zirconia friction

valuation of the Cytocom enerated by UHMWPE/ Tat&hi,

S. Niwa

Mechanical Engineering Laboratory, AIST, Ministry of International Trade and Industry, 1-2 Nanniki, Tsukuba-shi, Ibaraki 305, Japan

Tokyo BI-TECH Co. Etd, 8-9-12 Akasaka, Minato-ku, Tokyo 107, Japan

Abstract: Cytotoxicity of wear products generated by UHMWPE/yttria partially stabilized zirconia (YPSZ) friction and UHMWPE/surface-nitrided Ti-6AI-4V friction in pseudo-extracellular fluid (PECF) at 37 “C was evaluated. Though the amount of abraded UHMWPE against YPSZ was almost the same as that against nitrided Ti-6A1-4V, the wear products generated by UHMWPE/YPSZ friction significantly inhibited cell growth while those by UHMWPEjnitrided Ti-6AILTVfriction showed no growth inhibition. Dissolved Zr and Y ions were at least 70 times less than the amount causing growth inhibition. The cytotoxicity was caused mainly by the particles iess than 0.22pm in size. Amorphous zirconium-containing particles (5-20 nm) found in the PECF, formed presumably by stress corrosion, would be a responsible factor for the cytotoxicity.

stabilized zirconia (PSZ) has attracted special interest in recent years due to its superior fracture toughness and bending strength to alumina. Garvie et al. have reported the toughening mechanism of stressinduced tetragonal/monoclinic transformation in CaO-PSZ with a mean rupture strength of 650 MPa.* The material consists of a dispersion of tetragonal Zr02 within cubic Z rise polycrystalline Y203-PSZ (YPSZ)9 an -PSZtO have been also developed as the transformation toughened ceramics. The recent data of bending strength and fracture toughness for YPSZ are 900-1200 MPa and 9- 10 MNme312, respectively. li bility of PSZs was comparable to that i.e. no significant adverse soft tissue response was found after 6 month implantation in the paraspinalis muscles of rabbits12 and after 3 tion in the same muscles in rats.t’ these results, a 22mm-size YPSZ femoral head has recently been fabricated.13 On the other hand, Buchanan et aL2 and Riistlund et aLI have developed surface-modified Ti-6Al-4V by ion

TRODUCTION Wear products from bearing surfaces of prosthetic joints lead to many undesirable biological reacpolyethylene tions. Ultra-high-molecular-weight (UHMWPE) wear particles up to l-2mm in diameter are sometimes present in the tissues surrounding loose prostheses, implicating the contribution of the particles in the occurrence of aseptic loosening.’ Wear of metallic earing surface accelerates corrosion by the removal of a protective oxide film on the surface.2 Some of the released ions were reported to inhibit hydroxyapatite formation. 3>4 This would affect the normal osteoid mineralization and remodeling, which would also result in loosening of the implant. The wear of UHMWPE and the corrosion of metal have been reduced by applying ceramicsse7 or by modifying the metal surface.’ Among the ceramic bearing surfaces, partially *To whom correspondence

should be addressed.

203 Clinical Materi&

0267-6605/93/$6.00

8

1993 Elsevier Science Publishers Ltd, England

A. Ito, T. Tuteislzi, S. Niwu, S. Tange

204

implantation. Tateishi et al. have developed nitrided Ti-6Al-4V by a Nz gas nitrization technique.15 The N2 gas nitrization technique is characterized by its reduced running costs and little limitation on sample shape. In the present study, the wear resistance of nitrided Ti-6Al-4V and YPSZ against UHMWPE was evaluated by pinon-disk wear tests, and the cytotoxicity of the wear products was investigated.

