SINTERING AND AGEING PROPERTIES OF ...

Presented at the 33rd International Conference & Exposition on Advanced Ceramics and Composites (Hilton Daytona Beach Resort & Ocean Center, Daytona Beach, Florida, USA, 18-23 January 2009, The American Ceramic Society USA).

SINTERING AND AGEING PROPERTIES OF MANGANESE-DOPED Y-TZP CERAMICS S. Ramesh 1, S. Meenaloshini 1, C. Y. Tan 1, I. Sopyan 2 and W. D. Teng 3 1 Ceramics Technology Laboratory, Department of Mechanical Engineering, University Tenaga Nasional, Kajang, Selangor, Malaysia 2 Department of Manufacturing and Materials Engineering, Faculty of Engineering, International Islamic University Malaysia, Malaysia Material Characterization Department, 3 Ceramics Technology Group, SIRIM Berhad, Shah Alam, Selangor, Malaysia ABSTRACT The sinterability of 3 mol% yttria-tetragonal zirconia polycrystals (Y-TZP) containing small amounts of manganese oxide (MnO2) as sintering aid was investigated over the temperature range of 1250°C to 1500°C. Sintered samples were evaluated to determine bulk density, Young’s modulus, Vickers hardness and fracture toughness. In addition, the tetragonal phase stability of selected samples was evaluated by subjecting the samples to hydrothermal ageing in superheated steam at 180°C/10 bar for periods up to 525 hours. The research showed that the addition of MnO2, particularly ≥ 0.3 wt%, was effective in aiding densification and improving the matrix stiffness and hardness when compared to the undoped Y-TZP sintered at temperatures below 1350°C. On the other hand, the fracture toughness of Y-TZP was unaffected by MnO2 addition except for the 1 wt% MnO2-doped Y-TZP samples sintered above 1400°C. The hydrothermal ageing resistance of Y-TZP was significantly improved with the additions of MnO2 in the Y-TZP matrix. INTRODUCTION Yttria-tetragonal zirconia polycrystals ceramics (Y-TZP) possessed excellent physical and mechanical properties, thus making it an attractive candidate for a host of engineering applications 1,2. One of most successful applications of Y-TZP ceramics is found in orthopedics, with femoral heads for total hip replacement. It has been reported that more than half a million patients in Europe have been implanted with zirconia femoral heads since 1985 3,4. The excellent mechanical properties of Y-TZP can be attributed to a remarkable phenomenon known as transformation toughening, which increases its crack propagation resistance. In this mechanism, the stress of a propagating crack will be absorbed by the metastable tetragonal grains and causes it to transform into the monoclinic symmetry accompanied by ~4% volume expansion. As a result, compressive strain is generated at the crack tip thus making it difficult for further crack advancement5. Typically, Y-TZP exhibits strengths and toughnesses of more than 1 GPa and 6-10 MPam1/2, respectively6. However, one of the major limitations of Y-TZP ceramic is its susceptibility to ageing-induced tetragonal (t) to monoclinic (m) phase transformation when exposed in steam environment. Kobayashi et al.7 was the first to observe that Y-TZP ceramics exhibited a slow t to m phase transformation, starting at the free surface followed by the formation of microcracking and eventually accompanied by a severe reduction in mechanical properties. Many other workers have since reported the ageing phenomenon in Y-TZP and attempted to suppress this devastating effect of ageing8-11 although the underlying mechanism that governs ageing has not been unequivocally resolved12. Nevertheless, many studies were devoted towards grain boundary modification through the use of sintering additives or dopants since it is believed that during ageing monoclinic nucleation starts at grain boundary regions and propagate inwards into the grain4,13,14. In general, small amounts of dopants could promote densification and substantially control the microstructure as well as to enhance the mechanical properties of the sintered Y-TZP body15-22. For instance, studies by Kenellopoulous and Gill22 revealed that doping of CuO enhances the densification 1


