In 1976, ZrO₂ was revealed to display high toughness (R. C. Garvie and Hannink, 1975), which ushered in a decade of exceptional ceramic toughening development, culminating in materials having toughness on the order of 20 MPa m^1/2 (Evans, 1990). Undoped ZrO₂ displays three phases at different temperature: the monoclinic (m) phase of undoped ZrO₂ is thermodynamically stable at temperatures below 1,170°C; it is a tetragonal (t) phase from 1,170 to 2,370°C, and a cubic (c) phase above 2,370°C until the melt occurs at 2,706°C (French et al., 1994). Because the t-to-m phase transformation in zirconia may be used to toughen the material and enhance its mechanical properties, it has sparked a lot of technological interest in zirconia and composites incorporating zirconia (Heuer, 1987).

Using ZrO₂ alone have the shortage of low-temperature degradation (LTD) (Kohorst et al., 2012), so Al₂O₃ are doped to exert synergistic effect. Two composite materials can be made in the ZrO₂-Al₂O₃ system as follows: Al₂O₃ reinforced with ZrO₂ particles and denominated as zirconia toughened alumina (ZTA), or ZrO₂ reinforced with Al₂O₃ particles, named alumina toughened zirconia (ATZ) (Nevarez-Rascon et al., 2009). According to the findings of De Aza et al. (Ritchie, 1988; De Aza et al., 2002) and Gregori et al. (Gregori et al., 1999), the fracture toughness of the ceramic matrix material rose in both cases. ZTA can be considered a new generation of toughened ceramic, with toughness exceeding 12 MPa m^1/2, compared to 3 MPa m^1/2 for commercial alumina ceramic. The inclusion of discrete zirconia grains in the alumina matrix as a second phase allows the former to behave intrinsically, that is, to undergo the t-to-m transformation or to remain in the metastable tetragonal state after cooling of the composites from fabrication temperatures (Wang and Stevens, 1989). ZTA is microstructurally separated into four classes by Wang et al. (Figure 2A) (Wang and Stevens, 1989): (I) alumina with dispersed unstabilized zirconia (Figures 2A,a); (II) alumina with dispersed partially stabilized zirconia (PSZ) (Figures 2A,b); (III) alumina with PSZ agglomerates (Figures 2A,c); (IV) alumina-zirconia duplex structures (Figures 2A,d).

Except for these basic forms of ZrO₂-Al₂O₃ materials, the shapes of both components could be altered, or other metal compounds could be added, to achieve better toughening effect. Aguilar-Elguézabal and Bocanegra-Bernal (2013) investigated the fracture toughness of an Al₂O₃(n)-70 wt% ZrO₂ (ZrO₂ with 3 mol% yttria, TZ-3Y)n nanocomposite with addition of 2.5% Al₂O₃ whiskers. The fracture toughness was enhanced 62, 28, and 7% over pure Al₂O₃, the composite without additions of whiskers, and pure TZ-3Y for medical applications, respectively. They achieved a maximum fracture toughness of 6.9 MPa m^1/2 with an average grain size of 0.4 ± 0.17 μm, and found microcracking, crack deflection and phase transformation toughening mechanisms could contribute to the improvement of fracture toughness. Additionally, the Al₂O₃ whiskers did not affect other intrinsic properties during the toughening process. Santos et al. (2021) not only added Al₂O₃ in ZrO₂, but also replaced the conventional tetragonal zirconia stabilized with 3 mol% yttria (3Y-TZP) with ceria and yttria stabilized zirconia, namely (Ce, Y)-TZP. And this new composite received higher fracture toughness, elastic modulus and hardness than those obtained for 3Y-TZP. In the meantime, its values of Weibull modulus (m > 10) and flexural strength (> 950 MPa) were similar to the 3Y-TZP ceramics, with high resistance to degradation in saliva, indicating adequate properties for dental application. They thought several different toughening mechanisms acting simultaneously in the material. Apart from the above ingredients of Santos et al.’s study, Gommeringer et al. (2019) reinforced 1Y6Ce-TZP materials with different amounts of alumina and/or strontium hexaalumina by slip casting and pressureless sintering at different temperatures. Then these materials underwent accession of their mechanical properties, microstructure, phase composition, and low-temperature degradation stability, which showed that they exhibited a high fracture resistance of 10–12 MPa m^1/2. Partial substitution of the alumina dispersion by strontium hexaaluminate can improve the strength, while pres.erving toughness, hardness and low-temperature degradation resistance. The residual alumina provides grain growth inhibition. In summary, the improvement of structure or/and composition of the second phase is helpful to the toughening of ZrO₂-Al₂O₃ ceramic.

