北京化工大学：《材料导论》课程阅读材料（无极非金属）Zirconia_as_a_ceramic_biomaterial.pdf_P6-P9

6 C.Piconi.G.Maccauro Biomaterials 20 (1999)1-25 [35]promoting phase transition for local stress concen- a maximum transformation of 72%at 37C in a tim tration or variation of the yttrium/zirconium ratio. ranging from 7 to 30 yr,encompa sing a THR expected It is worthwhile to remark that the strength degrada- lifetime.But it must be remarked that such results were ceramics characterized at the start ol an M-p nt o occurred in all the mate a low Weibull modulus (m rials but one,where strength remained the same after the 700 MPa measured in three-point bending tests.Bending teatment.This vanan behavious to the di erenc he samples tes was control like yttr th on. 46 10n rain tested 6.Table 5 contains a summary of the results of ageing tests reported by several authors. tests were performed in saline at 37,50,95C for 36 The strength degrada ion in wet environments of zir months,and in an autoclave at 121C for st bending tests were performed on 8 mm gauge ples Modulus Of Rupture (MOR)of Mg-PSZ samples main- Alumina was used as a control.Samples tested at 37C tained for 1000h in a boiling saline solution. On the n vitro and in vivo did not show significant dillerences monoclini 产微 nd,the e in the s The dev opment o the surfac ely 2 d 5 m Bending strength variation of TZP samples implanted with the T-M surface transformation,an increase in the pending strength of samples was but the start ths increase in the bending strength was observed after ransition after 3 vr The corresponding increase in be months,associated with%M-phase formation on the nding strength was about 10% samples xperimental reported hase was 69 mol%in samples aged at anc 391 0,41 whil 121 after 500 h only.being 80 molafter 1000 h. for 100d did not induce significant variations in the Zr-OH bonds were identified by FTIR in samples strength of TZP samples.Also,Ichikawa et al [42]did aged for 960 h in water at 121C.This suggests that the not observe variatic in TZI ohase trans on in th material tested depends on mech he Wista s ageing in line,or subcu- pro et al.[341.the Conflicting results were reported by Drummond [43. tion of Zr-OH bonds being the transition initiator 44]and by Thomson and Rawlings [45].Drummo ond Nev ertheless,no microcracks were observed by SEM estudy on ageing of TZ [49 aterial samples 1920 h in saline a the e authors to time interval 140-304 d,and no significant correlatior dict that the bending stength of the material will main- with the testing environment was lound.Also contro tain for 80 yr a value higher than 800 MPa. n air, d s e ults of ag in vater for three ye ove the the T-M ntimal com 03 atio to the different from tetragonal,probably cubic.This fact Sato and Shimada model [31.32]to be 25.2 kcal mol- makes the samples tested not representative of TZP for On this basis authors can calculate that the flexural implants strength of the material maintained for 50yr in water a omson and R lings [45]reported the M-phase as more than a ate for

[35] promoting phase transition for local stress concentration or variation of the yttrium/zirconium ratio. It is worthwhile to remark that the strength degradation rate is not the same for all TZP ceramics. As it was reported by Swab [33], in the ten materials tested in presence of water vapour at low temperature, different levels of strength degradation occurred in all the materials but one, where strength remained the same after the treatment. This variability in ageing behaviour is related to the differences in equilibria of microstructural parameters like yttria concentation and distribution, grain size, flaw population and distribution in the samples tested [36]. Table 5 contains a summary of the results of ageing tests reported by several authors. The strength degradation in wet environments of zirconia was studied from the early phases of the development of zirconia for biomedical applications [37]. Garvie et al. [38] reported a reduction up to 14% of the Modulus Of Rupture (MOR) of Mg—PSZ samples maintained for 1000 h in a boiling saline solution. On the other hand, the content of monoclinic phase in the surface of the specimens of the same material implanted in paraspinal muscles of rabbits, although rather high (32%), did not show significant variations. Bending strength variation of TZP samples implanted in the marrow cavity and in the paraspinal muscles of NZW rabbits or maintained in a saline solution at 37°C for 12 months was investigated by Kumar et al. [39]. An increase in the bending strength was observed after six months, associated with 2% M-phase formation on the surfaces of samples. Experimental data reported by Schwartz [26] and by Christel [40, 41] are in agreement with those of Kumar’s [39]. Christel [40, 41] showed that gamma sterilization or ageing in Ringer’s solution for 100 d did not induce significant variations in the strength of TZP samples. Also, Ichikawa et al. [42] did not observe variation in the bending strength of TZP samples after 12 months ageing in air, saline, or subcutaneous tissues of Wistar rats. Conflicting results were reported by Drummond [43, 44] and by Thomson and Rawlings [45]. Drummond performed an extensive study on ageing of TZP [43]. Reduction in MOR of about 20% was observed in TZP samples after ageing for 730 d in Ringer’s, saline