of Y-TZP material through a mechanism involving a transient liquid phase due to the low melting point of CuO in ZrO2 matrix. Therefore, the inclusion of transition metal oxides such as manganese oxide (MnO2) is likely to affect hydrothermal ageing resistance if they aid sintering at relatively low temperatures (i.e. ≤ 1300ºC) and result in densification without grain growth. EXPERIMENTAL PROCEDURES The starting powder was a co-precipitated, spray-dried 3 mol% yttria-zirconia (Y-TZP) supplied by Kyoritsu, Japan. Varying amounts of high purity MnO2 (0.05, 0.1, 0.3, 0.5 and 1 wt%, BDH, UK) were mixed with the Y-TZP powders by a wet colloidal technique, using zirconia balls as the milling media and ethanol as the mixing medium. The slurry was oven dried and sieved to obtain soft, ready-to-press powder. Disc (20 mm diameter) and rectangular bar (4 × 13 × 32 mm) green samples were compacted at 0.3 MPa and isostatically pressed at 200 MPa. Consolidation of the particles by pressureless sintering was performed in air using a rapid heating furnace (ModuTemp, Australia), at various temperatures ranging from 1250ºC to 1500ºC, maintained at the soak temperature for 2 h before cooling to room temperature. The sintered samples were ground on one face by SiC papers of 120, 240, 600, 800 and 1200 grades successively, followed by polishing with 6 µm and 1 µm diamond paste to produce an optical reflective surface. The bulk density of the sintered samples was measured based on Archimedes’ principle using an electronic balance retrofitted with a density determination kit (Mettler Toledo, Switzerland). The Young’s modulus by sonic resonance was determined for rectangular samples using a commercial testing instrument (GrindoSonic: MK5 “Industrial”, Belgium). The instrument permits determination of the resonant frequency of a sample by monitoring and evaluating the vibrational harmonics of the sample by a transducer; the vibrations are physically induced in the sample by tapping. The modulus of elasticity or Young’s modulus is calculated using the resonant frequency of the sample23. Fracture toughness (KIc) and Vickers hardness measurements (Future Tech., Japan) were made on polished samples using the Vickers indentation method. The indentation load was kept constant at 98.1 N and a loading time of 10 s was employed. The value of KIc was computed using the equation derived by Niihara et al.24. Average values were taken from at least five measurements. Phase analysis by X-ray diffraction (XRD: Geiger-Flex, Rigaku Japan) of the powders and solid samples were carried out under ambient conditions using Cu-Kα as the radiation source. The fraction of the monoclinic phase was determined using the {111} peaks in accordance to the relationship developed by Toraya et al.25. The hydrothermal ageing experiment was performed in an autoclave containing superheated steam (180°C/10 bar). The extent of surface monoclinic development was evaluated by XRD analysis. RESULTS AND DISCUSSION The XRD results of the as-received Y-TZP powder indicated the presence of ~20% monoclinic phase content. Similar phase content was also found in all the doped powders, thus indicating that the dopant had negligible effect on phase stability of the zirconia powder. Upon sintering up to 1500ºC, all the samples, with the exception of the 1 wt% MnO2-doped material, exhibited a fully tetragonal phase. The tetragonal grains of the 1 wt% MnO2-doped Y-TZP started to become unstable when sintered at 1500°C, with ~3% monoclinic phase being detected by XRD in the sintered body. The bulk density variation with sintering temperature for undoped and MnO2-doped Y-TZPs is shown in Figure 1. The beneficial effect of MnO2 in enhancing the density has been revealed. Density of Y-TZPs sintered < 1400°C was significantly improved by the addition of up to 0.5 wt% MnO2. In particular, Figure 1 shows that Y-TZP samples containing ≥ 0.3 wt% MnO2 exhibited ~98% theoretical density (the theoretical density of Y-TZP is taken as 6.1 Mgm-3) if compared to the undoped ceramic (~91% theoretical density), when sintered at 1250°C. The bulk density of the undoped (0), 0.05 wt% and 0.1 wt% MnO2-doped exhibited a similar trend with increasing sintering temperature i.e. the bulk 2


density increases to a maximum at 1350°C for both MnO2-doped samples and at 1450°C for the undoped material before remaining almost constant with further increase in temperature as shown in Figure 1. In contrast, the addition of 1 wt% MnO2 showed an opposite trend with increasing sintering temperature above 1300ºC. The declined in the density for the 1 wt% MnO2-doped samples sintered above 1450ºC could be in part attributed to the monoclinic phase development in this sample upon sintering to room temperature.

Figure 1. Effect of sintering temperature and MnO2 addition on the relative density of Y-TZP. The variation of Young’s Modulus of sintered samples with increasing sintering temperature is shown in Figure 2. The Young’s modulus variation with sintering temperature is in good agreement with the bulk density trend as shown in Figure 1. The results showed that the Young’s Modulus of the sintered body increased to a maximum of ≥ 200 GPa with increasing bulk density up to 6 Mgm-3. In general, Y-TZP containing up to 0.5 wt% MnO2 exhibited E values ≥ 200 GPa regardless of sintering temperature. However, Y-TZP that had 1 wt% dopant addition portrayed an inconsistent modulus trend. In contrast, the undoped Y-TZP recorded the lowest Young’s modulus when sintered below 1300°C. The effect of sintering temperatures and MnO2 additions on the room temperature Vickers hardness of Y-TZPs is shown in Figure 3. The results clearly show the beneficial effect of MnO2 in enhancing the hardness of Y-TZP at low sintering temperatures. The hardness of the undoped Y-TZP was at its lowest (~9.7 GPa) when sintered at 1250ºC and soon increases rapidly to ~12.8 GPa at 1300ºC before reaching a maximum of ~13.7 GPa at 1400ºC. The hardness of the samples however decreased slightly with further sintering, down to ~13.2 GPa when sintered at 1500ºC. In comparison, the hardness of all the MnO2-doped samples was higher than the undoped ceramic when sintered at 3


1250ºC and 1300ºC as shown in Figure 3. The hardness trend of the 0.05 and 0.1 wt% MnO2-doped samples was similar, i.e. increased rapidly from ~11.3 GPa and ~12.0 GPa, respectively when sintered at 1250ºC to attained values of above 13 GPa at 1300ºC. However, for sintering beyond 1300ºC, the hardness trend of both doped materials was in agreement with that of the undoped ceramic.