ZrO₂ and Al₂O₃ could also be added to toughen ceramics except for ZTA or ATZ. Gali et al. (2018) investigated ZrO₂ toughened mica glass ceramics for dental restorations. It was recorded that mica glass ceramics with 20 wt% yttria stabilized zirconia (YSZ) reached Vickers hardness of 9.2 GPa, elastic modulus of 125 GPa, indentation toughness of 3.6 MPa m^1/2, and chemical solubility of 30 g/cm² (well below the permissible limit), respectively. The phenomenon of transformation toughening of YSZ was observed in enhancing the toughness properties of ceramic. Because of its osseointegration property, Calcium silicate (CaSiO₃, CS), has been investigated as a bioactive biomaterial for bone tissue repair and replacement (Mehrali et al., 2014). Long et al. (2008) successfully fabricated β-CaSiO₃/ZrO₂ (3Y) nanocomposites via spark plasma sintering. Adding ZrO₂ could enhance the phase transitional temperature of CaSiO₃ and inhibit its phase transition. When the β- CaSiO₃/ZrO₂ mole ratio was 60/40, the fracture toughness and strength of the nanocomposites were as high as 4.08 MPa m^1/2 and 395 MPa, respectively. Shuai et al. (2016) used zinc oxide whisker (ZnOw) to improve CaSO₄/bioglass scaffolds which were prepared via selective laser sintering (SLS). And the ZnOw improved fracture toughness and compression strength significantly. The combination of several toughening mechanisms, including whiskers pull-out, crack bridging, crack deflection, and crack branching, might be attributed to the increase of mechanical characteristics. Shirazi et al., (2014) investigated the effect of Al₂O₃ on α-CaSiO₃ ceramic and found that 15 wt% of Al₂O₃ addition at 1,250°C enhanced fracture toughness as well as hardness of CaSiO₃.

Apart from ZrO₂-Al₂O₃ system, other reinforcement-ceramic matrix systems have also been studied. Rasmussen et al. (2004) mixed core particles of cesium (Cs₂O) -containing synthetic leucite with two cesium-free matrix porcelains, a commercial porcelain (VP) and a synthesized leucite-based porcelain (NP). The toughness of both types of composite materials was dependent on Cs₂O content of the added core particles, and the ceramic containing 0.75 mol% Cs₂O reached a maximum toughness. The possible toughening mechanism was transformation toughening. Luo et al. (2002) modified glass-infiltrated Al₂O₃ with 0.5 wt% MgO. And the fracture toughness of this composite (5.12 MPa m^1/2) is obviously higher than the partially sintered Al₂O₃ (0.58 MPa m^1/2), MgO-modified partially sintered Al₂O₃ (1.86 MPa m^1/2), or glass-infiltrated Al₂O₃ (3.91 MPa m^1/2).

Most above mentioned second phases for toughening are metal compounds rather than metals alone. Besides, the research of Fujieda et al. (2012) and Uno et al. (2013) showed that metal nanocomposites (NPs) could also be used to increase the fracture toughness of ceramic. Fujieda et al. (2012) found that the addition of silver (Ag) and platinum (Pt) NPs improved the mechanical properties, including the fracture toughness and the Young’s modulus, of commercial ceramic. The fracture toughness of AgNPs was higher than that of PtNPs. This could be due to the difference in positive compressive stress generation, which is projected to be greater for silver due to its higher coefficient of thermal expansion. Uno et al. (2013) researched the effects of adding AgNPs on the toughening of dental porcelain, Noritake Super (NS) Porcelain AAA. When the concentration of Ag in the solution was 500 and 1,000 ppm (Ag500 and Ag1000), toughness values of 1.54 MPa m^1/2 and 1.51 MPa m^1/2 were observed, respectively, which were higher than that of the control (1.36 MPa m^1/2). The inclusion of Ag500 and Ag1000 NPs, on the other hand, resulted in a color shift. Some AgNPs remained nanoparticles, whereas others interacted with matrix elements and turned into silver ions. As a result of the ion exchange reaction and differential thermal expansion of the silver metal nanoparticles, a residual compressive stress was created. Swain S.K et al. (2016) developed a novel biocompatible β-tricalcium phosphate (β-TCP)-based composited by reinforcing ceramic matrix with 30 vol% of a biodegradable iron-magnesium (Fe-Mg) metallic phase, scilicet β-TCP-15Fe15 Mg and β-TCP-24Fe6 Mg (vol%) composites. Both these composites had increasing mechanical properties.

