DENTAL MATERIALS 24(2008)289-298 availableatwww.sciencedirect.com materials ° Science Direct ElseVierjournalhomepagewww.intl.elsevierhealth.com/journals/dema Systematic review article Stabilized zirconia as a structural ceramic. an overview J. Robert Kelly a,, Isabelle Denry Department of Reconstructive Sciences, Center for Biomaterials, University of Connecticut Health Center, Farmington, CT USA b Department of Restorative and Prosthetic Dentistry, The Ohio State University, Columbus, OH USA ARTICLE INFO A BSTRACT This review introduces concepts and background from the ceramics engineering literature Received 24 April 2007 regarding metastable zirconia ceramics to establish a context for understanding current and Accepted 11 May 2007 emerging zirconia-based dental ceramics e 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. Zirconia ramiCs Dental ceramics ghening Stabilized tetragonal R-curve Dispersion-toughened zirconia Partially stabilized zirconia Toughened zirconia polycrystalline Y-TZP 1. Toward improved reliability: (1)flaw control and(2)flaw tolerance 2. Zirconia polymorphs: temperature dependence and transformation strains 3. Three distinct zirconia ceramics: terminology, processing and microstructures 3.1 rtially stabilized zi 3.3. Single-phase, polycrystalline t-zro2 4. Mechanism(s) and consequences of transformation 293 5. R-curve behavior: description, definition and implications 6. Strength versus toughness Cyclic fatigue of transformation-toughened ceramics Presentend nt the annua cessio hef th cecnatde ms f ental Materias, e Etobein 2t55 06 30 1615. sa Tll fax:+18606791370 E-mail address: Kelly@nsol. uchc. edu .R Kelly) 0109-5641/$-see front matter e 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. doi:10.1016/ denta2007.05005
dental materials 24 (2008) 289–298 available at www.sciencedirect.com journal homepage: www.intl.elsevierhealth.com/journals/dema Systematic review article Stabilized zirconia as a structural ceramic: An overview J. Robert Kelly a,∗, Isabelle Denry b a Department of Reconstructive Sciences, Center for Biomaterials, University of Connecticut Health Center, Farmington, CT USA b Department of Restorative and Prosthetic Dentistry, The Ohio State University, Columbus, OH USA article info Article history: Received 24 April 2007 Accepted 11 May 2007 Keywords: Zirconia Ceramics Dental ceramics Transformation toughening Stabilized tetragonal R-curve Dispersion-toughened zirconia Partially stabilized zirconia Toughened zirconia polycrystalline Y-TZP abstract This review introduces concepts and background from the ceramics engineering literature regarding metastable zirconia ceramics to establish a context for understanding current and emerging zirconia-based dental ceramics. © 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. Contents 1. Toward improved reliability: (1) flaw control and (2) flaw tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 2. Zirconia polymorphs: temperature dependence and transformation strains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 3. Three distinct zirconia ceramics: terminology, processing and microstructures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 3.1. Dispersion-toughened ceramics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 3.2. Partially stabilized zirconia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 3.3. Single-phase, polycrystalline t-ZrO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292 4. Mechanism(s) and consequences of transformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 5. R-curve behavior: description, definition and implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 6. Strength versus toughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 7. Cyclic fatigue of transformation-toughened ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 Presented at the annual session of the Academy of Dental Materials, October 23–25, 2006, Sao Paulo, Brazil. ˜ ∗ Corresponding author at: UConn Health Center, 263 Farmington Avenue, Farmington, CT 06030-1615, USA. Tel.: +1 860 679 3747; fax: +1 860 679 1370. E-mail address: Kelly@nso1.uchc.edu (J.R. Kelly). 0109-5641/$ – see front matter © 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dental.2007.05.005
DENTAL MATERIALS 24(2008)289-298 8. Low temperature degradation of 3Y-TZP 8.1. Cubic phase and accelerated aging 2.31%; t-m approximately 4.5%. Sintered structures trans 1. Toward improved reliability: ( 1)flaw control and (2) flaw tolerance forming from t to m on cooling from sintering temperatures (approximately 1300-1500C)undergo spallation with por- tions crumbling into multi-grained powders Two research paths aimed at increasing the structural relia Beginning about 1972, the ceramic engineering commu- bility of ceramics have been pursued over the past 30 years. nity was discovering that alloying with lower valance oxides The first involved efforts to minimize the number and size such as Cao, Mgo, La203, and Y203, disfavored the strained of critical flaws based on the well-accepted Griffith flaw- m phase at room temperature and favored more symmetric ize/strength relationship for linearly elastic brittle fracture c"and t lattice structures(with'indicating metastability)[6] These cand t' phases are analogous to those in pure zirco- (1) nia but have dopant ions substituted on Zrt sites and have a fraction of oxygen sites vacant to retain charge neutrality where of is the fracture strength, c the critical flaw radius, 14]. The amount of dopant required for full cubic stabiliza- Kic