Figure 2. Young’s modulus variation with sintering temperature for undoped and MnO2-doped Y-TZP. Similar observation was noted for the hardness trend of 0.3 and 0.5 wt% MnO2-doped Y-TZPs with increasing sintering temperature. Both materials exhibited very high hardness of ~13.2 GPa (0.3 wt%) and 13.6 GPa (0.5 wt%) at 1250ºC. As the sintering temperature increased to 1300ºC–1350ºC, both materials exhibited similar hardness of 13.5 GPa. However, sintering beyond 1350ºC was detrimental as the hardness of both ceramics declined, with the 0.5 wt% MnO2-doped Y-TZP being more affected as shown in Figure 3. In contrast, the hardness of the 1 wt% MnO2-doped Y-TZP did not change significantly for sintering up to 1400ºC. The hardness of the sample was observed to fluctuate between 13.3–13.4 GPa when sintered between 1250ºC–1400ºC, before decreasing rapidly with further sintering. The 1 wt% MnO2-doped Y-TZP exhibited the lowest hardness for sintering above 1400ºC as shown in Figure 3 and this can be associated with the relatively low bulk density exhibited by the samples as depicted in Figure 1 and also to the m phase development for samples sintered above 1450ºC.

4


Figure 3. Variation in Vickers hardness of Y-TZPs as a function of sintering temperature.

The variation in the fracture toughness of undoped and MnO2-doped Y-TZPs is shown in Figure 4. It has been found that the fracture toughness of the undoped and up to 0.5 wt% MnO2-doped Y-TZPs fluctuated between the range of 4.6 to 5.2 MPam1/2 and did not vary significantly with increasing temperatures. Since the transformation toughening mechanism is related closely with the transformability of the t grains6, the fracture toughness can be used as a means of indication of the state of stability of the tetragonal grains in the zirconia matrix. In general, a high fracture toughness would indicate that the t grain was in a metastable state and responded immediately to the stress field of a propagating crack, such as induced during the indentation test18. In the present work, Y-TZP containing up to 0.5 wt% MnO2 did not show any indication of enhance toughness which supports the theory that the t grains were stable. On the other hand, in the case of the 1 wt% MnO2-doped samples, the fracture toughness was found to increased from 4.8 MPam1/2 at 1400ºC to 5.3 MPam1/2 at 1450ºC and then rapidly to > 7 MPam1/2 when sintered at 1500ºC.

5


Figure 4. Effect of MnO2 on the fracture toughness of Y-TZPs. The present work shows that the tetragonal phase stability was not disrupted in Y-TZP containing up to 0.5 wt% MnO2 and for sintering of up to 1450ºC. Above this regime, spontaneous phase transformation was observed upon cooling from sintering to room temperature. Based on this observation, it is believed that the presences of high MnO2 dopant (> 0.5 wt%) could have reacted with yttria at temperatures above 1450ºC to form a Mn-rich glassy phase. As a result of the dissolution of Y2O3 in the zirconia matrix, the minimum amount of stabilizer required for stabilization of the t phase was reduced, resulting in spontaneous phase transformation to the m phase upon cooling as evident in the present work for the 1 wt% MnO2-doped samples sintered above 1450ºC. This is also in agreement with the fracture toughness results which indicated that the tetragonal grains of the 1 wt% MnO2-doped Y-TZP sintered at 1500ºC was in the metastable state and responded immediately to stresses resulting from indentation, i.e. enhanced transformation toughening effect, as evident from the high KIc value measured for this sample (> 7 MPam1/2) as compared to the 1450ºC sintered sample (~ 5.3 MPam1/2). The ageing behaviour of both undoped and MnO2 doped Y-TZPs were studied by examining the amount of surface monoclinic content developed in samples sintered at 1350ºC, after exposure at different intervals of ageing duration (up to 525 h). Figure 5 shows the amount of surface monoclinic content developed in samples with different percentages of dopant addition after exposure in superheated steam. The results show that the undoped Y-TZP exhibited the worst ageing resistance and the t grains transformed to the monoclinic symmetry within few hours of exposure. The undoped ceramic attained about 92.6% m content after ageing for 24 h. The additions of up to 0.3 wt% MnO2 was not effective in resisting hydrothermal attack and the samples degraded severely within 50-100 h of exposure. On the other hand, the 0.5 wt% and 1 wt% MnO2-doped Y-TZPs exhibited improved ageing resistance and the samples recorded about 60% and 20% monoclinic content, respectively after exposure for 525 h. This result demonstrated that the addition of 0.5-1 wt% MnO2 was beneficial in 6


delaying of ageing-induced t to m phase transformation in Y-TZP ceramics. More work is in progress to elucidate the actual role of MnO2 on the t phase stability and microstructure development of Y-TZP.

Figure 5. Effect of hydrothermal ageing in superheated steam on the monoclinic phase development in Y-TZP sintered at 1350°C.

CONCLUSION The present work shows that the addition of manganese oxide was beneficial in enhancing the densification of Y-TZP ceramics, particularly when sintered at low temperatures (