solutions or distilled water at 37°C. Reduction takes place in the time interval 140—304 d, and no significant correlation with the testing environment was found. Also control specimens, maintained in air, showed similar behaviour. The samples tested contained 5.5—8.5 wt% Y2 O3 , slightly above the optimal composition, and contained phases different from tetragonal, probably cubic. This fact makes the samples tested not representative of TZP for implants. Thomson and Rawlings [45] reported the M-phase as reaching 10% after 18 months ageing in Ringer’s solution. They calculated that the M-phase might reach a maximum transformation of 72% at 37°C in a time ranging from 7 to 30 yr, encompassing a THR expected lifetime. But it must be remarked that such results were obtained on TZP ceramics characterized at the start of the test by an M-phase content of approximately 5%, and by a rather high defect population, indicated by a low Weibull modulus (m"6.5) and MOR below 700 MPa measured in three-point bending tests. Bending strength shows little variations during the test, showing that material strength in the samples tested was controlled by defects more than by phase transitions. Shimizu et al. [46] tested TZP samples (grain size 0.25 lm, density 6 g cm~3) in vitro and in vivo. In vitro tests were performed in saline at 37, 50, 95°C for 36 months, and in an autoclave at 121°C for 960 h. Samples were tested in vivo in subcutaneous tissue and in the tibial marrow of JW rabbits for 30 months. Three-point bending tests were performed on 8 mm gauge samples. Alumina was used as a control. Samples tested at 37°C in vitro and in vivo did not show significant differences. The development of the monoclinic phase on the surface of the samples was only observed 90 d after the beginning of the test, reaching approximately 2 and 5 mol% after 12 and 30 months, respectively (Fig. 5). In correspondence with the T—M surface transformation, an increase in the bending strength of samples was observed, but the starting value was recovered after 30 months. In samples tested at 50°C, 16 mol% of the surface underwent a T—M transition after 3 yr. The corresponding increase in bending strength was about 10%. The monoclinic phase was 69 mol% in samples aged at 95°C after 27 months, while in samples aged in an autoclave at 121°C the monoclinic phase was about 50 mol% after 500 h only, being 80 mol% after 1000 h. Zr—OH bonds were identified by FTIR in samples aged for 960 h in water at 121°C. This suggests that the phase transition in the material tested depends on mechanisms similar to the ones proposed by Sato and Shimada [31, 32] or by Yoshimura et al. [34], the formation of Zr—OH bonds being the transition initiator. Nevertheless, no microcracks were observed by SEM in identical material samples aged 1920 h in saline at 121°C. The activation energy of the transition process in the material tested was calculated to be about 21.5 kcal mol~1. This result allowed the authors to predict that the bending stength of the material will maintain for 80 yr a value higher than 800 MPa. Results of ageing tests in water for three years were recently reported [47]. The activation energy of the T—M transformation process was calculated according to the Sato and Shimada model [31, 32] to be 25.2 kcal mol~1. On this basis authors can calculate that the flexural strength of the material maintained for 50 yr in water at 37°C will be more than adequate for orthopaedic or dental implants. Recently Chevalier et al. [4] reported the results of a study on the T—M transformation kinetics 6 C. Piconi, G. Maccauro / Biomaterials 20 (1999) 1—25

Fig. 5. Tetragonal to monoclinic transformation of TZP in vivo and in saline (Reprint with permission from Shimizu K, Oka M, Kumar P et al., Time-dependent changes in the mechanical properties of zirconia ceramic. J Biomed Mat Res 1993;27:729—34.). Fig. 6. TZP ball heads ultimate compression load after clinical use, in comparison to the acceptance values of production batches (Reprint with permission from Cale´ s B, Stefani Y, Mechanichal properties and surface analysis of retrieved zirconia femoral hip joint heads after an implantation time of two to of two to three years. J Mat Sci Mater Med 1994;5:376—80.). in TZP obtained from coprecipitated powders. These results allow to predict a 25 yr ageing period at 37°C to reach 20% monoclinic content in their samples. The activation energy measured (log kJ mol~1) is of the same order of the one measured by Shimizu et al. [46]. Not only the yttria content but also the yttria distibution plays a role on T—M phase transition in TZP materials. The stabilizing oxide is introduced in zirconia during the early steps of the powder manufacturing process, i.e. coprecipitation of yttrium and zirconium salts. A different approach to the introduction of stabilizing oxide in ceramic powders consists a coating zirconia grains with yttria, thus obtaining an yttria gradient in the material. The effects of this yttria distribution on TZP hydrothermal stability was investigated by Richter et al. [29]. Samples obtained by coprecipitated and coated powders following the same preparation and sintering schedule were treated in an autoclave in the presence of water vapour at 140°C up to 120 h. The development of M-phase in the samples made out of the two materials follow a