraction patterns (SADPs) of a-Ti and Ti2ZrO along the zone axes of [111]Ti2ZrO// 1 2 1 3a-Ti; (c) an

SADP of c-ZrO2xalong the zone axis of [111].

(5)

(a 5 5.10 A˚). The EDS in Fig. 6(d) indicates that Y2Ti2O7

con-tained 26.4 at.% Y, 33.4 at.% Ti, and 40.2 at.% O. (4) Formation of TiO, Ti2ZrO, and a-Ti(O, Zr)

Oxygen-deﬁcient ZrO2was designated as ZrO2x, where x is the

degree of oxygen deﬁciency. The value of x increased as the Ti content increased and ranged from 0.61 to 0.92, as listed in Table III. As a reference, stoichiometric ZrO2and pure Ti were

also annealed under the same sintering conditions. The reference ZrO2 became oxygen deﬁcient and its composition was

mea-sured as ZrO2x(x 5 0.39), while the reference Ti dissolved a

negligible amount of oxygen ( 0.2 at.%). The concentration of yttria in ZrO2increased with the Ti content (Table III).

Togeth-er with the fact that the Y2O3was nearly insoluble in TiO and

a-Ti(Zr, O) (Figs. 3(d) and 4(c)), it was inferred that the stabi-lizer of ZrO2(yttria) was mainly retained in residual ZrO2 as

ZrO2particles were gradually dissolved into Ti. Based on XRD

and TEM/EDS results, it was the oxygen released from ZrO2

that resulted in the formation of a-Ti(O), Ti2ZrO, and/or TiO. If

the supply of oxygen from ZrO2was sufﬁcient, for example, in

the sintered 10T90Z and 30T70Z, TiO was formed. Otherwise, a-Ti(O, Zr) and Ti2ZrO were formed such as in the specimens

containingo50 mol% ZrO2.

Based on commonly cited Ellingham diagrams, Ti should not reduce ZrO2. Such diagrams have led to erroneous conclusions

in some studies.8,26Lin et al.8and Bannister and Barnes26 hy-pothesized that Ti was oxidized by residual oxygen in the cham-ber, resulting in the formation of TiO or TiO2. TiO could be

formed when the vacuum was in the range of 101–102torr. If the vacuum was poor, TiO2was formed and resulted in no

sta-bilization of ZrO2. While the vacuum level was better than

5 104torr, no TiO could be formed and therefore no stabi-lization of ZrO2was observed. In fact, the Ellingham diagram

does not accurately describe the reactions in the Ti–ZrO2system

because it does not account for solution formation or the for-mation of intermediate compounds like a-Ti(Zr,O), TiO, Ti2ZrO, and/or ZrO2x. Based on the present results, it was

be-lieved that the reaction between Ti and ZrO2was controlled by

reduction–dissolution mechanisms. This conclusion was consis-tent with those observed in the Ti/ZrO2diffusion couples,

con-ducted at high temperatures ranging from 11001 to 15501C by Lin and Lin.16,17,20,21

(5) Grain Growth of ZrO2

The grain size of m- or t-ZrO2in the sintered 10T90Z was about

0.3 mm (Fig. 3(a)) similar to the original ZrO2particle size. The

formation of TiO effectively retarded the grain growth of ZrO2

in the sintered 10T90Z. This was consistent with previous re-sults, which indicated that the pinning of the intergranular TiO decreased the grain size of ZrO2and improved its strength and

thermal shock resistance.8 However, the grain size of m- or t-ZrO2 increased to about 1–2 mm in the sintered 50T50Z

(Fig. 4(a)). In the sintered 70T30Z, the c-ZrO2was found

in-stead of m- or t-ZrO2and its grain size grew to about 2.5–4 mm

(Fig. 5(a)). Thus, the ZrO2grain size was affected by the Ti

con-tent in Ti/ZrO2composites. It was believed that oxygen

diffu-Fig. 6. (a) Transmission electron micrograph (bright-ﬁeld image) of a-Ti(Zr, O) and Y2Ti2O7in the 90T10Z composite after sintering at 15001C/1 h; (b)

and (c) the selected area diffraction patterns s of the Y2Ti2O7along the zone axes of [111] and [001], respectively; (d) an energy-dispersive spectrum of