critical stress intensity in mode I opening and Y is the tion is substantial; 8 mo1% in the case of the dopant Y2O3 crack shape factor. This research, part of a discipline known with one oxygen vacancy created for every two yttrium ions as ceramics processing, continues to investigate a multitude [7]. Partial stabilization of tetragonal zirconia can occur at of steps including powder fabrication (to control chemi- dopant concentrations of 2-5mo1% depending on grain size, 1 homogeneity particle size, size uniformity, etc ) particle to be discussed below. These metastable c and I phases have dispersion in processing media, powder consolidation and prolonged stability at room temperature given that cation and ho mobility in zirconia is quite low and that the oxygen vacan sities), sintering control, and"flaw kind"finish machining cies are locally ordered 17]. Recent consistent, but apparently (or finishing to net-shape avoiding the need for machining) independent work, attributes tetragonal metastability solely [1]. Second were efforts controlling ceramic microstructures to the presence of the oxygen vacancies that allow both anion to increase their resistance toward crack propagation, and cation relaxations to occur dependent on their vacancy to increase toughness. Counter to the Griffith relationship, proximity 14, 7,8. Overall, three mechanisms are discussed for increased strength and increased toughness do not gener ally correlate in transformation-toughened ceramics as will Y203 and CeO2: ()dopants inducing oxygen vacancies that are beelaborated on later. However, suchhigh toughness ceramics generally trivalent (e.g. Gd", Fe3+, Ga3+, and Y3); (i)tetrava- of lower strength are appealing for structural use due to their lent dopants being undersized or oversized with respect to the damage tolerance. Table 1 lists five major ceramic toughen ing mechanisms along with engineering material examples 2 One of these mechanisms, transformation toughening along with microcracking and deflection mechanisms are the Table 1-Ceramic toughening mechanisms [2] toughening mechanisms now prominent in zirconia-based or Mechanism Highest toughnes Example zirconia-containing ceramics and are the topic of this paper. aterials Transformation 2. Zirconia polymorphs: temperature dependence and transformation strains Microcracking Al O3/ZrO? SinG/sic Pure zirconia is monoclinic (m) at room temperature and pressure. With increasing temperature the material trans Metal dispersion AlO/Al forms to tetragonal (t), by approximately 1170 C and then Al2O3/Ni to a cubic (c) fluorite structure starting about 2370C with melting by 2716 C[3, 4]. These lattice transformations are Whiskers/platelets SigNa/Sic martensitic, characterized by(1)being diffusionless (i.e. involv SigN4/SigN Al2 O3/Sic ing only coordinated shifts in lattice positions versus transport of atoms),(2) occurring thermally implying the need for a Fibers CAS/SiC temperature change over a range rather than at a specific temperature and,(3) involving a shape deformation [S]. This Al2O3/SiC C/SiC transformation range is bounded by the martensitic start(Ms) and martensitic finish temperatures. Volume changes on cool Al2O3/Al2O ing associated with these transformations are substantial enough to make the pure material unsuitable for applica- a Calcium aluminum silicate glass ceramic. tions requiring an intact solid structure: c- t approximately Lithium aluminum silicate glass ceramic
290 dental materials 24 (2008) 289–298 8. Low temperature degradation of 3Y-TZP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 8.1. Cubic phase and accelerated aging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 1. Toward improved reliability: (1) flaw control and (2) flaw tolerance Two research paths aimed at increasing the structural reliability of ceramics have been pursued over the past 30 years. The first involved efforts to minimize the number and size of critical flaws based on the well-accepted Griffith flawsize/strength relationship for linearly elastic brittle fracture: f = KIC Y √c (1) where f is the fracture strength, c the critical flaw radius, KIC critical stress intensity in mode I opening and Y is the crack shape factor. This research, part of a discipline known as ceramics processing, continues to investigate a multitude of steps including powder fabrication (to control chemical homogeneity, particle size, size uniformity, etc.), particle dispersion in processing media, powder consolidation and packing (creating high and homogeneous greenware densities), sintering control, and “flaw kind” finish machining (or finishing to net-shape avoiding the need for machining) [1]. Second were efforts controlling ceramic microstructures to increase their resistance toward crack propagation, i.e. to increase toughness. Counter to the Griffith relationship, increased strength and increased toughness do not generally correlate in transformation-toughened ceramics as will be elaborated on later. However, such high toughness ceramics of lower strength are appealing for structural use due to their damage tolerance. Table 1 lists five major ceramic toughening mechanisms along with engineering material examples [2]. One of these mechanisms, transformation toughening along with microcracking and deflection mechanisms are the toughening mechanisms now prominent in zirconia-based or zirconia-containing ceramics and are the topic of this paper. 