quite different evolution: in ‘coprecipitated’ samples one can observe a fast increase in M-phase content, reaching 80 vol% after 24 h. The successive evolution of the transformation is slower, the amount of M-phase reaching 90 vol% after 120 h of treatment. In ‘coated’ samples, the evolution of the M-phase appears to be progressive, reaching 60 vol% after 120 h of treatment. The thickness of the monoclinic layer after 120 h approximately 120 lm in TZP made out of coprecipitated powder, and around 5 lm when made out of coated powder. This result was achieved using precursors sintered to full density at a relatively low temperature without glassy additives, resulting in a TZP completely tetragonal with grains less than 0.5 lm in size [48, 49]. Cale´ s et al. [50] reported the first results on mechanichal behaviour of THR zirconia ball heads after clinical use. Tests were performed on four ball heads retrieved from patients after 22, 24, 27, 39 months. The results obtained are very relevant as the ball heads were subjected not only to the action of the body environment, but also to the physiological cyclic loading. Three out of the retrieved ball heads were subjected to static compression tests, while bar samples for the bending test were obtained from the ball retrieved after 22 months. The UCL of the retrieved ball heads was within the acceptance values characteristic of each production batch (Fig. 6). The experimental values obtained from the bending tests performed on test bars obtained from the 22 months implanted ball head, as well as results of tests performed on test bars machined from current production TZP ball heads aged in Ringer’s solution and in animals at 37°C for 2 yr [23] did not show significant differences, the bending strength remaining unchanged. The effects of the combination of stress and a wet environment on TZP stability were also reported [5]. TZP ball heads, 032 mm, were maintained in Ringer’s solution for 3, 6 and 12 months under static loads of 10, 20 and 30 kN fitted in Ti6Al4V tapers. An axial bore in tapers allowed Ringer’s solution to reach the top of the conical bore in the ball head. Compression tests did not show significant variations in UCL. Whatever the time in Ringer’s solution and the applied load, the average UCL is (129.5$6.5) kN, which corresponds to the average UCL measured before the test (132 kN). XRD analysis did not show variations in the contents of the monoclinic phase up to 12 months. Ageing of samples of the same material for 2 yr was performed by implanting then in muscles of rats and rabbits, and in the femur of rabbits and sheeps. Fracture toughness measured by the microindentation do not show significant variations, KIC 8 C. Piconi, G. Maccauro / Biomaterials 20 (1999) 1—25

C.Piconi.G.Maccauro/Biomaterials 20(1999)1-25 9 4.Wear and inner bearing surfaces bet virgin and retrie 4.1.Zirconia on zirconia TZP ball heads,nor in density or hardness were observed after 18 months of clinical use.Similarly no difference There is clear experimental evidence that the wear rate were obs of the couple zirconia/zirconia is too high to diase this n of 10 in the predominant tetragonal structure [52.53]was [56.571.show the disastrous amounts of wear of this observed. ceramic couple,up to 5000 times the wear of the The above results presented by different authors,con alumina/alumina one.Recently TZP/TZP wear was the of this [33]reporte e improvements degradation of Ps in wetenvironments depends on the studies)The TZPTZP couple was investi material microstructure,and can be controlled by acting into account the effects of environment,sliding speed on the materal manufacturing process and on the pre e that e.tha metho rd Iso 647 able to mai tor 61).These authors onfirmed the sults baned in wet environments for exp ected implant lifetimes.but general conclusions about the stability of TZP must be Sliding of a pair made of low thermal conductivity as this behaviour speculiar to each material and cturing techn eads to an increase in surface temperature.For ology the yns the wet environment this process may lead to cracking 3.3.Impact tests grain pullout and catastrophic abrasive wear.Neverthe less,the work recently published by Chevalier et al.[165 opens again the res earch in his nel In pin-on- in this field g rde magnitude lower than the wear rate of alumina/alumina Tateishi and Yunoki [28].Bodies growing in weigh pair.These results were not replicated using bovine serum werdeepoL22 m as lubricant 028 m lumina ball heads (on Cocr Zirconia/UHMWPE got)failed un der some 15Jimpact.The role exerted by the spigot The wear of the couple UHMWPE/zirconia was studied by many authors.The results obtained are sum 6.Data are scattered over several or Wear rates of UHMWPE against zirconia five times less than against alumina were observed in ring on disc 3.4.Fatigue resistance tests carried out in conformity to ISO 6474[59].due to h muc nner grain siz onia than alumina(8x from 5 to 10 kN at 30 Hz are reported by Tateishiet al. number of cycles,for alumina surface roughness passing from R=0.06 um to R.=0.22 um.Low residual poro ure.More interest soin zirconi e induced UHM wea ing orted by Ca L39 s tnan a Differen creases from 1s to 90kN.and shows a tendency to Surface zhness and p increase to infinity for loads less than 28 kN.It was finishing ma produce different wear rates.It was hy observed experimentally that TZP 22.2 mm ball head value below 0m202 to50 million cycles with load pothesized that the existence of a threshold changes can influence only a little the wear rate [65]