MATERIALS AND METHODS Materials Three YPSZ disks (40mm+ x 6mm) for wear tests designated as YPSZl, YPSZ2 and YPSZ3 were kindly provided independently by three ceramic companies in Japan manufacturing YPSZ components for prosthetic joints. These contained 3mol% of Yz03 as a stabilizing agent. Nitrided Ti-6Al-4V disks were prepared by heating Ti-6Al-4V ELI disks at 850 “C for 16 h in a N2 gas atmosphere. The nitrided layer was 40,um thick. The nitrogen content of the layer continuously decreased from the surface toward the inside, with TIN phase at the surface. The surface finish of these disks was typical of that achieved on actual prosthetic joints, i.e. 0.007, 0.005, 0.006 and O.OlOpmRa for the YPSZl, 2, and 3 and the nitrided Ti-6Al-4V, respectively. Methods Pin-on-disk wear tests of UHMWPE against the YPSZ and the nitrided Ti-6Al-4V were performed in pseudo-extracellular fluid (PECF)16 at 37 “C following ASTM standard F732-8217 using a reciprocating pin-on-disk wear machine. For each disk four cylindrical UHMWPE pins material, (9 mm4 x 13 mm) were presoaked in the PECF for 2 weeks to fully absorb the PECF. One of the pins was then pressed with a static pressure of 3.54 MPa against the disk in a chamber filled with the PECF (55ml). The other pins, contained in the chamber without contact with the disk, were control soak specimens for weight change correction due to fluid absorption. Distilled water was periodically added to the PECF by an automatically controlled pump to prevent the compositional change of PECF by evaporation. The stroke and speed of the motion of UHMWPE pins were 25 mm and 1 cycle/s, respectively. The PECF was

exchanged by new one at every 25 x lo4 cycles. The weight change of UHMWPE pins was measured, in most cases, at every 25 x lo4 cycles. The PECF used for the 25 x lo4 cycles’ wear was sterilized by autoclaving at 120 “C for 30min, or by irradiating ultra-violet light for 2 h. Some volume of the PECF was filtered by a 0.22 pm pore size filter and a DIAFLO% ultrafiltration membrane YMlO (> 10 000 molecular weight cutoff) to eliminate wear particles. The PECF was mixed in the volume proportion 1 : 4 with Eagle MEM supplemented with 10% foetal bovine serum. The control was a 1 : 4 mixture of PECF unused for wear tests and Eagle MEM supplemented with 10% foetal bovine serum. Mouse fibroblast cells L929 suspended in these media of 3.0ml at a concentration of 1.67 x lo4 cells/ml were inoculated in 35 mm diameter cell culture dishes. The cells were cultured in a 5% COz atmosphere at 37 “C for 4 days. In each day monolayers in the dishes were trypsinized and cell counts were performed. Growth parameters used were r.g.r.1 = N1/NI,,,,,,r r.g.r.4 = N4/iV4controi where NI : number of cells in a test dish 1 day after inoculation Nl control: number of cells in the control dish

1 day after inoculation N4 : number of cells in a test dish 4 days after inoculation N4 control

..

number of cells in the control dish 4 days after inoculation

The cytotoxicity tests were performed three times independently for each disk, using three PECFs used for distinct 25 x lo4 cycles’ friction. Chemical composition of the filtrate by the 0.22 pm filter was analyzed by ICP with a detection limit being 0.1 mg/liter. For evaluating Zr ion effects on cell growth, the L929 cells were cultured in 4: 1 mixtures of Eagle MEM, and PECFs supplemented with Zr ions (0.044 10 mg/liter) by adding zirconium oxychloride. For Y ions, yttrium chloride was directly dissolved in the Eagle MEM medium (O.Ol7.0mg/liter) because yttrium chloride was not soluble in the PECF. The cells cultured for 3 days were fixed with 2.5% glutaraldehyde and stained with 0.1% crystal violet.

Cytocompatibility

of wear particles

205

The s.f.1, representing the degree of pseudopods extension, equals 1 for a circular cell and becomes larger for a more irregularly shaped cell with pseudopods. The s.f.2, representing the degree of elongation of a cell, equals I for a circular cell and becomes larger for a more elongated cell.]* The wear particles were observed by a Model CM30 transmission electron microscope (200 kV). Each wear particle was identified by EDX, EELS and/or electron diffraction. The amount of each kind of wear particles was semiquantitatively determined by coaanting the particles. 300

200

108

400

RESULTS

Cycles (xl 0 4 ) Fig.