Both metal and non-metal reinforcements have their advantages and disadvantages. Metal fillers like ZrO₂ or Al₂O₃ have esthetic color but unexpected LTD or lower strength, while non-metal fillers like nanocarbon materials possess high elasticity modulus but susceptible to high sintering temperature. Proper combination of metal and non-metal components could be a promising approach to toughen bioceramics. Cesar et al. (2005) modified six kinds of dental porcelains with the addition of leucite. Leucite (K₂O·Al₂O₃·4SiO₂) is a potassium–aluminum–silicate phase that can be incorporated into the dental porcelain in two ways: the incongruent melting of potash feldspar (K₂O·Al₂O₃·6SiO₂) or as a synthetic powder. For the materials investigated, the higher the leucite content in the porcelain, the higher the fracture toughness, implying that they would have better clinical performance. Xiang et al. (2007) fabricated mica-based glass-ceramics containing needle-like fluorapatites. They found that the ceramics with higher fluorapatite content (mainly containing CaO and P₂O₅) presented higher fracture toughness and Vickers hardness, due to a large amount of needle-like fluorapatite crystals and the fine microstructure. Luo et al. (2002) fabricated MgO-modified glass infiltrated Al₂O₃ for CAD/CAM. There are two main functions of glass infiltration: the first is to eliminate almost all porosities, which are prone to crack initiation; secondly, the difference in the coefficients of thermal expansion between Al₂O₃ and glass produces compressive stresses, further enhancing strength at the Al₂O₃-glass interface. Furthermore, MgO improved grain size uniformity, controlled grain growth, and promoted homogeneous wetting of Al₂O₃ grains by the liquid through an alternation in interfacial energies, allowing the manufacture of high-density ceramics.

The underlying toughening mechanism of controlling grain size is mainly fine-grained toughening. The reinforcement fillers located at the grain boundary provide a pinning effect, which can not only inhibit grain growth but also promote filler dispersion (Karthiselva et al., 2018). The toughening effect of fine-grain reinforcement is quite obvious according to the Hall-Petch formula. And the vital factors influencing the efficiency of fine-grained toughening include the content of second phase, the geometric shape, as well as the temperature and impurities (Liu et al., 2020).

Transformation toughening mainly occurs in ceramics consisting of ZrO₂, and could sometimes happen for other components as well. When an external tensile stress is applied to a crack tip, metastable tetragonal ZrO₂ inclusions scattered in a ceramic matrix are known to convert to their thermodynamically stable monoclinic form (R. C. Garvie and Hannink, 1975; Lange, 1982). The stress to initiate transformation is lower than the fracture stress, and transformation happens preferentially and proceeds until the transformable t phase is exhausted (Hannink et al., 2000). Therefore, the phase transformation, which is accompanied by a volume expansion of 4% and shear strain of 6%, creates a compressive stress that slows and eventually stops the crack propagation, while the strain energy associated with any net shear component of the transformation strain in the transformation zone contributes an effective increase in the energy of fracture (Wang and Stevens, 1989). The contribution of stress-activated transformation, ∆K_cT, to the toughness is usually expressed in the form Δ K c T = η E ∗ e T V f h 1 / 2 1 − ν (2) Where η is a factor depending on the zone shape at the crack tip and the nature of the stress field in that zone, E^∗ the effective modulus of the material, e_T the dilatational strain, V_f the transformed volume fraction of the particles, h the width of the transformation zone from the crack surface, and νthe Poisson ratio (Hannink et al., 2000).