2. Zirconia polymorphs: temperature dependence and transformation strains Pure zirconia is monoclinic (m) at room temperature and pressure. With increasing temperature the material transforms to tetragonal (t), by approximately 1170 ◦C and then to a cubic (c) fluorite structure starting about 2370 ◦C with melting by 2716 ◦C [3,4]. These lattice transformations are martensitic, characterized by (1) being diffusionless (i.e. involving only coordinated shifts in lattice positions versus transport of atoms), (2) occurring athermally implying the need for a temperature change over a range rather than at a specific temperature and, (3) involving a shape deformation [5]. This transformation range is bounded by the martensitic start (Ms) and martensitic finish temperatures. Volume changes on cooling associated with these transformations are substantial enough to make the pure material unsuitable for applications requiring an intact solid structure: c→t approximately 2.31%; t→m approximately 4.5%. Sintered structures transforming from t to m on cooling from sintering temperatures (approximately 1300–1500 ◦C) undergo spallation with portions crumbling into multi-grained powders. Beginning about 1972, the ceramic engineering community was discovering that alloying with lower valance oxides, such as CaO, MgO, La2O3, and Y2O3, disfavored the strained m phase at room temperature and favored more symmetric c* and t* lattice structures (with * indicating metastability) [6]. These c* and t* phases are analogous to those in pure zirconia but have dopant ions substituted on Zr4+ sites and have a fraction of oxygen sites vacant to retain charge neutrality [4]. The amount of dopant required for full cubic stabilization is substantial; 8mol% in the case of the dopant Y2O3 with one oxygen vacancy created for every two yttrium ions [7]. Partial stabilization of tetragonal zirconia can occur at dopant concentrations of 2–5mol% depending on grain size, to be discussed below. These metastable c* and t* phases have prolonged stability at room temperature given that cation mobility in zirconia is quite low and that the oxygen vacancies are locally ordered [7]. Recent consistent, but apparently independent work, attributes tetragonal metastability solely to the presence of the oxygen vacancies that allow both anion and cation relaxations to occur dependent on their vacancy proximity [4,7,8]. Overall, three mechanisms are discussed for stabilization of t-ZrO2 with the most common dopants being Y2O3 and CeO2: (i) dopants inducing oxygen vacancies that are generally trivalent (e.g. Gd3+, Fe3+, Ga3+, and Y3+); (ii) tetravalent dopants being undersized or oversized with respect to the Table 1 – Ceramic toughening mechanisms [2] Mechanism Highest toughness (MPam1/2) Example materials Transformation ∼=20 ZrO2 (MgO) HfO2 Microcracking ∼=10 Al2O3/ZrO2 Si3N4/SiC SiC/TiB2 Metal dispersion ∼=25 Al2O3/Al Al2O3/Ni WC/Co Whiskers/platelets ∼=15 Si3N4/SiC Si3N4/Si3N4 Al2O3/SiC Fibers ≥30 CASa/SiC LASb/SiC Al2O3/SiC SiC/SiC SiC/C Al2O3/Al2O3 a Calcium aluminum silicate glass ceramic. b Lithium aluminum silicate glass ceramic.
DENTAL MATERIALS 24(2008)289-298 oxide cations(e.g. Ti++, Ge++, Ce4+); or, (ii) dopants resulting 3.1. Dispersion-toughened ceramics In minor, stabilization mechanism involves matrix constraint of The simplest material utilizes the dispersion of zirco- t-zrO2 grains held within non-transforming materials nia particles in another matrix and appears to be the The"absorption of energy"during the room temper- least widely published and commercially important. These ature t-m transformation in partially stabilized zirconia dispersion-toughened materials, such as Zro2-toughened alu (Cao-ZrO2; described in more detail below) was recognized mina(Al2O3)or ZrO2-toughend mullite(3Al2O3 2SiO2)have as a strengthening mechanism in 1975 [10]. In the 1975 been termed ZTA and ZTM [11]. In contrast with the other two publication reference was made to similarities between trans- classes, stability of the t phase to room temperature does not forming zirconia and austenite-martensite strengthening primarily involve the use of dopants but is controlled instead of TRIP steels(transformation-induced plasticity)[10]. TRIP by particle size, particle morphology and location (intra-or steels contain retained austenite when they have carbon con- intergranular, see Fig. 1a). In ZTA, for example, particles above centrations in excess of 1 wt % Some mechanical properties, a critical size will transform to monoclinic symmetry upon in particular toughness and ductility, rely on the diffusionless cooling to room temperature [12]. Since this t-m transfor- transformation of this austenite into high-carbon martensite mation is known to be martensitic, a useful way to describe nduced by stress and strain. Three features, in particular, particle size effects has been to examine their influence on vere seen shared with strengthened steels leading Garvie et the martensitic start(Ms) temperature; essentially all t-phase al. [10] to term these new zirconia compositions"ceramic stabilization can be viewed as decreasing the Ms to below steel":(1)three allotropes,(2)martensitic transformations, room temperature. Such investigation has suggested that the and ()metastable phases. Stabilized zirconia and steels also particle size effect is likely due to difficulties in nucleating share similarities in elastic moduli and coefficients of thermal the transformation, although considerations have also been expansion Garvie et al. further predicted that a vast range of given to the possible effects of surface and strain energy and ceramic materials with different properties would be devel- chemical free energy driving forces[12. Within dental mate oped analogously to iron systems [10] rials the sole commercial example of a dispersion-toughened ceramic is In-Ceram Zirconia (Vita Zahnfabrik) which is an interpenetrating composite of 30% glass and 70% polycrys 3. Three distinct zirconia ceramics: talline ceramic consisting of Al2O3 ZrO2 in a vol. ratio of terminology, processing and microstructures As foreseen by Garvie wide latitude was found in the applica- 3. 