1. Wear

of

UHMWPE against Ti-6A1-4V.

YPSZ

and

nitrided

The microscopic images of stained cells were captured by a TV camera for low illumination intensity (min. 0.3 lux). The images were AD-converted into 512 x 512 pixels with 256 illumination intensity levels, then processed into binary (blackand-white) images by selecting an appropriate threshold intensity value so that the cells were colored black. Perimeter, adhesion area and shape factors of the cells were calculated from the binary image using a PIAS LA500 image analyzer, where the shape factors were s.f.1 =

(perimeter)2 47r(adhesion area)

and s,f.2 = n(diameter

of circumscribed circle/2)2 adhesion area

Figure 1 shows the weight of abra by YPSZ and nitrided Ti-&AI-4V. The abrasion rate of UHMWPE against YPSZ2 and YPSZ3 was nearly equivalent to that against nitrided Ti-6Al-4V, though the rate was 1~9 times higher against YPSZl. Therefore, the abrasion rate of UHMWPE against YPSZ was similar to that against nitrided Ti-6Al-4V when YPSZ was properly manufactured. The PECF used for UH~WP~~~~ inhibited initial cell adhesion for the first day as well as cell growth rate in the 2nd to 4th day, though no significant growth inhibition was observed for the PECF used for UHMWP~/n~trided Ti-6AI-4V friction (Figs 2 and 3). Cell response to the UV sterilized PECF was essentially the same as that to the autoclaved PECF, though the former was somewhat less toxic than the ialter. The filtrates of PECFs used for U~~WP~/YPS~ friction, obtained by using a 0.22 pm pore size filter, had still ? ?No

filtration Pore size 0.22 pm filtration ? ?Ultrafiltration

(a

No filtration Pore size 0.22 pm filtration Ultrafiltration

0

1.5 /

0

Nitrided Ti-6AI-4V

Fig. 2. Relative

growth

PSZl

rate at 1st (a) and 4th day (b) for autoclave-sterilized PECF nitrided Ti-6A1-4V friction.

PSZ2

PSZ3

used for UHMWPE/YPSZ

Nitrided Ti-6AI-4V

and LJHMWPEj

206

A. Ito, T. Tuteishi,

? ?No filtration

S. Niwa, S. Tunge

(a

0

Pore size 0.22 firn filtration ? ?Ultrafiltration

T

T

T

-

PSZI

Fig. 3. Relative

PSZ2

growth

PSz3

--r

Nitrided Ti-6AI-4V

PSZl

Ti-6AI-4V

rate at 1st (a) and 4th day (b) for UV-sterilized PECF used for UHMWPEjYPSZ Ti-6Al-4V friction.

considerable cytotoxicity. Initial cell adhesion was only slightly improved as indicated by r.g.r. 1. The number of cells at the 4th day was still 12-76% of that of the control. The ultrafiltration of the PECFs gave comparable r.g.r.1 and r.g.r.4 values with the control. The chemical analysis of the PECF filtered by 0.22pm pore size filter demonstrated no higher contents of Zr and Y ions than 0.1 mg/liter. The r.g.r.4 values for the zirconium and yttrium solutions made from zirconium oxychloride and yttrium chloride were all larger than 0.85 in the range of Zr 0.04 to lO.Omg/liter and Y 0.01 to 7.0 mg/liter, respectively. Table 1 shows the results of image analysis of cell morphology. The cells cultured with the PECF for UHMWPEjYPSZ friction had less s.f.2 than the control cells. On the other hand, the cells cultured with the PECF for UHMWPE/nitrided Ti-6Al4V friction had less perimeter, adhesion area and s.f.1 than the control. UHMWPE wear particles by the friction against YPSZ were columnar or filmy in shape and 1440 pm in size. YPSZ wear particles were semiangular in shape and 6-20 ,um in size. On the other hand, UHMWPE wear particles against nitrided Ti-6Al4V were semispherical in shape and 0.6-2pm in Table 1. Morphological parameters PECF used for UHMWPEjYPSZl Ti-6A1-4V friction con tvol