Microcracks in ceramic materials can be subdivided as residual microcracks and stress-induced microcracks. Take the ZrO₂ ceramic as an example, the residual microcracks are those caused by the volume expansion and shear strain associated with t-to-m phase transformation mentioned in section above, while the stress-induced microcracks are from the volume expansion and shear strain associated with the subsequent stress-induced transformation during fracture process. During phase transformation, tangential stresses are generated around the transformed m-ZrO₂ particles. Microcrack toughening mainly relies on two shielding sources. One is the main-crack and microcrack interaction which causes stress redistribution ahead of the main-crack tip and thus lowers the continuum stiffness of the microcracked material. The other is the stress redistribution at the main-crack tip due to the release of residual stress when a microcrack is nucleated (Hutchinson, 1987). Microcracks form at the intersections of the fillers and the matrix as a result of this, which can absorb fracture energy by extending in the stress field of a propagating crack or deflecting the propagating crack (Wang and Stevens, 1989). Microcracks behind the main-crack tip provide the most shielding, while microcracks ahead of the main-crack tip contribute no shielding effect. What’s more, compressive normal residual stress may attribute to an anti-shielding effect at the main-crack tip, which should be paid attention for the design of ceramic materials (Chen, 2005).

Crack deflection refers to the condition that the reinforcement fillers will generally deflect the crack at a certain angle when the crack propagates in the ceramic matrix and meets fillers (Zou et al., 2018). So the fracture energy of crack propagating is consumed and the rate of crack propagating is reduced, as a result, the fracture toughness is increased (Zou et al., 2018). Specifically, deflection toughening occurs whenever interaction between the crack front and the minor phase produces a non-planar crack, subject to a stress intensity lower than that experienced by the corresponding planar crack. And the non-planar crack arises either from the existence of weakened interfaces or from residual strains present in the material. Considering shape, the rod of high aspect ration is the most effective morphology to deflect propagating cracks, which could account for four-fold increases in fracture toughness (Faber and Evans, 1983).

Crack bifurcation occurs at the matrix-filler interface. The primary crack runs vertically, whereas the bifurcation crack runs parallel to the matrix-filler interface. When a micro-crack forms at the end of the interface and can be deflected at 90° to extend to the ceramic matrix, the bifurcation crack is ended, which helps stress propagation at the corner. The new fracture then spreads parallel to the primary crack (Cheng et al., 2018).

Take advantage of crack bridging, several types of discontinuous reinforcing fillers have been successfully applied to form toughened ceramics including second-phase whiskers (Becher, 1990), platelets (Becher, 1991), as well as elongated, plate-like, and large matrix grains (Himsolt et al., 1979; Mussler et al., 1982). Crack bridging happens during ceramic cracking. The two ends of the reinforcement filler link the crack surface and begin to distort as the distance between the two crack walls grows, which imposes a closure force on the crack. The two segments of fillers are tightly linked to the matrix throughout the process and consume the crack propagation energy.

Before reaching the reinforcement’s deformation limit, a pull-out may occur, consuming the fracture energy, or filler-matrix debonding with crack bridging may occur. Defects in this mechanism alter the mobility of fillers, which in turn affects energy dissipation (Liu et al., 2020). The extent of pullout of discontinuous reinforcement is commonly quite limited, which is due both to the short length of such phases and the fact that bonding and clamping stresses often discourage pull out. However, pullout cannot be ignored as even short pull-out lengths contribute to the toughness achieved (Li, 2012).

Biomedical applications require ceramic materials with excellent biocompatibility to ensure device acceptance and patient safety, which is the prior recommend compared with toughness. First, no short- or long-term cytotoxicity is expected for biomaterials. Nieto et al. (2017) reviewed studies on graphene reinforced ceramic matrix composites (GNP-CMCs) and found all studies showed no cytotoxic effects induced by GNPs. After evaluating the long-term biocompatibility of a new ZTA, mainly in terms of DNA damage, mutagenicity and cancerogenetic potential in mammalian cells, Maccauro et al. (2009) found no long-term carcinogenic effect. Besides, toughening ceramic strategies should not affect the function of cells. In in vitro tests, the β-TCP-15Fe15Mg composite fabricated by Swain S. K et al. (2016) was shown to support the attachment and proliferation of osteoblast and endothelial cells and the cells exhibited characteristic markers for bone formation and angiogenesis, respectively. For the second phase addition strategy, proper concentration of reinforcement is a critical factor to be controlled. According to the evaluation of Zhang et al. (2019), hydroxyapatite ceramic with 3 wt% ZrO₂ showed good cell viability and no cytotoxicity, and the mouse bone mesenchymal stem cells continued to proliferate on the ceramic surface observed for 5 days. Shuai et al. (2016) doped ZnO whisker (ZnOw) in the ceramic matrix, and cells attachment, extension and interconnection behavior of MG63 cells were enhanced with the ZnOw increasing from 1 wt% to 5 wt%. However, when ZnOw were further increased to 10 wt%, a sharp decrease of cell expansion degree was observed.

