Strength and Reliability of Surface Treated Y-TZPDental Ceramics

Tomaz Kosmac,1 Cedomir Oblak,2 Peter Jevnikar,2 Nenad Funduk,2 Ljubo Marion2

1 Ceramics Department, Jozef Stefan Institute, Ljubljana, Slovenia

2 Department of Prosthodontics, Medical Faculty, University of Ljubljana, Ljubljana, Slovenia

Received 3 November 1999; 15 February 2000 accepted 16 February 2000

Abstract: This work was undertaken to evaluate the effects of dental grinding and sand-blasting on the biaxial flexural strength and Weibull modulus of various Y-TZP ceramicscontaining 3 mol% yttria. In addition, the susceptibility of pristine and mechanically treatedmaterials to low-temperature degradation under the conditions adopted for testing the chem-ical solubility of dental ceramics was investigated. The results revealed that surface grindingand sandblasting exhibit a counteracting effect on the strength of Y-TZP ceramics. Dentalgrinding lowered the mean strength and Weibull modulus, whereas sandblasting provided apowerful method for strengthening, but at the expense of somewhat lower reliability. Thefinest-grained material exhibited the highest strength after sintering, but it was less damagetolerant than tougher, coarse-grained materials. Upon extraction with the acetic acid solutionand the ammonia solution, a significant amount of tetragonal zirconia had transformed tomonoclinic, but extensive microcracking and attendant strength degradation had not yetoccurred. Standard grade Y-TZP ceramics are more resistant in an alkaline than in an acidicenvironment, and there was a strong grain-size dependence of the diffusion-controlled trans-formation. Since a special Y-TZP grade containing a small amount of alumina exhibited thehighest damage tolerance and superior stability in an acidic environment, this material showsconsiderable promise for dental applications.© 2000 John Wiley & Sons, Inc. J Biomed Mater Res(Appl Biomater) 53: 304–313, 2000

Keywords: tetragonal zirconia; dental grinding; sandblasting; surface strengthening; aging

INTRODUCTION

Yttria partially stabilized tetragonal zirconia (Y-TZP) hasbecome increasingly popular as an alternative high-toughnesscore material in dental restorations because of its biocompat-ibility and attractive mechanical properties. Compared toother dental ceramics, Y-TZP ceramics have superiorstrength, fracture toughness, and damage tolerance due to astress-induced transformation toughening mechanism, whichoperates in this particular class of ceramics. Below about700°C, the tetragonal grains in Y-TZP ceramics are thermo-dynamically unstable and may transform into the stable mon-oclinic form under an externally applied stress exerted bygrinding, impact or fracture.1,2 The volume dilatation associ-ated with the tetragonal-to-monoclinic (t3m) phase transfor-mation produces compressive stresses, making crack propa-gation more difficult, and intrinsic flaws and surface damageless detrimental to strength.

Zirconia dental restorations are produced by dry- or wet-shaping of ceramic green bodies, which are then sintered tohigh density. To achieve the perfect fit between the prostheticwork and the prepared tooth structure, a final adjustment bydental grinding is usually required, whereas sandblasting isused to improve the bond between the luting agent and theprosthetic material.3-6 Because Y-TZP ceramics exhibitstress-induced transformation, it is likely that the surface ofthe metastable tetragonal zirconia ceramic might be trans-formed, i.e., constrained, and also damaged, which influencesthe mechanical properties and reliability of the material.7,8 Ina previous study,9 we have shown that dental grinding usinga coarse-grit diamond burr at a high rotation speed lowers themean strength and reliability, whereas sandblasting improvesthe mean strength, at the expense of somewhat lower reli-ability. The counteracting effects of surface grinding andsandblasting on the flexural strength of two standard gradeY-TZP ceramics were discussed in terms of an effectivelength of surface flaws acting as the stress concentratorsrelative to the depth of the stress-induced surface compres-sive layer, which contributes to strengthening.

It has been reported that severe machine grinding, which isnearly equivalent to dental grinding, is less effective in ini-

Correspondence to:Tomaz Kosmac, Ph.D., Ceramics Department, Jozef StefanInstitute. Jamova 39, SI-1000 Ljubljana, Slovenia (e-mail: tomaz.kosmac@ijs.si)

© 2000 John Wiley & Sons, Inc.