2. Partially stabilized zirconia tion of the zirconia t-m transformation in ceramics, leading to development of three different materials each having an These materials are the most widely studied,commer- associated terminology [10]. These three classes are listed cially important, microstructurally complex, and in the case in Table 2 along with some dental examples; the first two of Mg-doped some of the toughest of the transformation of these are at least two-phase materials with t-ZrO2 as the toughened ceramics. In these ceramics t-ZrO2 intra-granular minor phase(dispersed and precipitated respectively) and precipitates exist within a matrix of stabilized c-zrO2. Sta- the last is essentially a single-phase t-zrO2. The origin and bilization involves dopant addition, such as with Cao, Mgo, details of stabilization of the t phase differs among these three La2O3, and Y203, in concentrations lower than that required toughened microstructures; photomicrographs of representa- for full c-Zro2 stabilization. Full stabilization is purpose- tive materials are presented in Fig. 1. The three materials share fully not achieved in these materials, hence the historical that stabilization of t occurs and that toughness involves the derivation for the term "partially stabilized zirconia"or martensitic t- m transformation PSZ, to which the relevant dopant is often appended: Ca PSZ, Mg-PSZ, Y-PSZ, etc. [11]. Precipitates are fully coherent with the cubic lattice, forming on a nanometer scale with lenticular morphology(approximately 200nm diameter and Table 2-Forms of transformation-toughened zirconia 75nm thick) parallel to the three cubic axes (refer 1. Zirconia(dispersed phase)toughened ceramics; e.g, ZTA Fig. 1b)[12, 13]. Following sintering or solution annealing in (alumina), ZTM(mullite) the cubic solid solution single-phase field(approximately Dental example: >1850C), precipitates are nucleated and grown at lower In-Ceram ita Zahnfabrik) temperatures(approximately 1100C)within the two-phase 2. Partially stabilized zirconia(PSZ; e.g. Ca-PSZ, Mg-PSZ, Y-psz) tetragonal solid solution plus cubic solid solution phase fiel Lenticular (ens shaped) tetragonal precipitates in a cubic matrix a process termed"aging"[12, 13]. Aging optimization(time- Denzir-MDentror 3. Tetragonal zirconia polycrystals (TZP; e.g. Y-TZP, Ce-TZP and phase stability [14]. Metastability can be lost when tetrag Nominally 98% tetragonal, fine grain size onal precipitates are too small(they will not transform) and Dental examples. when precipitates grow too large spontaneous transformation DC Zirkon(DCS Precident, Schreuder Co can occur to m with twinning and microcracking [14] Complex decomposition and tertiary precipitation pro- Lava(3M ESPE cesses have also been reported to occur with aging of Mg-PS In-Ceram YZ(Vita Zahnfabrik) [15] along with the development of quite some range of phys
dental materials 24 (2008) 289–298 291 oxide cations (e.g. Ti4+, Ge4+, Ce4+); or, (iii) dopants resulting in charge-compensations (YNbO4, YTaO4) [9]. Another, more minor, stabilization mechanism involves matrix constraint of t-ZrO2 grains held within non-transforming materials. The “absorption of energy” during the room temperature t→m transformation in partially stabilized zirconia (CaO–ZrO2; described in more detail below) was recognized as a strengthening mechanism in 1975 [10]. In the 1975 publication reference was made to similarities between transforming zirconia and austenite–martensite strengthening of TRIP steels (transformation-induced plasticity) [10]. TRIP steels contain retained austenite when they have carbon concentrations in excess of 1 wt.%. Some mechanical properties, in particular toughness and ductility, rely on the diffusionless transformation of this austenite into high-carbon martensite induced by stress and strain. Three features, in particular, were seen shared with strengthened steels leading Garvie et al. [10] to term these new zirconia compositions “ceramic steel”: (1) three allotropes, (2) martensitic transformations, and (3) metastable phases. Stabilized zirconia and steels also share similarities in elastic moduli and coefficients of thermal expansion. Garvie et al. further predicted that a vast range of ceramic materials with different properties would be developed analogously to iron systems [10]. 