of L929 cells cultured with and UHMWPE/nitrided

PSZl

Nitvided Ti-6Al-4

Number of cells Perimeter

503 0.20

554 0.18

538 0.13

(mm) Area (pm2) s.f.1 s.f.2

1000 3.70 5.35

1000 3.17 3.61

600 2.63 4.19

V

and UHMWPE/nitrided

size, smaller than the UHMWPE particles against YPSZ. Nitrided Ti-6Al-4V wear particles were 0.03-7 pm in size. Also, 0.4-2p.m aggregates of fine particles (5-20nm) as shown in Fig. 3 were observed. These aggregates, though Zr was detected by EDX, did not have the electron diffraction pattern of YPSZ. The diffraction pattern showed no crystalline phase but a broad halo of 4.1 A corresponding to (110) diffraction of UHMWPE. The number of each kind of wear particles was estimated as UHMWPE (against YPSZ) > YPSZ = UHMWPE (against nitrided TidA14V) >>nitrided Ti-6A14V The estimation was reliable especially for the particles larger than 5 pm in size because the particles were larger than the pores (2-5,~~rn) of micro-grid mesh.

DISCUSSION The tissues surrounding lose prosthetic joints in humans have demonstrated the presence of wear particles from bearing surfaces.’ Because the majority of wear particles found had been UHMWPE, only little attention is focused on the wear of ceramic components in vitro and in vivo.The abrasion resistance of MgO-PSZ against grinding media in water has appeared to be one order of magnitude higher than alumina. l9 On the other hand, Toni et al.*O reported a 4800 times higher wear rate of YPSZjYPSZ friction than alumina/alumina friction in Ringer’s solution at 37 “C. In the present study, TEM observation demonstrated much more

Cytocompatibilit_y

of wear particles

207

KeV Fig. 4. Aggregates of amorphous and Zr-containing particles found in the wear particles for UHMWFEjYPSZ fricrion (left) with the election diffraction pattern and the EDX spectrum (right). Coexistence of UHMWPE was shown by the electron diffraction halo corresponding to ( I 10) diffraction of UHMWPE.

YPSZ wear particles than nitrided Ti-6Al-4V wear part&&s, though the weight loss of UHMWPE against YPSZ was similar to that against nitrided Ti-6Al-4V. Therefore, nitrided Ti-6A1-4V is more wear resistant than YPSZ in PECF at 37 “C. This could be due to higher Mohs and Knoop hardness of TIN than that of Zr02.21T22 The PECF used in UHMWPE/nitrided Ti-6Al4V friction showe no growth inhibition though the cell shape was slightly changed. On the contrary the PECF used in UHMWPEjYPSZ friction significantly inhibited cell growth. Concentration of dissolved Zr and Y ions during the wear test (< 0.1 mg/liter) were at least 70 times less than that causing cell growth inhibition. Moreover, the ultrafiltration regained cytocompatibility of the PECF. The cytotoxicity was, therefore, caused not by dissolved ions but by fine particles, especially those smaller than 0.22 pm in size. A similar cell response in the two distinct sterilizing methods demonstrated that the particles were originally formed during UH~WPE/Y~SZ friction, not by a chemical reaction during autoclaving. The 5-20 nm particles, smaller than most viruses, shown in Fig. 4 are so-called ‘ultrafine particles’ with their physical and chemical properties being quite different from those of bulk. Some of the unique properties of ultrafine particles, caused by their high surface energy, are high reactivity, adsorption property and catalytic activity. Therefore, it is not surprising if these particles react undesirably to biological macromolecules or membranes in the cells or the culture medium as was reported for amorphous