(6)

sion through a vacancy mechanism was predominant for the grain growth of ZrO2in the Ti/ZrO2composites. Lin and Lin21

reported that the rate of reduction–dissolution between ZrO2

and Ti increased with temperature, leading to a significant in-crease in oxygen vacancies in ZrO2, which increased diffusion

velocities and enhanced the grain growth of ZrO2x. Previous

studies indicated that a-Zr was segregated to the ZrO2 grain

boundaries by the exsolution of zirconium from metastable ZrO_2xin the region far away from the original interface dur-ing cooldur-ing.17,21_{However, no intergranular a-Zr was found in}

Ti/ZrO2 composites as zirconium was readily dissolved in the

abutting Ti to form a-Ti(Zr, O) in this study.

According to the foregoing discussion, the reaction products, crystal structures, and morphology in Ti/ZrO2composites

sin-tered at 15001C/1 h are summarized in Table IV. The interfacial reactions between Ti and ZrO2powders are briefly described as

follows: During sintering, Zr and O diffused into Ti, whereas Ti simultaneously diffused into ZrO2. The TiO was formed in the

sintered 10T90Z and 30T70Z, because a sufﬁcient amount of oxygen, released from ZrO2, was dissolved in the Ti. For the

sintered 50T50Z, 70T30Z, and 90T10Z, a-Ti(Zr, O) and Ti2ZrO

were formed because smaller amounts of oxygen were supplied by ZrO2. Both t- and m-ZrO2x were found in the sintered

10T90Z, 30T70Z, and 50T50Z, while c-ZrO2xwas stabilized by

a high concentration of the retained yttria and oxygen vacancies in the sintered 70T30Z. In the sintered 90T10Z, ZrO2particles

were almost dissolved in a-Ti(Zr, O), which resulted in the pre-cipitation of Y2Ti2O7.

IV. Conclusions

(1) Powder mixtures of Ti and 3 mol% Y2O3-PSZ in

var-ious ratios were sintered at 15001C for 1 h in an argon atmo-sphere. The microstructures and reaction products of the Ti/ ZrO2composites depended on the Ti/ZrO2ratio.

(2) Ti reacted with and was mutually soluble in ZrO2,

re-sulting in the formation of a-Ti(O, Zr), Ti2ZrO, and/or TiO.

Oxygen atoms, contained in a-Ti(O, Zr), Ti2ZrO, and/or TiO,

were extracted from ZrO2, whereby oxygen-deﬁcient ZrO2was

produced.

(3) In the specimens withr30 mol% Ti, the relatively small Ti/ZrO2ratio led to the formation of TiO as oxygen could be

sufﬁciently supplied by excess ZrO2.

(4) Relatively large Ti/ZrO2ratios resulted in the formation

of a-Ti(Zr, O) as well as Ti2ZrO in the specimens with 50

mol% Ti, with no TiO being found.

(5) Both m-ZrO2xand t-ZrO2xwere found in specimens

withr50 mol% Ti; however, c-ZrO2xwas formed in the

spec-imens with 70 mol% Ti as it contained a high concentration of retained yttria and oxygen vacancies in ZrO2.

(6) In the specimen with 90 mol% Ti, ZrO2particles were

almost dissolved in Ti, being accompanied of simultaneous pre-cipitation of Y2Ti2O7.

Acknowledgments

The authors would like to express their sincere gratitude to Mr. Ming-Chung Li and Wei-Chen Wang for preparing the Ti/ZrO2composite specimens.

References 1

R. C. Garvie, R. H. J. Hannink, and R. T. Pascoe, ‘‘Ceramic Steel,’’ Nature (London), 258, 703–4 (1975).

2

R. H. J. Hannink, P. M. Kelly, and B. C. Muddle, ‘‘Transformation Tough-ening in Zirconia-Containing Ceramics,’’ J. Am. Ceram. Soc., 83 [3] 461–87 (2000).

3

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] 197–207 (1956).