3. Three distinct zirconia ceramics: terminology, processing and microstructures As foreseen by Garvie wide latitude was found in the application of the zirconia t→m transformation in ceramics, leading to development of three different materials each having an associated terminology [10]. These three classes are listed in Table 2 along with some dental examples; the first two of these are at least two-phase materials with t-ZrO2 as the minor phase (dispersed and precipitated respectively) and the last is essentially a single-phase t-ZrO2. The origin and details of stabilization of the t phase differs among these three toughened microstructures; photomicrographs of representative materials are presented in Fig. 1. The three materials share that stabilization of t occurs and that toughness involves the martensitic t→m transformation. Table 2 – Forms of transformation-toughened zirconia 1. Zirconia (dispersed phase) toughened ceramics; e.g., ZTA (alumina), ZTM (mullite) • Dental example: In-Ceram zirconia (Vita Zahnfabrik) 2. Partially stabilized zirconia (PSZ; e.g. Ca-PSZ, Mg-PSZ, Y-PSZ) • Lenticular (lens shaped) tetragonal precipitates in a cubic matrix • Dental example: Denzir-M (Dentronic AB) 3. Tetragonal zirconia polycrystals (TZP; e.g. Y-TZP, Ce-TZP • Nominally 98% tetragonal, fine grain size • Dental examples: DC Zirkon (DCS Precident, Schreuder & Co) Cercon (Dentsply Prosthetics) Lava (3M ESPE) In-Ceram YZ (Vita Zahnfabrik) 3.1. Dispersion-toughened ceramics The simplest material utilizes the dispersion of zirconia particles in another matrix and appears to be the least widely published and commercially important. These dispersion-toughened materials, such as ZrO2-toughened alumina (Al2O3) or ZrO2-toughend mullite (3Al2O3·2SiO2) have been termed ZTA and ZTM [11]. In contrast with the other two classes, stability of the t* phase to room temperature does not primarily involve the use of dopants but is controlled instead by particle size, particle morphology and location (intra- or intergranular; see Fig. 1a). In ZTA, for example, particles above a critical size will transform to monoclinic symmetry upon cooling to room temperature [12]. Since this t→m transformation is known to be martensitic, a useful way to describe particle size effects has been to examine their influence on the martensitic start (Ms) temperature; essentially all t-phase stabilization can be viewed as decreasing the Ms to below room temperature. Such investigation has suggested that the particle size effect is likely due to difficulties in nucleating the transformation, although considerations have also been given to the possible effects of surface and strain energy and chemical free energy driving forces [12]. Within dental materials the sole commercial example of a dispersion-toughened ceramic is In-Ceram Zirconia (Vita Zahnfabrik) which is an interpenetrating composite of 30% glass and 70% polycrystalline ceramic consisting of Al2O3:ZrO2 in a vol.% ratio of approximately 70:30. 3.2. Partially stabilized zirconia These materials are the most widely studied, commercially important, microstructurally complex, and in the case of Mg-doped some of the toughest of the transformationtoughened ceramics. In these ceramics t-ZrO2 intra-granular precipitates exist within a matrix of stabilized c-ZrO2. Stabilization involves dopant addition, such as with CaO, MgO, La2O3, and Y2O3, in concentrations lower than that required for full c-ZrO2 stabilization. Full stabilization is purposefully not achieved in these materials, hence the historical derivation for the term “partially stabilized zirconia” or PSZ, to which the relevant dopant is often appended: CaPSZ, Mg-PSZ, Y-PSZ, etc. [11]. Precipitates are fully coherent with the cubic lattice, forming on a nanometer scale with lenticular morphology (approximately 200 nm diameter and 75 nm thick) parallel to the three cubic axes (refer to Fig. 1b) [12,13]. Following sintering or solution annealing in the cubic solid solution single-phase field (approximately >1850 ◦C), precipitates are nucleated and grown at lower temperatures (approximately 1100 ◦C) within the two-phase tetragonal solid solution plus cubic solid solution phase field; a process termed “aging” [12,13]. Aging optimization (timetemperature-transformation) involves both precipitate size and phase stability [14]. Metastability can be lost when tetragonal precipitates are too small (they will not transform) and when precipitates grow too large spontaneous transformation can occur to m with twinning and microcracking [14]. Complex decomposition and tertiary precipitation processes have also been reported to occur with aging of Mg-PSZ [15] along with the development of quite some range of phys-
292 DENTAL MATERIALS 24(2008)289-298 dispersed t-Zro2 phase lenticular t-zro, precipitates on cubic faces a. Ziconia-toughenend alumina(ZTA b Mg partially-stabilized zirconia(Mg-PSz) Yttria-stabilized tetragonal zirconia polycrystalline(y-tZP Fig. 1- Microstructural features of the three major categories of transformation-toughened zirconia Compiled with permission from Heuer[11] and Matsui et al.22 ical properties, for example, KIc ranging from 3 to 16MPa m12 8 phase is thought to directly explain the improved mechan with lower values occurring in both un-aged and over-aged ical properties of Mg-PSZ during aging[18]. One commercial ceramic [16]. Specific insights into bulk phase developments Mg-PSZ appears to be available as a dental ceramic( Denzir-M and microstructural control over toughening in Mg-PSZ came Dentronic AB, Sweden) and has received attention during with the advent and application of neutron diffraction; X-ray vitro testing of fixed partial dentures [191 penetration in these materials is extremely limited compared to centimeter depths possible with neutrons. Aging in Mg-PSZ 3.3. Single-phase, polycrystalline t-Zro2 involves the development of a complex microstructure form- ing from essentially a two-phase c+t starting system. These In 1977 it was reported that fine grain ZrO2 (gener changes include (1)tertiary t-phase precipitation,(2)some ally <0.5um) with small concentrations of stabilizing Y203 C-m transformation,(3)limited orthorhombic (o) phase for- could contain up to 98% of the metastable t phase fol- mation, and most critically,(4)growth of