precipitated silica.23 It was unclear by the TE,M observation whether UHMWPE particles smaller than 0.22,~~rnin size were present in the PECF or not due to the limit in identification by EEkS and electron diffraction. However, such UII particles, if present, may have little effect on cell growth in the present experiments, because UHMWPE floats on the surface of culture medium (density = 0.92 - 0.95) and may have little chance to interact with the cells. Chemical reaction of water to YPSZ induces tetragonal-to-monoclinic phase transformation, resulting in degradation of YPSZ in mechanical properties.24-26 Recently, the existence of amorphous zirconium hydroxide and/or yttrium hydroxide formed by the reaction has been confirmed in grain boundaries and on grain surface in the degraded YPSZ.27-29 The reaction occurs typically in the range 100-200 “C. However, even at body temperature, the reaction slowly proceeds, i.e. as the consequence of the reaction for 5-10 months, 20% of the tetragonal phase in the surface layer transforms to monoclinic phase.30 Zirconium hydroxide is precipitated by aqueous reactions usually as gelatinous aggregate with each particle being much less than 0.1 pm in size. many glasses and ceramics, mechanical stimulation, especially tensile strength, accelerates the formation of hydroxide.3*132 Therefore, the amorphous and Zr-containing particles as shown in Fig. 4 are most likely such amorphous zirconium hydroxide formed during the friction in alkaline PECF.

A. Ito, T. Tateishi, S.

208

The present cytotoxicity is not specific to YPSZ. Wear particles less than 0.22 ,um in size generated by UHMWPE/alumina friction under the same exalso inhibit L929 cell perimental conditions growth.33 Sasada et a1.34 carried out wear experiments of SUS 3 16/SUS 316 and Ti-GAI-4V/Ti6Al-4V in air and in cell culture medium, reporting that (I) the wear products generated in the medium had cytotoxicity while those generated in air did not, (2) filtration by a 0.22p.m pore size filter only slightly improved the cytocompatibility, and (3) the chemical composition of released ions were different from those of the metals considerably lacking in the amount of Fe and Ti ions. These results are essentially the same as those in the present study, and strongly suggest the presence of amorphous iron and titanium hydroxide which are precipitated by an aqueous reaction as ultrafine particles under neutral and alkali conditions.

CONCLUSION The cytotoxicity of wear particles by the friction of UHMWPE against YPSZ and nitrided Ti-6Al-4V was evaluated. Though the weight of abraded UHMWPE against YPSZ was similar to that against nitrided Ti-6Al-4V, the wear particles by UHMWPEjYPSZ friction significantly inhibited cell growth while those by UHMWPE/nitrided Ti-6Al-4V friction showed no growth inhibition. Filtration of the PECF used for UHMWPE/YPSZ friction by a 0.22 ,um pore size filter little improved the cytocompatibility of the PECF. Concentration of dissolved Zr and Y ions during the wear test (< 0.1 mg/liter) were at least 70 times less than that causing cell growth inhibition. Moreover, the ultrafiltration regained cytocompatibility of the PECF. The cytotoxicity was, therefore, caused mainly by fine particles less than 0.221_~rnin size. Fine amorphous zirconium-containing particles found in the PECF would be a responsible factor for the cytotoxicity.

ACKNOWLEDGEMENTS We are grateful to Dr M. Shiraishi and Mrs K. Ito in National Institute for Pollution and Resources for use of TEM, and to Dr M. Akao in Tokyo Medical and Dental University for helpful discussions.