Biocompatibility of toughened ceramic has been studied mostly in terms of in vitro cytocompatibility, but fewer in vivo analysis has been conducted, which is a vital process before being considered for clinical applications. Petre et al. (2019) investigated the in vivo biocompatibility of both resorbable and nonresorbable PVA fiber-reinforced calcium phosphate cements. And these fibers were well dispered in the matrix without inducing inflammatory responses or other adverse reactions. Moreover, the inclusion of PVA fibers did not negatively affect the ingrowth of new bone into the PVA fiber-reinforced cements. The in vivo test of Kucko et al. (2019) also showed that PVA fibers do not compromise the excellent osteocompatibility of calcium phosphate cements. Nevertheless, further in vivo studies should be performed to investigate the biocompatibility of various types of toughened ceramics in more detail.

Degradability is significant for biomaterials like bone substitutes, which determines the final repair effect of bio-implants; while biomaterials for other applications, such as dental restorations or artificial joint, require to be stable rather than degradable. Therefore, most degradability tests of toughened ceramic are conducted in bone substitute ceramic composites. Significantly influenced by mechanical properties of biomaterials, degradability is mostly tested by soaking materials in the simulated body fluid (SBF) (Xiang et al., 2007). Toughening strategies, especially second phase addition may change the degradation rate and degradability. Li et al. (2015) found that the addition of ZnO or ZrO₂ decreased the degradation rate, and improved the degradability of wollastonite. Similarly, Feng et al. (2014) observed the degradation rate of calcium silicate ceramic scaffolds decreased with the increase of hydroxyapatite whisker content. Moreover, toughened bone substitute ceramic possesses the apatite-forming ability in SBF, and osteoblast-like cells spread well on the scaffolds and proliferated with increasing culture time (Feng et al., 2014; Mehrali et al., 2014; Li et al., 2015). However, soaking in the SBF tests only the in vitro bio-degradability, and in vivo degradability should also be increasingly studied in the future.

Toughness and strength are typically exclusive properties, which means the improvement of toughness usually comes with a decrease in strength (Launey and Ritchie, 2009). If the decrease of strength is within proper range, the toughening methods are still acceptable, because of the following reasons. On the one hand, the strength of most ceramic is high enough to satisfy clinical requirements. On the other hand, the best strength is not the highest but the strength that matches those of the human tissues. Therefore, toughening is a more essential issue than strengthening for ceramics of biomedical applications and receives extensive attentions. Toughening strategies from second phase addition, surface modification to manufacturing process improvement have been studied to solve this problem. Considering different clinical application scenarios, ceramics for biomedical applications could be classified as esthetic-demanding and non-esthetic-demanding: dental restorations are the representatives of the former type of ceramic, in which esthetic sometimes is even paid more attention than toughness; while the latter, such as dental implants and artificial joints which could not be seen, mainly requires promising mechanical properties, especially toughness. What’s more, studies concerning esthetic, including color and translucency, are still limited. Adding reinforcing fillers as second phase to ceramic matrix is one of the most widely used strategies, and even commercial ceramic products have been fabricated based on this strategy; in contrast, surface modification strategies gain less studies and applications. Toughening based on ZrO₂-Al₂O₃ system or mica glass ceramics (Gali and Kumar, 2019), and using ion-exchange (chemical methods) tends to preserve excellent esthetics, while adding colorful fillers like AgNPs (Fujieda et al., 2012) would affect ceramic color. For non-esthetic demanding applications, high toughness is the first consideration, which is fine achieved by fiber/whisker-shape (Aguilar-Elguézabal and Bocanegra-Bernal, 2013) or nanocarbon adding (Siddiqui et al., 2018).