an anion-ordered lowing sintering [20]. High strengths coincided with high acancy phase termed delta( a )having the composition tetragonal phase content and low strengths coincided with Mg2ZrsO12[17, 18]. This 8 compound nucleates on the broad high monoclinic phase content [20). Subsequent investigation lenticular tetragonal-cubic phase boundary and grows with revealed that the highest strengths(700 MPa)and toughness onsumption of c-ZrO2. Although the t-8-phase boundary is KIc 6-9 MPa m1)were only found below a critical average coherent, some lattice parameter mismatch exists leaving the grain size(<0.3 um)[21]. This critical grain-size phenomenon t phase increasingly susceptible to transformation as the 8 indicated a strength/toughness mechanism beyond the sim- layer thickens [18].Growth of this Mg-rich 8 phase also appears ple flaw-related grain-size effects generally recognized for to occur with Mg depletion of the t precipitates. It has been polycrystalline ceramics. However, many attributes are still calculated roughly that the stress required for the t-m trans- shared with other polycrystalline materials, including the formation decreases from 470 to 70 MPa with aging, in a linear simplicity of processing; norequirement for"aging"heattreat fashion with &-phase formation [18].Thus, the precipitation of ments, and reciprocally, relative insensitivity to follow-on heat
292 dental materials 24 (2008) 289–298 Fig. 1 – Microstructural features of the three major categories of transformation-toughened zirconia. Compiled with permission from Heuer [11] and Matsui et al. [22]. ical properties, for example, KIC ranging from 3 to 16 MPam1/2 with lower values occurring in both un-aged and over-aged ceramic [16]. Specific insights into bulk phase developments and microstructural control over toughening in Mg-PSZ came with the advent and application of neutron diffraction; X-ray penetration in these materials is extremely limited compared to centimeter depths possible with neutrons. Aging in Mg-PSZ involves the development of a complex microstructure forming from essentially a two-phase c + t starting system. These changes include (1) tertiary t-phase precipitation, (2) some c→m transformation, (3) limited orthorhombic (o) phase formation, and most critically, (4) growth of an anion-ordered vacancy phase termed delta (ı) having the composition Mg2Zr5O12 [17,18]. This ı compound nucleates on the broad lenticular tetragonal-cubic phase boundary and grows with consumption of c-ZrO2. Although the t-ı-phase boundary is coherent, some lattice parameter mismatch exists leaving the t phase increasingly susceptible to transformation as the ı layer thickens [18]. Growth of this Mg-rich ı phase also appears to occur with Mg depletion of the t precipitates. It has been calculated roughly that the stress required for the t→m transformation decreases from 470 to 70 MPa with aging, in a linear fashion with ı-phase formation [18]. Thus, the precipitation of ı phase is thought to directly explain the improved mechanical properties of Mg-PSZ during aging [18]. One commercial Mg-PSZ appears to be available as a dental ceramic (Denzir-M, Dentronic AB, Sweden) and has received attention during in vitro testing of fixed partial dentures [19]. 3.3. Single-phase, polycrystalline t-ZrO2 In 1977 it was reported that fine grain ZrO2 (generally < 0.5m) with small concentrations of stabilizing Y2O3 could contain up to 98% of the metastable t phase following sintering [20]. High strengths coincided with high tetragonal phase content and low strengths coincided with high monoclinic phase content [20]. Subsequent investigation revealed that the highest strengths (∼=700 MPa) and toughness (KIC ∼= 6–9 MPam1/2) were only found below a critical average grain size (<0.3m) [21]. This critical grain-size phenomenon indicated a strength/toughness mechanism beyond the simple flaw-related grain-size effects generally recognized for polycrystalline ceramics. However, many attributes are still shared with other polycrystalline materials, including the simplicity of processing; no requirement for “aging” heat treatments, and reciprocally, relative insensitivity to follow-on heat
DENTAL MATERIALS 24(2008)289-298 293 treatments (e.g. porcelain firings in dental usage)and the much broader grain size range, approximately 8-0.25 um[26] opportunity to explore chemistry-based powder fabrication Both grain size and the test temperature in relationship to the and nano-scale microstructures Ms temperature control the size of the transformation zone (t These materials consist primarily of equiaxed grains of or h)associated with a growing crack(to be discussed momen t-zro2 sintered (and sometimes hot isostatically pressed or tarily); with rr directly influencing toughness mechanisms HIP'ed)to 96-99.5% of theoretical density. It has recently For 2 mol% Y-TZP, AK was clearly shown to decrease as tem- been reported that some grain-boundary segregation of Y3+ peratures moved away(higher from the Ms temperature[26] ions occurs during late-stage sintering[22]. It is energetically Transformation zone size of a crack extended at two different favorable for the cubic phase to form at yttria-rich triple junc- temperatures indicated an rr of slum at 295K and =10 um ons and grain boundaries at temperatures between 1300 and at 80K(with Ms presumed to be below 80K)(26] 1500 C[22]. The implications for hydrolytic stability(low tem- Nano-scale transformation-toughened Zro2 is alread erature degradation) by having minor volume fractions of appearing in the