304

tiating the t3m transformation in TZP materials10-12than inother zirconia toughened ceramics. According to Swain andHannink,13 extensive heat is generated during severe machin-ing such that locally developed temperatures exceed the m3t transformation temperature and the reverse m3 t transfor-mation occurs. As a result, tetragonal grains on the groundsurface are partitioned and highly distorted, but the amount oftransformed monoclinic zirconia contributing to the strength-ening is negligible. These authors also showed that the for-ward t3 m transformation can be retained upon gentle handgrinding at lower speed and grinding force, but the strengthresponse was not examined. The effect of grinding parame-ters on the strength of Y-TZP was investigated by Xu et al.,14

but the results were not related to the amount of transformedzirconia on the ground surfaces. A noticeable improvement inthe strength of Y-TZP has been observed only upon finemachining using a 25mm diamond wheel, whereas coarsegrinding resulted in strength reduction. The coarser the dia-mond wheel they used, the lower the measured strength.

Erosion studies of Y-TZP ceramics revealed a plasticresponse regime with an extensive transformation zonearound the damaged area.15-17 A biaxial compressive state isgenerated within this zone, which hinders the nucleation ofradial cracks during penetration and also reduces the effectivestress field during subsequent fracture in flexure. The depth ofimpact damage was attributed primarily to lateral cracks thatevolve during unloading15 and was found to scale with thekinetic energy of the impacting particles.17 Since lateralcracks are less detrimental to strength than radial cracks,Y-TZP ceramics show a high degree of impact-damage tol-erance.

While most current dental ceramics exhibit excellentchemical durabilityin situ, the transformability of Y-TZPceramics under isothermal conditions warrants attention.When exposed to an aqueous environment above 100°C overlong periods of time, Y-TZP ceramics start transformingspontaneously into the monoclinic structure.18 This t3mtransformation is diffusion-controlled and is accompanied byextensive microcracking, which ultimately leads to strengthdegradation.19,20 Ever since Y-TZP has been considered forbiomedical applications, its stabilityin vitro and in vivo hasbeen investigated, too. Thus, Drummond21 observed a signif-icant drop in strength of sintered Y-TZP between 140–304daysin vitro aging regardless of the aging solution used. Thestrength reduction was ascribed to the t3m transformation,but no data on the amount of transformed zirconia weregiven. In contrast, no degradation of cast Y-TZP material

containing a small amount of alumina and silica was reportedafter 12 months of implantation,22 indicating that composi-tion, microstructure, materials prehistory, and testing condi-tions play an important role in the low-temperature degrada-tion of Y-TZP ceramics. There are, however, no literaturedata on the t3m transformation of Y-TZP under the extrac-tion conditions adopted for testing the chemical solubility ofdental ceramics, as specified in ISO 6872:1996(E).

This study was designed to evaluate the effect of dentalgrinding and sandblasting on the biaxial flexural strength andreliability of Y-TZP ceramics containing 3 mol% yttria. Inaddition, the susceptibility of pristine and mechanicallytreated materials to low-temperature degradation under theconditions adopted for testing the chemical solubility of den-tal ceramics was investigated.

MATERIALS AND METHODS

The Y-TZP ceramics used in this study contained 3 mol%yttria in the solid solution. Two standard grades (TZ-3YB andTZ-3YSB) were of the same chemical composition, but theydiffered in their mean grain size, whereas the third material(TZ-3YS-E) also contained 0.25% of alumina addition tosuppress the t3m transformation during aging. Materialswere fabricated from commercially available ready-to-pressY-TZP powders, by uniaxial dry pressing and pressurelesssintering in air. Starting powders, sintering conditions, andthe main characteristics of sintered materials are listed inTable I. The relative density of sintered specimens exceeded99% of the theoretical value and they were 100% tetragonal,but they differ in the mean grain size, strength, and fracturetoughness. The difference in the mean strength among mate-rials was statistically insignificant, whereas the differences inthe mean fracture toughness were found to be significant (p ,0.05).