Niwa, S. Tange

REFERENCES 1. Bell, R. S., Schatzker, J., Fornasier, V. L. & Goodman, S. B., A study of implant failure in the Wagner resurfacing anthroplasty. J. Bone Jt Surg., 67A (1985) 1165-75. 2. Buchanan, R. A., Rigney, Jr, E. D. & Williams, J. M., Ion implantation of surgical Ti-6Al-4V for improved resistance to wear-accelerated corrosion. J. Biomed. Mater. Res., 21 (1987) 355-66. 3. Blumenthal, N. C. & Posner, A. S., In vitro model of aluminum-induced osteomalacia: inhibition of hydroxyapatite formation and growth. Calcif. Tiss. Znt., 36 (1984) 439-41. N. C. & Cosma, V., Inhibition of apatite 4. Blumenthal, formation by titanium and vanadium ions. J. Biomed. Mater. Res., 23, Al (1989) 13-22. 5. Kawauchi, K., Kuroki, Y., Saito, S., Ohgiya, H., Sato, S., Kondo, S., Hirose, I., Obara, S., Aso, H., Yamano, K., Sasada, S. & Norose, S., Total hip endoprostheses with ceramics head and HDPE socket, clinical wear rate. In Orthopaedic Ceramic Implants, Vol. 4, Proc. Jap. Sot. Brthop. Ceramic Implants, eds H. Oonishi & Y. Ooi, 1984, pp. 253-7. 6. Semlitsch, M., Lehmann, M. & Weber, H., New prospects for a prolonged functional life-span of artificial hip joints by using the material combination polyethylene/aluminium oxide ceramic/metal. J. Biomed. Mater. Res., 11 (1977) 537-52. I. Tateishi, T., Terui, A. & Yunoki, H., Friction and wear properties of biomaterials for artificial joint. In Bioceramics Vol. 2, Proc. 2nd Zmter.Sym. on Ceramics in Medicine, ed. Heimke, G., German Ceram Sot., Cologne, 1990, pp. 145-51. 8. Garvie, R. C., Hannink, R. H. & Pascoe, R. T., Ceramic Steel? Nature, 258 (1975) 703-4. 9. Gupta, T. K., Bechtold, J. H., Kuznicki, R. C., Cadoff, L. H. & Rossing, B. R., Stabilization of tetragonal phase in polycrystalline zirconia. J. Mater. Sci., 12 (1977) 242I6. 10. Hannink, R. H. J. & Garvie, R. C., Sub-eutectoid aged Mg-PSZ alloy with enhanced thermal up-shock resistance. J. Mater. Sci., 17 (1982) 2637-43. 11. Christel, P., Meunier, A., Heller, M., Torre, J. P. & Peille, C. N., Mechanical properties and short-term in-vivo evaluation of yttrium-oxide-partially-stabilized zirconia. J. Biomed. Mater. Res., 23 (1989) 45-61. 12. Garvie, R. C., Urbani, C., Kennedy, D. R. & McNeuer, J. C., Biocompatibility of magnesia-partially stabilized zirconia (Mg-PSZ) ceramics J. Mater. Sci., 19 (1984) 3224-8. 13. Tateishi, T. & Yunoki, H., Research and development of advanced biocomposite materials and application to the artificial hip joint. Bull. Mech. Eng. Lab. Jap., 45 (1987) l-9. 14. Rastlund, T., Thomsen, P.: Bjursten, L. M. & Ericson, L. E., Difference in tissue response to nitrogen-ionimplanted titanium and c.p. titanium in the abdominal wall of the rat. J. Biomed. Mater. Res., 24 (1990) 847-60. 15. Ito, A., Tateishi, T., Ushida, T., Aoyagi, J. & Honma. T.. Improvement of Ti and Ti-6Al-4V corrosion resistance by nitrization. J. Jap. Sot. Biomaterials, 9 (1991) 236 -42. 16. Homsy, C. A., Hodge, ScD, R., Gordon, N., Braggs, E. E. & Estrella, M., Biochemical engineering materials screening and monitoring. J. Biomed. Mater. Res., 3 (1969) 235-45. 17. ASTM F732-82, Standard practice for reciprocating pin on flat evaluation of friction and wear properties of polymeric materials for use in total joint prosthesis. Annual Book of ASTM Standard,y,1982.