literature and in limited commercializa Zro2 segregated at grain boundaries along with the depletion tion. It was reported in 2002 that the trend toward increased of Y3+ from t-Zro2 will be discussed later. phase stability with decreasing particle-size of t-ZrO2 could Properties of these essentially homogeneous ceramics are be overcome by adjusting the yttria dopant concentration[23] primarily a function of grain size, in that grain size controls the Whereas 3 mol% Y2O3 has been found to optimize toughness Ms temperature and the ease of transformation(and hence the in micrometer and sub-micrometer t-ZrO2, dopant concen- toughness effect). The closer the test or service temperature trations and critical grain size for nano-scale material were is to the Ms temperature(but still above it avoiding sponta- identified as 1.0 mol% for 90nm and 1.5 mol% for 110 nm; neous transformation) the less stable are the t grains and the both combinations reaching around 16 MPa m[23. As with higher the available toughness increment. For a given dopant micro-scale zirconias, strong grain size effects(decreasing concentration the toughness increment decreases as grains toughness with decreasing grain size) were exhibited in these become much smaller than the critical size, presumably due to nano-scale ceramics as well[23]. At least one commercial over-stabilization"of the grains, eliminating the t-m trans- nano-scale Ce-TZP containing 20%Al2 O3(Nanozir, Matsushita formation upon introduction of a crack[23] Electrical Works, Ltd. is being examined as a dental ceramic This grain size effect itself is controlled by the type [28 having a reported fracture toughness of approximately dopant and the dopant concentration that are essentially 20 MPa m12(E. Rothbrust, Ivoclar Vivadent, Inc, personal com- determining(1)the degree of tetragonality (i.e. crystal lattice munication). It may be that nano-scale t-zro2 will primarily length ratio of c/a being>1.0)and(2) the thermal expan- appear in a polycrystalline form due to the difficulties for sion anisotropy (c versus a directions) of the unit cells. The intra-granular precipitation and tertiary phase development compositional effect of the dopant can be represented by required in PSZ. materials the resultant unit cell tetragonality (c/a[24, 25]. In general higher tetragonality contributes to a less stable material char- cterized by an increased Ms temperature [24]. Based on4. Mechanism(s)and consequences of tragonality, at similar grain size and dopant concentra- transformation tions yttrium appears to be a stronger stabilizer than cerium and especially titanium [25]. Anisotropic thermal expansie Numerous factors are discussed in the literature as (1)nucle- for the c and a axes, can influence residual strains in t ating and driving the transformation and (2)controlling grains; higher residual stresses can lower the nucleation stress the consequences of transformation. Two main phenomena threshold for the t-m in the presence of crack-tip strain resulting from transformation include (1)increased resistance energy[26]. Linear thermal expansion coefficients()are avail- for growth of both short(<100 um)and long(>0.3mm)cracks obtained by direct measure oflattice parameters from 800 to with crack length (termed R-curve behavior) until generally room temperature. These measures indicate that anisotropy reaching a toughness plateau. These transforming is minimal (aa and ac crossover or, equivalently, aa/ac =1) at step away from the simple Griffith dependence on flaw size 4mol%Y2O3[27] and many have strengths that depend on the stress needed In its most basic form, a transformation-toughening con to trigger transformation rather than being flaw-size ser tribution(ak)has been defined as[26 tive. Quite non-linear behavior is exhibited by the toughest materials bordering on quasi-plasticity with measurable pre (2) failure deformation. Therefore, as will be discussed below, high strength and high toughness do not present in the same where ko is the fracture resistance inherent to the matrix with- material out transformation toughening. In general, the toughness con Driving forces and the role of temperature particularly for tribution from transformation(15 MPam1/2)exceeds that for the t-m transformation can be simply considered within microcracking (2-6 MPam 2) or deflection(-2-4MPamv2) a thermodynamic framework, as reproduced here following the work of Becher and Swain [26]. The total unit volume At a given temperat ransformation-toughening free energy change AFo for the transformation, including an ontribution(△K)for2 TZP decreased from approx applied stress is ain size decreased from 0.5 um; for 12 mol% Ce-TzP this same range occurred △Fo=△FcH+△Ue+△Us-△U1