After firing, disc-shaped specimens (15.56 0.03 mm indiameter and 1.56 0.03 mm thick) from each standard gradematerial were divided into groups of ten, and for each groupa different surface treatment was applied to the top surface ofthe specimens. A coarse-grit (150mm , ISO-No. 806 314 110534 018) and a fine-grit (50mm) diamond burr (ISO-No. 806314 110 514 018), both produced by Dica, Dendia, Austria,mounted on a high-speed handpiece, were chosen for surfacegrinding to simulate clinical conditions. Dry grinding wasperformed without water-spray cooling, whereas, for wetgrinding, a stream of water coolant was directed near the

TABLE I. Starting Y-TZP Powders, Sintering Conditions, and Main Characteristics of Sintered Ceramics

Starting Powder GradeaSintering

ConditionsPhase

CompositionMean Grain Size

in mmFlexural Strength

in MPa (SD)KIc in MPa m1/2

(SD)

TZ-3YB 1500°C/2h 100% t 0.31 1021 (89.5) 4.45 (0.05)TZ-3YSB 1550°C/4h 100% t 0.44 914 (58.3) 4.76 (0.11)TZ-3YS-E 1550°C/4h 100% t 0.57 987 (106) 5.18 (0.12)

a Tosoh, Tokyo, Japan.

305STRENGTH AND RELIABILITY OF Y-TZP CERAMICS

grinding edge. The grinding speed was 150,000 rpm. Thegrinding load, which has been estimated to be about 100 g,23

was applied by finger pressure. For sandblasting, discs weremounted in a sample holder at a distance of 30 mm from thetip of the sandblaster unit (Duostar Z2, Bego, Bremen, Ger-many), equipped with a nozzle of 5 mm in diameter. Sampleswere sandblasted for 15 s with 110mm fused alumina parti-cles at 4 bars. Before and after each surface treatment thesamples were analyzed by XRD (D 5000, Siemens, Germa-ny), using CuKa radiation. The relative amount of trans-formed monoclinic zirconia on the specimens’ surfaces wasdetermined from the integral intensities of the monoclinicM(111) and M(111), and the tetragonal T(111) peaks accord-ing to the method of Garvie and Nicholson.24

A universal testing machine (Model 4301, Instron Corp.,Canton, MA) was used for biaxial flexural strength determi-nation. Tests were performed according to ISO 6872 at aloading rate of 1 mm/min. Surface-treated specimens werefractured with the surface-treated side under tension. Theload-to-failure was recorded for each disc, and the flexuralstrength was calculated using the equations of Wachtman etal.25 The statistical significance of differences among datawas analyzed with ANOVA, followed by Tuckey’s HSD testat a significance level of 0.05. The variability of the flexuralstrength values was analyzed using the two-parameterWeibull distribution function. Fracture toughness was mea-sured by the indentation technique26 using a load of 200 N.

The stability of pristine and mechanically treated materialsin an aqueous environment was tested under the conditionsspecified in ISO 6872:1996(E) for testing the chemical solu-bility of dental ceramics. After extraction with the 4% aceticacid solution by refluxing for 16 h, specimens were analyzedby XRD for phase composition and tested for strength. Thesame protocol and experimental setup were used for testingthe stability of Y-TZP ceramics in an alkaline environment.In this case, samples were extracted with an ammonia solu-tion at pH 9.5.

RESULTS

During surface grinding, material was sequentially removedas the burr was randomly moved back and forth across thesurface; the process was accompanied by extensive sparking.On average, about 130mm of material was removed duringgrinding using a coarse-grit diamond burr, and about 70mmduring fine grinding. During sandblasting about 60mm ofmaterial was uniformly removed, but sparks were not ob-served during this operation. Microscopic examination of theground and sandblasted samples revealed that in both casesthe materials’ surface was in part plastically deformed, asshown in Figure 1. The depth of the intersecting grindinggrooves and parallel grit scratches, representing the mostcharacteristic feature of the ground surface morphology, var-ied with the diamond grit size. The eroded surface waswrinkled with sharp, randomly oriented scores, and surface

pits were readily observed on an otherwise plain sandblastedsurface.

The relative amounts of the monoclinic zirconia in surfacetreated samples are listed in Table II. For all materials, thehighest amount of the monoclinic phase was found aftersandblasting, regardless of whether the samples were groundprior to sandblasting or not. Almost negligible amounts of them phase were obtained after grinding, and there was nodifference between the ground groups. When grinding andsandblasting were applied in sequence, the amount of themonoclinic phase was found to be within the values obtainedby the treatment that was applied last.

The mean values of biaxial flexural strength in MPa andthe respective standard deviations are graphically representedin Figure 2. Dental grinding reduced the mean strength of allmaterials tested. However, statistically significant differences(p , 0.05) were found only between the control group and

Figure 1. SEM images showing Y-TZP surface morphology after (a)dry grinding using a 150 mm diamond burr and (b) sandblasting.