The ZrO₂-Al₂O₃ system is well studied and used in clinic, such as dental restorations and implants, as well as femoral heads and hip joints. Although the addition of Al₂O₃ helps to improve the problem of low-temperature degradation (LTD) of ZrO₂, it should still be paid enough attention during application, because most medical devices are applied in humid body environment which promotes degradation (Nguyen et al., 2009; Borchers et al., 2010). LTD refers to the t-to-m transformation of ZrO₂ induced by hydrothermal aging in the humid environments, and with LTD, the energy barrier for t-to-m transformation is decreased due to incorporation of water constituents into zirconia lattice (Guo, 2004). Non-metal materials, especially nanocarbon materials, show their potential as reinforcement fillers due to excellent mechanical properties. Nevertheless, the application of nanocarbon materials still faces some challenges. Firstly, it’s difficult for nanocarbon to disperse in ceramic matrix due to its large specific surface area, surface energy, van der Waals forces caused by the intermolecular electrical dipole moment, and interactions between functional groups, as well as easy aggregation and entanglement properties (Liang et al., 2018). Secondly, the interface invasion between nanocarbon and ceramic is regarded unsuitable due to the differences in surface tension and density. Furthermore, it is difficult to determine the temperature during sintering, because nanocarbon can be destroyed at high temperatures when interface bonding is diminished. What’s more, measurement of ceramic matrix composites with nanocarbon is not homogeneous, for the composites are extremely anisotropic (Liu et al., 2020). Therefore, combination of different types of second phase materials may be a possible approach to exert synergistic effect and, in the meantime, diminish their disadvantages.

In addition to the type of second phase, the complex transition zone between ceramic matrix and reinforcement fillers, namely the interface, also play an important role on toughening, which is the basis of toughening mechanisms, such as pull-out and crack bridging. Proper interface strength is neither too weak nor too strong, so that the toughening mechanisms could dissipate energy during the loading process appropriately (Liu J et al., 2013; Ramirez et al., 2014; Ahmad et al., 2015).

De Aza et al. (2002) pointed out that the idea of a well-defined stress limit, toughness (K_IC), must incorporate the concept of a threshold (K_I0) under which crack propagation does not occur. The K_IC usually used only represents the resistance to fast fracture. However, ceramic materials are susceptible to slow crack propagation at K_I values under K_IC, which is attributed to stress assisted corrosion at the crack tip, or any pre-existing defect in ceramic. The combined effect of high stresses at the crack tip and the presence of water or body fluid induce crack propagation in a subcritial manner (Chevalier et al., 1999). This slow crack propagation threshold may represent an intrinsic property of ceramic that gives information of its mechanical properties more realistic than the widely used toughenss, which represents only fast crack growth. The concept of slow crack propagation reminds us to pay attention to evaluations of mechanical properties, which may overestimate the behavior of ceramics.

Microscopic building blocks, weak interfaces, and architecture are responsible for the extraordinary qualities of highly mineralized natural materials. Bio-inspired concepts could pave the way for new techniques to improve the toughness of brittle materials with enticing features but limited uses due to their brittleness (Wondraczek et al., 2011). These new materials show that bio-inspired techniques can be used to achieve both strength and toughness, which are normally mutually exclusive qualities (Ritchie, 2011). Nature transforms brittleness into toughness through three “overarching features”: (I) stiff and hard construction blocks are separated by (II) weaker interfaces, which are organized in (III) specialized architectures (Mirkhalaf et al., 2014). It may seem contradictory to make a material tougher by adding weak interfaces, but it appears to be a ubiquitous and powerful method in natural materials. Mirkhalaf et al. (2014) used three-dimensional laser engraving to create weak interfaces inside the bulk of glass and then infiltrating the interfaces with polyurethane. The crack is channeled towards toughening configurations by the weak interfaces, which obstructs its propagation. The technology produces a bio-inspired glass that is 200 times tougher than non-engraved, “intact” glass (namely bulk glass, which was not engraved, but that is not devoid of the surface defects typical to glass exposed to air). Glass and ceramic are both brittle materials, therefore, above mentioned bio-inspired method could also be attempted in ceramics. These bio-inspired concepts have opened new pathways to improving the toughness of ceramics, and in the recent years, increasing numbers of synthetic composites inspired from biological materials have emerged (Munch et al., 2008; Espinosa et al., 2011). This is one of the development directions of the future, with still limited researches, which calls for more investigating efforts.