4

R. Ruh, ‘‘Reaction of Zirconia and Titanium at Elevated Temperatures,’’ J. Am. Ceram. Soc., 46 [7] 301–6 (1963).

5

R. Ruh, N. M. Tallan, and H. A. Lipsitt, ‘‘Effect of Metal Additions on the Microstructure of Zirconia,’’ J. Am. Ceram. Soc., 47 [12] 632–5 (1964).

6

A. Arias, ‘‘Thermal Shock Resistance of Zirconia With 15 mol% Titanium,’’ J. Am. Ceram. Soc., 49 [6] 334–8 (1966).

7

A. Arias, ‘‘Mechanism by Which Metal Additions Improve the Thermal Shock Resistance of Zirconia,’’ J. Am. Ceram. Soc., 49 [6] 339–41 (1966).

8

C. L. Lin, D. Gan, and P. Shen, ‘‘Stabilization of Zirconia Sintered With Titanium,’’ J. Am. Ceram. Soc., 71 [8] 624–9 (1988).

9

L. D. Teng, F. M. Wang, and W. C. Li, ‘‘Thermodynamics and Microstructure of Ti–ZrO2Metal–Ceramic Functionally Graded Materials,’’ Mater. Sci. Eng. A,

293, 130–6 (2000).

10

L. D. Teng, W. C. Li, and F. M. Wang, ‘‘Effect of Ti Content on the Mar-tensitic Transformation in Zirconia for Ti-ZrO2Composites,’’ J. Alloys Compds.,

319, 228–32 (2001).

11

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).

12

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

13

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

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

14

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).

15

J. Zhu, A. Kamiya, T. Yamada, W. Shi, K. Naganuma, and K. Mukai, ‘‘Surface Tension, Wettability and Reactivity of Molten Titanium in Ti/Yttria-Stabilized Zirconia System,’’ Mater. Sci. Eng. A, A327, 117–27 (2002).

16

K. L. Lin and C. C. Lin, ‘‘Microstructural Evolution and Formation Mech-anism of the Interface Between Zirconia and Titanium Annealed at 15501C,’’ J. Am. Ceram. Soc., 89 [4] 1400–8 (2005).

17

K. L. Lin and C. C. Lin, ‘‘Temperature Dependence of the Interfacial Reac-tion Between Titanium and Zirconia Annealed Between 11001 and 15501C,’’ J. Am. Ceram. Soc., 90 [3] 893–9 (2007).

18

K. F. Lin and C. C. Lin, ‘‘Interfacial Reactions Between Zirconia and Titanium,’’ Scripta Metall., 39 [10] 1333–8 (1998).

19

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).

20

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

Zirconia Interface After Reaction at 15501C,’’ J. Am. Ceram. Soc., 88 [5] 1268–72 (2005).

21

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).

22

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

23

M. Ru¨hle, N. Claussen, and A. H. Heuer, ‘‘Microstructural Studies of Y2O3

-Containing Tetragonal ZrO2Polycrystals (Y-TZP)’’; pp. 352–70 in Advances in

Ceramics, Vol. 12, Science and Technology of Zirconia II, Edited by N. Claussen, M. Ru¨hle, and A. H. Heuer. American Ceramic Society, Columbus, OH, 1984.

24

A. H. Heuer and M. Ru¨hle, ‘‘On the Nucleation of the Martensitic Transfor-mation in Zirconia (ZrO2),’’ Acta. Metall., 33 [12] 2101–12 (1985).

25

P. C. Wahlbeck and P. W. Gilles, ‘‘Reinvestigation of the Phase Diagram for the System Titanium–Oxygen,’’ J. Am. Ceram. Soc., 49 [4] 180–3 (1966).

26

M. J. Bannister and J. M. Barnes, ‘‘Solubility of TiO2in ZrO2,’’ J. Am. Ceram.

Soc., 69 [11] C-269–71 (1986). &

Table IV. Reaction Products, Crystal Structures, and Morphology in Ti/ZrO2Composites

Specimens Reaction products Crystal structures Morphology

90T10Z a-Ti(Zr, O) Hexagonal Irregular

Ti2ZrO Orthorhombic Lamellar

Y2Ti2O7 Pyrochlore Round