dental materials 24 (2008) 289–298 293 treatments (e.g. porcelain firings in dental usage) and the opportunity to explore chemistry-based powder fabrication and nano-scale microstructures. These materials consist primarily of equiaxed grains of t-ZrO2 sintered (and sometimes hot isostatically pressed or HIP’ed) to 96–99.5% of theoretical density. It has recently been reported that some grain-boundary segregation of Y3+ ions occurs during late-stage sintering [22]. It is energetically favorable for the cubic phase to form at yttria-rich triple junctions and grain boundaries at temperatures between 1300 and 1500 ◦C [22]. The implications for hydrolytic stability (low temperature degradation) by having minor volume fractions of c-ZrO2 segregated at grain boundaries along with the depletion of Y3+ from t-ZrO2 will be discussed later. Properties of these essentially homogeneous ceramics are primarily a function of grain size, in that grain size controls the Ms temperature and the ease of transformation (and hence the toughness effect). The closer the test or service temperature is to the Ms temperature (but still above it avoiding spontaneous transformation) the less stable are the t grains and the higher the available toughness increment. For a given dopant concentration the toughness increment decreases as grains become much smaller than the critical size, presumably due to “over-stabilization” of the grains, eliminating the t→m transformation upon introduction of a crack [23]. This grain size effect itself is controlled by the type of dopant and the dopant concentration that are essentially determining (1) the degree of tetragonality (i.e. crystal lattice length ratio of c/a being > 1.0) and (2) the thermal expansion anisotropy (c versus a directions) of the unit cells. The compositional effect of the dopant can be represented by the resultant unit cell tetragonality (c/a) [24,25]. In general higher tetragonality contributes to a less stable material characterized by an increased Ms temperature [24]. Based on tetragonality, at similar grain size and dopant concentrations yttrium appears to be a stronger stabilizer than cerium and especially titanium [25]. Anisotropic thermal expansion, for the c and a axes, can influence residual strains in t grains; higher residual stresses can lower the nucleation stress threshold for the t→m in the presence of crack-tip strain energy [26]. Linear thermal expansion coefficients (˛) are available for yttria-doped ZrO2 over a limited concentration range, obtained by direct measure of lattice parameters from 800 ◦C to room temperature. These measures indicate that anisotropy is minimal (˛a and ˛c crossover or, equivalently, ˛a/˛c ∼= 1) at 4.5mol% Y2O3 [27]. In its most basic form, a transformation-toughening contribution (KT) has been defined as [26]: Kc = Ko + KT (2) where Ko is the fracture resistance inherent to the matrix without transformation toughening. In general, the toughness contribution from transformation (≈15 MPa m1/2) exceeds that for microcracking (≈2–6 MPam1/2) or deflection (≈2–4 MPam1/2) mechanisms [9]. At a given temperature, the transformation-toughening contribution (KT) for 2mol% Y-TZP decreased from approximately 12–2.5 MPam1/2 as grain size decreased from 2 to 0.5m; for 12mol% Ce-TZP this same range occurred over a much broader grain size range, approximately 8–0.25m [26]. Both grain size and the test temperature in relationship to the Ms temperature control the size of the transformation zone (rT or h) associated with a growing crack (to be discussed momentarily); with rT directly influencing toughness mechanisms. For 2mol% Y-TZP, KT was clearly shown to decrease as temperatures moved away (higher) from the Ms temperature [26]. Transformation zone size of a crack extended at two different temperatures indicated an rT of ≤1m at 295 ◦K and ∼=10m at 80 ◦K (with Ms presumed to be below 80 ◦K) [26]. Nano-scale transformation-toughened ZrO2 is already appearing in the literature and in limited commercialization. It was reported in 2002 that the trend toward increased phase stability with decreasing particle-size of t-ZrO2 could be overcome by adjusting the yttria dopant concentration [23]. Whereas 3mol% Y2O3 has been found to optimize toughness in micrometer and sub-micrometer t-ZrO2, dopant concentrations and critical grain size for nano-scale material were identified as 1.0mol% for 90 nm and 1.5mol% for 110 nm; both combinations reaching around 16 MPam1/2 [23]. As with micro-scale zirconias, strong grain size effects (decreasing toughness with decreasing grain size) were exhibited in these nano-scale ceramics as well [23]. At least one commercial nano-scale Ce-TZP containing 20% Al2O3 (Nanozir, Matsushita Electrical Works, Ltd.) is being examined as a dental ceramic [28]; having a reported fracture toughness of approximately 20 MPam1/2 (F. Rothbrust, Ivoclar Vivadent, Inc., personal communication). It may be that nano-scale t-ZrO2 will primarily appear in a polycrystalline form due to the difficulties for intra-granular precipitation and tertiary phase development required in PSZ materials. 4. Mechanism(s) and consequences of transformation Numerous factors are discussed in the literature as (1) nucleating and driving the transformation and (2) controlling the consequences of transformation. Two main phenomena resulting from transformation include (1) increased resistance for growth of both short (≤100m) and long (≥0.3mm) cracks and, for many ceramics, (2) toughness continuing to increase with crack length (termed R-curve behavior) until generally reaching a toughness plateau. These transforming ceramics step away from the simple Griffith dependence on flaw size and many have strengths that depend on the stress needed to trigger transformation rather than being flaw-size sensitive. Quite non-linear behavior is exhibited by the toughest materials bordering on quasi-plasticity with measurable prefailure deformation. Therefore, as will be discussed below, high strength and high toughness do not present in the same material. Driving forces and the role of temperature particularly for the t→m transformation can be simply considered within a thermodynamic framework, as reproduced here following the work of Becher and Swain [26]. The total unit volume free energy change FO for the transformation, including an applied stress is: FO = FCH + Ue + US − UI (3)