C0.3N0.7Ti-SiC Toughed Silicon Nitride Hybrids with Non-Oxide Additives Ti3SiC2

In situ grown C0.3N0.7Ti and SiC, which derived from non-oxide additives Ti3SiC2, are proposed to densify silicon nitride (Si3N4) ceramics with enhanced mechanical performance via hot-press sintering. Remarkable increase of density from 79.20% to 95.48% could be achieved for Si3N4 ceramics with 5 vol.% Ti3SiC2 when sintered at 1600 °C. As expected, higher sintering temperature 1700 °C could further promote densification of Si3N4 ceramics filled with Ti3SiC2. The capillarity of decomposed Si from Ti3SiC2, and in situ reaction between nonstoichiometric TiCx and Si3N4 were believed to be responsible for densification of Si3N4 ceramics. An obvious enhancement of flexural strength and fracture toughness for Si3N4 with x vol.% Ti3SiC2 (x = 1~20) ceramics was observed. The maximum flexural strength of 795 MPa for Si3N4 composites with 5 vol.% Ti3SiC2 and maximum fracture toughness of 6.97 MPa·m1/2 for Si3N4 composites with 20 vol.% Ti3SiC2 are achieved via hot-press sintering at 1700 °C. Pull out of elongated Si3N4 grains, crack bridging, crack branching and crack deflection were demonstrated to dominate enhance fracture toughness of Si3N4 composites.

However, due to the high degree of covalent bonding, Si 3 N 4 -based ceramics are very difficult to densify through the solid-state sintering process. Therefore, effective approaches to ensure rapid consolidation and high mechanical performance of Si 3 N 4 -based ceramics are actively being explored, including gas pressure sintering (GPS) [11], hot-pressing sintering (HPS) [20][21][22][23][24][25][26], hot isostatic pressing sintering (HIP) [27], spark plasma sintering (SPS) [26,28,29], and microwave sintering [1,30], etc. However, considering the requirement of high gas pressures for gas pressure sintering and extra current devices for SPS with a significantly higher furnace costs, HPS allows the dense and complex-shaped parts with medium cost [2]. Previous considerable efforts have demonstrated that fully dense Si 3 N 4 ceramics with superior strength could be achieved through liquid phase sintering by (d 50 = 5 µm, purity > 98%) were kindly provided by Forsman Scientific Co., Ltd., Beijing, China. To investigate the effect of Ti 3 SiC 2 content on the mechanical properties, experiments were conducted with various amounts of Ti 3 SiC 2 powders (1 to 20 vol.%) embedded in α-Si 3 N 4 powders. To ensure the homogeneity of the mixed powders, α-Si 3 N 4 with x vol.% Ti 3 SiC 2 powders (x = 1~20) were wet ball-milled for 10 h by using ethanol as ball-milling media. The substance was dried at 80 • C, and sieved with a filter with a mesh size of 63 µm, then placed in a graphite die coated with BN powder to avoid reaction between the powder and graphite die. Hot-press (HP) sintering was performed on vacuum hot press sintering furnace (ZT-63-21Y, Shanghai Chenhua Technology Co., Ltd., Shanghai, China) with ramp of 10 • C /min 1600 • C and 1700 • C for 90 min in flowing nitrogen under 30 MPa uniaxial pressure during the whole cycle. After natural cooling to room temperature inside furnace, samples were polished and ultrasonic cleaned before characterization. For comparison, α-Si 3 N 4 powders with 2 wt.% Alumina (Al 2 O 3 , AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and 5 wt.% yttria (Y 2 O 3 , AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) were hot-pressed at the same sintering condition.