306 KOSMAC ET AL.

the ground fine-grained standard grade material for both wetand dry grinding using the 150 um diamond burr. The drop instrength after fine grinding (50mm diamond burr) of thismaterial was not significant, as was the strength degradationof the other two materials upon any grinding procedure.Sandblasting, in contrast, significantly (p , 0.05) increasedthe strength of all materials tested. The two standard gradematerials were also tested for both surface treatment proce-dures applied in a sequence. The strength values were signif-icantly lower (p , 0.05) when compared to the controlgroups, if samples were first sandblasted and then dry ground

in both materials. Dry grinding, rough or fine, followed bysandblasting also revealed higher strength than the controlgroups, whereas significant differences were found only incoarse-grained material.

Weibull statistical analysis of the biaxial flexural strengthdata yielded two characteristic parameters for each test group,the characteristic strength,so (which is the stress level at63.29% failure probability) and the Weibull modulus, m(which is the slope of the ln(ln1/1-P vs. lns plots). Theresults are graphically represented in Figure 3. For all mate-rials, the steepest plot was obtained with the control group. Ifcoarse grinding was involved in the surface treatment (wet ordry, prior to or after sandblasting), both the Weibull modulusand the characteristic strength significantly dropped. Forthose materials, which were subjected to fine grinding, thetwo characteristic Weibull parameters were between the con-trol and the coarse-ground groups. The Weibull modulus ofthe sandblasted groups was between the control and thecoarse-ground groups, too, but with a significantly highercharacteristic strength than that of the controls.

The relative amounts of the monoclinic zirconia on thesurface of the sintered Y-TZP samples after extraction withthe 4% acetic acid solution and the diluted ammonia solutionare shown in Figure 4. Although there was no measurablemass loss upon extraction, a significant amount of monocliniczirconia was detected on all extracted surfaces. The two

TABLE II. Relative Amounts of Monoclinic Zirconia (XM in %)Verified by X-ray Diffractometry

Surface Treatment

Relative Amounts of MonoclinicPhase in % (SD)

TZ-3YB TZ-3YSB TZ-3YS-E

Dry ground (150mm) 4.3 (0.6) 5.3 (0.6) 3.6 (0.4)Wet ground (150mm) 3.4 (0.7) 4.2 (0.4) n.m.Dry ground (50mm) 3.5 (0.6) 3.9 (0.7) n.m.Sandblasted 13.9 (1.2) 15.2 (2.3) 17.1 (1.8)Sandblasted1 dry ground

(150 mm)3.7 (0.9) 3.1 (1.2) n.m.

Dry ground (150mm)1 sandblasted

12.7 (1.9) 15.7 (1.3) n.m.

Figure 2. Mean biaxial flexural strength values for as-sintered and surface-treated Y-TZP ceramics.DG 50 — dry grinding with a 50 mm diamond burr, DG 150 — dry grinding with a 150 mm diamondburr, WG 150 — wet grinding with a 150 mm diamond burr, SB — sandblasting. Error bars representone SD from the mean.

307STRENGTH AND RELIABILITY OF Y-TZP CERAMICS

Figure 3. Weibull plots for the as-sintered and surface treated (a) fine grained standard grade, (b)coarse grained standard grade, and (c) corrosion resistant E-grade Y-TZP ceramics. (continued)

308 KOSMAC ET AL.

standard grade materials transformed more in the acidic en-vironment, whereas the Xm value of the E-Grade materialwas slightly higher after extraction with the ammonia solu-tion, but for each material there was no significant difference

in Xm between the two extracting media. The highest amountof transformed zirconia was observed with the coarse-grainedstandard grade material after extraction with the acetic acidsolution, which was found to be significantly higher than that

Figure 3. (continued) (c) Corrosion resistant Y-TZP E-grade ceramics.

Figure 4. Relative amounts of monoclinic zirconia (Xm in %) on the surface of as sintered Y-TZPceramics after extraction with 4% acetic acid solution and diluted ammonia solution (pH 5 9.6) byrefluxing for 16 h.

309STRENGTH AND RELIABILITY OF Y-TZP CERAMICS

obtained with the other two materials extracted with the samemedium. There was, however, no difference in the amount oftransformed zirconia between the materials after extractionwith the ammonia solution. Because coarse-grained standardgrade Y-TZP showed the highest transformability under iso-thermal conditions, this material was also tested for strength.The mean strength of as-sintered material after extractionwith the acetic acid solution and the ammonia solution was956 (117) MPa and 938 (87) MPa, respectively, which is notsignificantly higher than that of the control group. The rela-tive amount of transformed zirconia on sandblasted coarse-grained Y-TZP surfaces increased from an initial value of15.2% to 23% after extraction with acid, and to 22% afterextraction with the ammonia solution, respectively. The meanstrength of these samples dropped from an initial 1224(119)MPa to 1132(149) MPa after extraction with the acetic acidsolution and to 1176(129) MPa after extraction with theammonia solution.