Cytocompatibility

18. Ushida, T. L&Tateishi, T., Evaluation of cell morphology on bioceramics. In Bioceramics Vol. 1, Proc. of the 1st inter. Bioceramic Sym., Kyoto, eds. H. Oonishi, H. Aoki & K. Sawai. Ishiyaku EuroAmerica Inc., Tokyo, 1989, pp. 36-41. 19. Sturhahn; H. H.. Dawihl, W. & Thamerus, G., Anwendungsmiiglichkeiten und werkstoffeigenshaften von zirkonoxid-sintererzeugnissen, 2: Verschleissverhalten und anwendungsbeispiele. Ber. Dt. Keram. Ges., 52 (1975) 84-6. 20. Toni, A., Sudanese, A., Cattaneo, G. L., Ciaroni, D., Greggi, T., Dallari, D. & Giunti, A., Alumina vs zirconium oxide comparative wear test. Trans. 3rd World Biomaterials Congress, Kyoto 1988, 338 pp. 21. Shaffer. P. T. B. & Hausner, H. H., Plenum press handbooks of high-temperature materials, MO. 1, Materials Index. Plenum Press, New York, 1964, pp. 293, 383. 22. Abe, H., Kanno, T.: Kauai, M. & Suzuki, K., Engineering ceramics. In Ceramic Science, Vol. 5, eds T. Yamaguchi & II. Yanagida. Gihoudo, Tokyo, 1984, pp. 7, 24. 23. Pig&t, G. H. & Pinto, P. J., Effects of nonfibrous minerals in the V79-4 cytotoxicity test. Environmental Health Perspectives, 91 (1983) 173-9. 24. Kobayashi, K., Kuwajima, H. & Masaki, T., Phase change

and mechanical properties of ZrO,-Y,O, solid electrolyte after aging. Solid State lonics, 3/4 (1981) 489-95. 25. Sato, T. & Shimada, M., Crystalline phase change in yttria-partially-stabilized zirconia by low-temperature annealing. J. Am. &ram. Sot., 67 (1984) C212- 13.

of wear particles

209

26. Sato, T., Ohtaki,

S. & Shimada, M., Transformation of ytttria partially stabilized zirconia by low temperature annealing in air. J. Mater. Sci.; 20 (1985) 1466-70. 27. Lange, F. F., Dunlop, G. L. & Davis, B. I., Degradation during aging of transformation-toughened ZrOz-Y,Oj materials at 250 “C. J. Am. Ceram. Sot., 49 (1986) 237740. 28. Nakajima, K., Kobayashi, K. & Murata, Y., Phase stability of Y-PSZ in aqueous solutions. In Advances in Ceramics lfol. 12: Science and Technology of Zirconia 2. eds N. Claussen, M. Ruble & A. H. Heaer. The American Ceramic Society, Inc., Columbus, Ohio, 1984, pp. 399-407. 29. Yoshimura, M., Noma, T., Kawabata, K. & Somiya; S., Role of Hz0 on the degradation process of Y-TZP. J. Mater. Sci. Letters, 6 (1987) 465-7. 30. Thomson, I. & Rawlings, R. D., Mechanical behaviour

of zirconia and zirconia-toughened alumina in a simulated body environment. Biomuterials, II (I 990) 505-8. 31. Michalske, T. A. & Frieman, S. W., A molecular interpretation of stress corrosion in silica. Nature, 11 (1982) 51 l12. 32. Sato, T. & Shimada, M., Transformation

tetragonal

Zr02

polycrystals

of yttria-doped by annealing in water. J.

Am. Ceram. Sot., 68 (1985) 356-9. 33. Ito, A., Tateishi, T., Niwa, S. & Tange, S., Cytotoxicity

of wear debris by the friction of bioceramics against UHMWPE. Trans. of 1991 Annual Meeting of Ceram. Sot. Jap., 1991, 145 pp. 34. Sasada: T., Imaizumi, T., Morita, M. & Mabuchi, K., Evaluation of wear toxicity for biomedical materials through cell culture method. Zyunkalsu, 33 (1988) 288-93.