ranging (9$1) MPa m~1@2 whatever the site or time of implantation. Neither differences in the finish of outer and inner bearing surfaces between virgin and retrieved TZP ball heads, nor in density or hardness were observed after 18 months of clinical use. Similarly no differences were observed in the average bending strength of TZP bars aged for 300 d in SBF at 37°C and 60°C. In these samples the formation of monoclinic phase ((1%) in the predominant tetragonal structure [52, 53] was observed. The above results presented by different authors, con- firm the conclusions of the work by Swab [33] reported at the beginning of this section. The extent of strength degradation of TZPs in wet environments depends on the material microstructure, and can be controlled by acting on the material manufacturing process and on the precursors selected for ceramic manufacture. One can observe that there is experimental evidence that TZP ceramic is able to maintain good mechanical properties in wet environments for expected implant lifetimes, but general conclusions about the stability of TZP must be avoided, as this behaviour is peculiar to each material and of its manufacturing technology. 3.3. Impact tests Impact test constitutes a useful assay to evaluate the ability of a component to dissipate shock energy, i.e. its toughness. There is very limited information in this field: up to now the only results presented on this topic are due to Tateishi and Yunoki [28]. Bodies growing in weight were dropped from a 0.5 m height onto a ball head inserted in its spigot. 022.2 mm TZP ball heads (on Ti alloy spigot) failed under an impact of some 78 J, while 028 mm Alumina ball heads (on CoCr spigot) failed under some 15 J impact. The role exerted by the spigot material due to the differences in elastic properties of the two alloys and its influence on the results reported was not clearified. 3.4. Fatigue resistance Tests in Pseudo Extra Cellular Fluid (PECF) and in saline solution, with loads cycled from 1 to 12 kN and from 5 to 10 kN at 30 Hz are reported by Tateishi et al. [27, 28]. Tests were performed up to 10 million cycles on 022.2 mm TZP ball heads without failure. More interesting results are reported by Cale` s [54]. The number of cycles-to-rupture increase as the maximum load decreases from 15 to 90 kN, and shows a tendency to increase to infinity for loads less than 28 kN. It was observed experimentally that TZP 022.2 mm ball heads can withstand up to 50 million cycles with load cycled from 2.8 to 28 kN. 4. Wear 4.1. Zirconia on zirconia There is clear experimental evidence that the wear rate of the couple zirconia/zirconia is too high to use this ceramic couple in prosthetic joints. Early studies performed by Murakami and Ohtsuki [55], Sudanese et al. [56, 57], show the disastrous amounts of wear of this ceramic couple, up to 5000 times the wear of the alumina/alumina one. Recently TZP/TZP wear was the object of new interest, probably due to the improvements in TZP ceramics processing (reported after the previous studies). The TZP/TZP couple was investigated taking into account the effects of environment, sliding speed, and load on wear properties, using the ball on ring (pin on disk) method [58], by the ring on disk test in conformity to the standard ISO 6474 [59, 60], and on hip simulator [61]. These authors confirmed the results obtained previously. Sliding of a pair made of low thermal conductivity materials leads to an increase in surface temperature. For zirconia/zirconia pair the temperature may rise up to more than 100°C [58], enhancing the T—M phase transition in the wet environment. This process may lead to cracking, grain pullout and catastrophic abrasive wear. Nevertheless, the work recently published by Chevalier et al. [165] opens again the research in this field. In pin-on-disc tests performed using water as a lubricant, they observed zirconia/zirconia or zirconia/alumina wear rates one order of magnitude lower than the wear rate of alumina/alumina pair. These results were not replicated using bovine serum as lubricant. 4.2. Zirconia/UHMWPE The wear of the couple UHMWPE/zirconia was studied by many authors. The results obtained are summarized in Table 6. Data are scattered over several orders of magnitude. Wear rates of UHMWPE against zirconia five times less than against alumina were observed in ring on disc tests carried out in conformity to ISO 6474 [59], due to the much finer grain size of zirconia than alumina (8]) [26]. Other authors [62] found an increase of some 65—70% in UHMWPE volume loss, depending on the number of cycles, for alumina surface roughness passing from R! "0.06 lm to R! "0.22 lm. Low residual porosity in zirconia surface induced UHMWPE wear 40—50% less than alumina ceramics [63, 64]. Different finishing processes can have a big influence on wear. Surface roughness and porosity obtained from sample finishing may produce different wear rates. It was hypothesized that the existence of a threshold value below which surface roughness changes can influence only a little the wear rate [65]. C. Piconi, G. Maccauro / Biomaterials 20 (1999) 1—25 9