DISCUSSION

The first part of the present investigation revealed that surfacegrinding and sandblasting exhibit a counteracting effect onthe strength of yttria stabilized tetragonal zirconia (Y-TZP)ceramic. Surface grinding using diamond burrs at a highrotation speed lowered the mean strength and reliability,whereas sandblasting provided a powerful method forstrengthening, though at the expense of a somewhat lowerreliability. In our previous article,9 which covered only thestandard grade materials, this finding was explained by con-sidering two competing factors influencing the strength ofsurface-treated Y-TZP ceramics: residual surface compres-sive stresses, which contribute to strengthening; and mechan-ically induced surface flaws, which cause strength degrada-tion. With reference to Table II, grinding was less effective ininitiating the t3m transformation of zirconia, in agreementwith previously reported literature data. We assume that,under the grinding conditions used in our work, the locallydeveloped temperatures exceeded the m3t transformationtemperature (about 700 °C) below which the monocliniczirconia is thermodynamically stable, as already suggested.13

Nevertheless, the amount of transformed zirconia on theground surfaces was almost negligible and so must have beenthe surface strengthening effect, regardless of the grindingconditions used and the material tested.

The strength of ground materials is thus determined by thecritical defect size to initiate failure, which can be estimatedusing the Griffith strength relation27:

df 51

w

KIC

Îccr

, (1)

wheredf is the fracture stress,w is a geometric constant (52/p1/2 for surface line cracks),KIC is the fracture toughness,andccr is the critical defect size to initiate failure. Calculated

ccr values for sintered specimens before and after grinding arelisted in Table III. However, Eq. (1) does not take intoaccount any of the residual stresses that may exist in both theas-sintered and ground material before being loaded to frac-ture. Therefore, calculatedccr values should be regarded as aneffective length of strength controlling defects, which wouldresult in an equivalent strength of the as-sintered and groundmaterial without any residual surface stresses. With referenceto Table III, the effective length of strength determining flawsin the as-sintered Y-TZP ceramics increases with the grainsize. This is a commonly reported observation, which wasrelated to the inherent flaw population generated during sin-tering of fine- and coarse-grained ceramics.28,29

Surface grinding increases the effective critical defect size,presumably by generating radial surface cracks. Fracto-graphic examination of ground specimens indeed revealedthat failure originated from radial cracks extending severaltens ofmm (up to 50mm) from the grinding grooves into thebulk of the material.9 Further evidence of grinding-inducedsurface cracking was obtained by SEM analysis of groundsamples, prepared by a standard bonded-interface technique,as described elsewhere.30 An SEM micrograph of a polishedinterface perpendicular to the ground surface, showing about20–40mm long radial cracks, which caused material chip-ping during separation of the two bonded interfaces, is rep-resented in Figure 5. Also with reference to Table III, grind-ing-induced cracks extend deeper into the fine-grained Y-TZP than into the coarser-grained Y-TZP, presumably due tohigher toughness of the latter materials. Similarly, a greaterflaw tolerance of tougher coarse-grained TZP ceramics uponindentation was observed by Readey et al.31 By using waterspray during coarse-grinding, the calculatedccr of the dryground specimens was reduced by about 30%, indicatinglower stresses during wet grinding. Further reduction inccr

was obtained by using a finer diamond burr, in agreementwith recently published results by Xu et al.14 These authorsreported on a noticeable grit-size dependence of strengthdegradation upon machining of Y-TZP, but they could notfind any evidence of grinding-induced surface cracking.

Sandblasting, in contrast, is capable of transforming alarger amount of zirconia in the surface of Y-TZP ceramicsindicating that not only stresses, but also temperatures werelower during this operation. The effective length of surfaceflaws, which are introduced by sandblasting, does not seem toexceed significantly the thickness of the compressive surface

TABLE III. Effective Mean Critical Defect Size ccr for As-Sintered and Ground Zirconia Y-TZP Ceramic

Surface Treatment

Mean Critical Defect Sizeccr in mm(SD)

TZ-3YB TZ-3YSB TZ-3YS-E

As-sintered 15.2 (3.0) 20.9 (2.7) 22.0 (4.29)Dry ground (150mm) 55.9 (15.7) 37.2 (15.6) 34.99 (11.41)Wet ground (150mm) 41.7 (17.7) 34.4 (18.8) n.m.Dry ground (50mm) 28.8 (6.2) 25.9 (6.9) n.m.

