The microstructures of the polished surfaces of composites A, B, C, and D are shown in Fig. 1(a-d), respectively. From the figure, we can easily distinguish the zirconia grain (light colored) and the alumina grain (grey colored) due to the atomic number difference. The zirconia particles are well distributed in alumina matrix which could inhibit abnormal grain growth of alumina grains. In addition, based on the microstructure observation in higher magnification factors shown in Fig. 2(a), it is found that some flake-like grains exist in Zr4+/La3+ system, and the EDS characterization of such a grain indicates that is composed the element of O, Al and La. Meanwhile, EDS of the matrix grain indicates that the grain is composed of the element of O and Al (Fig. 2(b)). In order to clarify the phase of the flake-like grains, an XRD test was conducted and shown in Fig. 2(c). The XRD patterns of all sample show that Al2O3 and t-ZrO2 are the major phase, and m-ZrO2 minor phase, respectively. Moreover, different from the other three sample, a new phase LaAl11O18 occurs in sample A, which corresponds to the flake-like grain in Fig. 2(a). The generation of the LaAl11O18 phase could be explained as follows [19], which could also explain the high intensity of m-ZrO2 in sample A.

In addition, the microstructure of Zr4+/La3+ system is very dense without obvious pores, which could be further proved by the typical TEM image shown in Fig. 2(d). The weights increase, average grain size, density, relative density and Vickers hardness of the four researched systems combined with the referenced pure alumina and ZTA are listed in Table 1. It can be seen that the alumina grain growth is effectively inhibited after infiltration (from 2.21 µm to less than 1 µm). It can also be observed that the densities of all the infiltrated samples which are above 99% are higher than that of the sintered samples which has not been subjected to infiltration (96.21%). However, although the densities of the samples subjected to infiltration increased, considering the introduction of zirconia and the hardness of zirconia is lower than that of alumina, the hardness of all the composites subjected to the infiltration treatment are not much higher than the pure alumina.

The crack resistance as a function of the crack size of the Zr4+/La3+ system is shown in Fig. 3 (a). The fracture toughness of Sample B, C, and D remained basically unchanged versus the crack size, the phenomenon of which is common in brittle materials. However, it can be seen that the fracture toughness of sample A increased with the increase of the crack size. Herein, the toughness of the composites can be divided into three parts using the following equation [20, 21]

Where is fracture toughness of the Al2O3 matrix, is the contribution from the ZrO2, and is the contribution from the LaAl11O18 platelets. As can be seen in XRD, only a little m-ZrO2 exist in sample B, C, and D. The intensity of m-ZrO2 phase is higher than that in B, C and D while the fracture toughness of Sample A is higher than the other four, which illustrate the effect of phase transformation toughen mechanism is small for Sample A. Hence, to understand the enhanced , we can focus on the residual stress caused by thermal expansion misfit. The global residual stress of composites can also be computed as following equations [22].

And,

Where *σ*m and *σ**l *are residual stress in ZTA matrix and LaAl11O18 platelets, *α*m and *α*l (*α*m = 8.7 × 10-6/K,*α*l = 7.7 × 10-6/K) are thermal expansion coefficients of ZTA matrix and LaAl11O18 platelets, respectively. *ν*m and *ν*l are Poisson’s ratios of ZTA matrix and LaAl11O18 platelets, *E*m and *E*l are Young’s modulus of of ZTA matrix and LaAl11O18 platelets [23], *f**m *is the volume fraction of the ZTA matrix, Δ T is the difference in temperature over which the stress is locked in. From the calculation of equations above, the residual stress of LaAl11O18 versus temperature is shown in Fig. 4(a), and the compressive stress in LaAl11O18 platelets is 689 MPa, while the tensile stress of ZTA matrix is 28 MPa. The thermal expansion coefficient differences between ZrO2 (10.5×10−6/℃), Al2O3 (8.5×10−6/℃), and LaAl11O18 (7.7×10− 6/℃) impose compressive stress on Al2O3 grain boundaries, which strengthens the grain boundary and leads to more energy consumed when crack passes through flake-like LaAl11O18 grains (Fig. 3(b)). As a result, the fracture toughness of the Zr4+/La3+composite system would be tremendously improved [24]. A schematic of the interaction between the propagating crack and the LaAl11O18 grain in the presence of highly localized tensile or compressive stresses at LaAl11O18 grain surface is shown in Fig. 4(b). It can be noted that since LaAl11O18 grain are placed in “hoop-compression”, the crack would be attracted to pass through them, and thus result in transgranular fracture. Due to that more fracture work would be drastically consumed when transgranular fracture occurs, the fracture toughness of the Zr4+/La3+ system would be thus drastically enhanced. Therefore, the effect of residual stresses on the fracture toughness was confirmed by satisfactory agreement between the theoretical calculation and the crack path characteristics determined experimentally above.

To explore the contribution of residual stress on toughening of ZTA composites, an empirical equation proposed by Ramachandran and Shetty [25] was introduced to describe the measured toughness enhanced behaviors,

Where *K*∞, *K*0 and λ are adjusted parameters. In the case of the present ZTA and co-doped ZTA, *K*0 is the intrinsic fracture resistance defined as the fracture toughness just to start crack extension in the absence of shielding. *K*∞ is the steady-state fracture toughness when a steady-state process zone has been developed behind the crack tip, and λ is a fitting constant while c is the crack length (the sum of a and l.). It can be seen that experimental data can be fit well by Eq. (4). And the values of *K*0 and *K*∞ were 5.53 MPa·m1/2 and 17.76 MPa·m1/2, respectively, which indicate that residual stress contributes a lot to toughening. That is to say, the measured fracture toughness values under the different 4 loads could be fitted as a rising R-curve behavior with the steady-state fracture toughness of 17.76 MPa·m1/2. The enormous enhancement of the crack resistance versus the crack size could be attributed to thermal expansion misfit and flake-like LaAl11O18 in the Zr4+/La3+ system.