310 KOSMAC ET AL.

layer, otherwise the strength of the material would have beenreduced instead of increased. Microscopic examination ofsandblasted samples [see Fig. 1(b)] indicated that a substan-tial erosive wear has occurred during sandblasting. Sincelateral crack chipping is the most prevalent mechanism in-volved in the erosive wear of ceramics,17 lateral cracks couldbe expected in these samples. This was later confirmed bymicroscopic examination using a bonded-interface technique.An example, illustrating lateral crack chipping, is shown inFigure 6. In order to evaluate the contribution of surfacecompressive stresses and lateral cracks to the overall strength,sandblasted samples were annealed above the m3t transfor-mation temperature for 1 h and tested for strength. If it isassumed that, along with a propagating reverse transforma-tion, the residual surface compressive stresses are released,this should be reflected in a lowering of the mean strength ofthe sandblasted samples.

The annealing temperature was determined experimen-tally, by monitoring the reduction in monoclinic zirconiacontent on the surface of sandblasted samples with tempera-ture, as shown in Figure 7, for the coarse grained material.The monoclinic zirconia content starts decreasing at anneal-ing temperatures higher than 350 °C and dropped below 2%after annealing at 900 °C, where the reverse transformationwas completed. Mean biaxial flexural strength values of thesandblasted coarse-grained samples before and after anneal-ing at these two characteristic temperatures are represented inFigure 8, together with the mean strength value of the corre-sponding control group. After annealing at 900°C, the meanstrength of the sandblasted coarse-grained samples droppedfrom 1224(119) MPa to 820(110) MPa, which is consideredto be the “Griffith strength” of the sandblasted material, i.e.,the strength of damaged but unconstrained material. Thebalance thus reflects the contribution of a thin compressivelayer to the overall strength of sandblasted samples. Themean strength of annealed samples was about 10% lower than

that of the control group, but with a noticeably higher stan-dard deviation. The difference in mean strength between thecontrol and annealed group is statistically insignificant, indi-cating that sandblasting introduces surface flaws, which arenot detrimental to the strength of Y-TZP ceramics. However,under clinical conditions where the material is exposed tothermal and mechanical cycling in a chemically active aque-ous environment over long periods, these impact flaws maygrow to become stress intensifiers, facilitating fracture initi-ation at lower levels of applied stress. Because coarse-grainedceramics exhibit stronger crack resistance toughening, it isreasonable to expect this effect to be more pronounced infine-grained Y-TZP.

Because surface grinding and sandblasting exhibit a coun-teracting effect on the strength of Y-TZP ceramics, sandblast-ing of ground ceramic pre-forms is strongly recommended. Inthis way some of the larger grinding-induced cracks areremoved, while at the same time, surface compressivestresses are produced to strengthen the material. Sandblastingfollowed by grinding, in contrast, appears to be at least asdetrimental a surface treatment as grinding alone. Duringsubsequent grinding, the surface compressive layer is re-moved and grinding cracks are introduced, which causestrength degradation.

The influence of various mechanical treatments on thereliability of sintered Y-TZP ceramics can best be discussedin terms of the Weibull parameters represented in Figure 3.Coarse grinding not only reduces the mean strength, but alsolowers the reliability. It is interesting to note that the Weibullmodulus of all groups in which the coarse-grinding wasinvolved was nearly the same, ranging from 4–6. Subsequentsandblasting does not increase the Weibull modulus of thealready ground Y-TZP ceramic, but it can shift the normal-ized strength (so) toward very high values, which shouldresult in the lowest failure probability of all the groundmaterials for any externally applied bending stress level.

Figure 6. SEM micrograph of a polished interphase perpendicular tothe sandblasted surface of a fine-grained Y-TZP, showing lateralcrack chipping.

Figure 5. SEM micrograph of a polished interphase perpendicular tothe ground (150 mm grit) of a fine-grained Y-TZP. Arrows indicategrinding induced radial cracks.

311STRENGTH AND RELIABILITY OF Y-TZP CERAMICS

The reliability of ground Y-TZP ceramics can also besignificantly improved by using a finer diamond burr.Therefore, if sandblasting is not foreseen as a finishingoperation, fine-grinding is recommended to improve themean strength and reliability of coarse ground Y-TZPceramics. Sandblasting has the potential of increasing thecharacteristic strength of both as-sintered materials, whichat least compensates for a somewhat lower reliability ofthe sandblasted ceramics.

In the second part of our work it was demonstrated that,upon extraction by refluxing under the conditions adopted for

testing the chemical solubility of dental ceramics, a signifi-cant amount of tetragonal zirconia has transformed to mon-oclinic. However, the extraction time was not long enough toreveal statistically significant differences in the Xm valuesbetween materials and the extracting media, and strengthdegradation has not yet occurred. Therefore, based on theresults in Figure 4, only some general conclusions regardingthe aging properties of Y-TZP ceramics can be made. Stan-dard grade Y-TZP ceramics are more susceptible to low-temperature degradation in an acidic environment than in analkaline environment.

Figure 7. The reduction in monoclinic phase content on the surface of sandblasted coarse-grainedstandard grade Y-TZP with annealing temperature.

Figure 8. Mean biaxial flexural strength of the sandblasted coarse-grained standard grade Y-TZPsamples before and after annealing at 350°C/1 h and 900°C /1 h. Error bars represent one SD fromthe mean.

312 KOSMAC ET AL.

In the 4% acetic acid solution, the diffusion-controlledtransformation of standard grade zirconia strongly dependson the grain size. A larger grain size, which resulted in higherdamage tolerance of the mechanically treated standard gradeY-TZP ceramics, may thus turn out to be a disadvantageduring prolonged aging below 100°C in an acidic environ-ment. It can be expected that, along with propagating trans-formation during aging, the mean strength of the Y-TZPceramics initially increase. However, once the amount oftransformed monoclinic is large enough to cause surfacemicrocracking, the strength starts to drop drastically, as de-scribed in the literature.18 Pre-existing monoclinic zirconia inthe surface of the sandblasted Y-TZP hinders the propagationof the diffusion-controlled transformation during subsequentexposure to aqueous environments. In spite of this fact,low-temperature strength degradation of sandblasted materialis likely to occur sooner, because the total amount of trans-formed zirconia is higher than that in the surface of theunconstrained material after the same exposure time.

Compared to standard grade Y-TZP ceramics, the E-gradematerial seems to be more stable in acidic environments,whereas in alkaline environments this material does not seemto be superior. This was latter confirmed by prolonged iso-thermal aging of samples for 200 h at 80°C in an acidic andalkaline environment. After aging in an ammonia solution,about 16–18% of the monoclinic zirconia was observed onthe specimens’ surface, but there was no significant differ-ence among the three materials. After aging in the 4% aceticacid solution, however, the relative amount of monocliniczirconia in the coarse- grained standard grade material ex-ceeded 30%. Considerably less (21%) zirconia had trans-formed in the surface of the fine-grained material of the samecomposition and even less (14.7%) monoclinic zirconia wasobserved on the surface of the E-grade material. Since thelatter material exhibited the largest grain size, its superiorphase stability in the acetic acid solution was attributed to thepresence of a small amount of alumina, which seems to bevery effective in hindering the diffusion-controlled transfor-mation of Y-TZP in an acidic environment. Based on itssuperior resistance to degradation after aging in the aceticacid solution and the high strength and fracture toughness, theE-Grade material might be suggested for dental applications.

REFERENCES

1. Garvie RC, Hannink RH, Pascoe RT. Ceramic steel? Nature1975;258:703–704.

2. Subbarao EC. Zirconia — an overview. Adv Ceram 1981;3:1–24.

3. Koutayas SO, Kern M. All-ceramic posts and cores: The state ofthe art. Quint Int 1999;30:383–392.

4. Luthardt R, Herold V, Sandkuhl O, Reitz B, Knaak JP, Lenz E.Kronen aus Hochleistungskeramik. Dtsch Zahnarztl Z 1998;53:280–285.

5. Wohlwend A, Studer S, Scha¨rer P. The zirconium oxide abut-ment: An all-ceramic abutment for the aesthetic improvement ofimplant superstructures. Quintess Dent Technol 1997;1:63–74.

6. Meyenberg KH, Lu¨thy H, Scha¨rer P. Zirconia posts: A newall-ceramic concept for nonvital abutment teeth. J Esthet Dent1995;7:73–80.

7. Gupta TK. Strengthening by surface damage in metastabletetragonal zirconia. J Am Ceram Soc 1980;63:117–121.

8. Green DJ. A technique for introducing surface compression intozirconia ceramics. J Am Ceram Soc 1983;66:C178–C179.

9. Kosmacˇ T, Oblak C, Jevnikar P, Funduk N, Marion L. Theeffect of surface grinding and sandblasting on flexural strengthand reliability of Y-TZP zirconia ceramic. Dent Mater 1999;15:426–433.

10. Gross V, Swain MV. Mechanical properties and microstructureof sintered and isostatically pressed yttria partially stabilizedzirconia (Y-PSZ). J Aust Ceram Soc 1986;22:1–12.

11. Urabe K, Nakajima A, Ikawa H, Udagava S. Characteristicmicrostructure of CeO2-Y2O3 dopped tetragonal zirconia poly-crystals. Adv Ceram 1988;24:345–355.

12. Virkar AV, Matsumoto RLK. Ferroelastic domain switching asa toughening mechanism in tetragonal zirconia. J Am CeramSoc 1986;69:C224–C226.

13. Swain MV, Hannink RHJ. Metastability of the martensitictransformation in a 12 mol% ceria-zirconia alloy: grindingstudies. J Am Ceram Soc 1989;72:1358–1364.

14. Xu HKK, Jahanmir S, Ives LK. Effect of grinding on strength oftetragonal zirconia and zirconia-toughened alumina. Mach SciTech 1997;1:49–66.

15. Lange FF, Evans AG. Erosive damage depth in ceramics: Astudy on metastable tetragonal zirconia. J Am Ceram Soc 1979;62:62–65.

16. Chen I-W. Implications of transformation plasticity in ZrO2

containing ceramics: II. Elastic — plastic indentation. J AmCeram Soc 1986;69:189–194.

17. Ritter EJ. Particle impact damage of engineering ceramics.Ritter JE, editor. Erosion of ceramic materials. Trans Tech;1992. p 555–578.

18. Sato T, Shimada M. Transformation of yttria-doped tetragonalZrO2 polycrystals by annealing in water. J Am Ceram Soc1985;68:356–359.

19. Li JF, Watenabe R. Phase transformation in Y2O3 partiallystabilized ZrO2 polycrystals of various grain size during low-temperature aging in water. J Am Ceram Soc 1997;17:897–903.

20. Kim DJ. Influence of aging environment on low-temperaturedegradation of tetragonal zirconia alloys. J Euro Ceram Soc1997;17:897–903.

21. Drummond JL. In vitro aging of yttria-stabilized zirconia. J AmCeram Soc 1989;72:675–676.

22. Ichicawa Y, Akagawa Y, Nikai H, Tsuru H. Tissue compatibil-ity of a new zirconia ceramics in vivo. J. Prosthet Dent 1992;68:322–326.

23. Siegel SC, von Fraunhofer JA. The cutting efficiency of dentaldiamond burs: A comparative study. J Am Dent Assoc 1996;127:637–644.

24. Garvie RC, Nicholson PS. Phase analysis in zirconia systems.J Am Ceram Soc 1972;55:303–305.

25. Wachtman JB, Capps W, Mandel J. Biaxial flexure tests ofceramic substrates. J Mater 1972;7:188–194.

26. Chantikul P, Anstis GR, Lawn BR, Marshall DB. A critical eval-uation of indentation techniques for measuring fracture toughness.II. Strength method. J Am Ceram Soc 1981;64:539–543.

27. Lawn BR. Fracture of brittle solids, 2nd ed. Cambridge: Cam-bridge University Press, 1993.

28. Rice RW. Strength grain size effects in ceramics. Proc BritCeram Soc 1972;20:205–207.

29. Seidel J, Claussen N, Ro¨del J. Reliability of alumina ceramics:effect of grain size. J Euro Ceram Soc 1995;15:395–404.

30. Guiberteau F, Padture NP, Lawn BR. Effect of grain size onHertzian contact damage in alumina. J Am Ceram Soc 1994;77:1825–1831.

31. Readey MJ, Mc Callen CL, Mc Namara PD, Lawn BR. Corre-lations between flaw tolerance and reliability of ceramics. JMater Sci 1993;28:2748–2752.

313STRENGTH AND RELIABILITY OF Y-TZP CERAMICS