Characterizations
Before microstructure and mechanical performance characterizations, all hot-pressed samples with diameter of 50 mm were cut into bars and cuboids with help of inside diameter slicer. To reduce surface roughness of samples and guarantee sufficient experimental precision, the abrasive SiC papers with grit size of P320, P600, P1200 and P2000 were used in chronological order during the polishing process. Final manual polishing was carried out with polishing cloth containing alumina suspension with particle size 0.3 µm. The bulk density of each sample was determined according to the Archimedes principle in distilled water. X-ray Diffraction (XRD) patterns were recorded on X'pert PRO (PANalytical B.V., Almelo, Netherlands). Phase identification and quantitative analysis were performed on MDI Jade software (version 6.0, MDI, Livermore, CA, USA) according to Rietveld method. The microstructures of polished surfaces and fracture surfaces were observed using scanning electron microscopy (SEM, Nova NanoSEM 230, FEI Company, Hillsboro, OR, USA) with an energy dispersive X-ray (EDX) analyzer. The Vickers hardness was performed on micro hardness tester (VTD 512, Beijing Weiwei Technology Co., LTD, Beijing, China) under load of 9.8 N with a dwell time of 10 s, and determined by the Vickers diamond indentation method using the following equation: where P is the indentation load on the polished surface and d is the average diagonal length of the Vickers indentation. For accuracy, 11 Vickers indentations on each specimen were applied. After indentation, the microstructures were immediately observed by optical microscopy (ECLISPE LV150N, Tokyo, Japan). As a simple way of estimating toughness, indentation techniques were applied from observed corner cracks and calculated Vickers hardness using the Anstis equation: where E is the Young's modulus and c is the half-length of cracks formed by the indentation.  Figure 1 illustrates the density of Ti 3 SiC 2 filled Si 3 N 4 ceramics as a function of Ti 3 SiC 2 volume fraction. For Si 3 N 4 ceramics which HP sintered at 1600 • C without aids, the density is only 2.58 g·cm −3 . Partial densification may be attributed to the residual SiO 2 liquid phase during firing at high-temperature which always present on Si 3 N 4 powder particles. A remarkable increase to 3.11 g·cm −3 could be observed for Si 3 N 4 ceramics filled with only 5 vol.% Ti 3 SiC 2 when sintered at the same temperature. The enhanced density is even more noticeable than the Si 3 N 4 ceramics with 7 wt.% Y 2 O 3 -Al 2 O 3 aids. These observed results demonstrate Ti 3 SiC 2 to be a effective sintering aid to densify Si 3 N 4 ceramics. However, further increase in Ti 3 SiC 2 content dose not bring any appreciable consolidation.

Results and Discussion
Materials 2020, 13,1428 4 of 13 g·cm −3 could be observed for Si3N4 ceramics filled with only 5 vol.% Ti3SiC2 when sintered at the same temperature. The enhanced density is even more noticeable than the Si3N4 ceramics with 7 wt.% Y2O3-Al2O3 aids. These observed results demonstrate Ti3SiC2 to be a effective sintering aid to densify Si3N4 ceramics. However, further increase in Ti3SiC2 content dose not bring any appreciable consolidation. As expected, higher sintering temperature 1700 °C could further promote densification of Si3N4 ceramics filled with Ti3SiC2. Besides, experimental points are inclined to distribute on a straight line with coefficient of determination (R 2 ) above 0.99. To investigate the composition evolution after sintering, XRD patterns of Si3N4 ceramics filled with different volume fraction of Ti3SiC2 sintered at 1600 and 1700 °C, as well as 7 wt.% (Al2O3-Y2O3) densified Si3N4 ceramics are shown in Figure 2. As seen in Figure 2a, both α and β phase of Si3N4 could be detected when sintering temperature is 1600 °C, which suggests only a partial transformation of α phase to the more stable β phase. In contrast, when further improving sintering temperature to 1700 °C, all diffraction peaks of α-Si3N4 phase disappear (see in Figure 2b). This completely transformation of α to β-Si3N4 phase is believed to be essential to the enhancement of densification and mechanical performance [1,52]. Another important feature should be noted here is that the characteristic diffraction peaks of the raw Ti3SiC2 powder nearly disappear completely after sintering. This could be ascribed to the fact As expected, higher sintering temperature 1700 • C could further promote densification of Si 3 N 4 ceramics filled with Ti 3 SiC 2 . Besides, experimental points are inclined to distribute on a straight line with coefficient of determination (R 2 ) above 0.99. To investigate the composition evolution after sintering, XRD patterns of Si 3 N 4 ceramics filled with different volume fraction of Ti 3 SiC 2 sintered at 1600 and 1700 • C, as well as 7 wt.% (Al 2 O 3 -Y 2 O 3 ) densified Si 3 N 4 ceramics are shown in Figure 2. As seen in Figure 2a, both α and β phase of Si 3 N 4 could be detected when sintering temperature is 1600 • C, which suggests only a partial transformation of α phase to the more stable β phase. In contrast, when further improving sintering temperature to 1700 • C, all diffraction peaks of α-Si 3 N 4 phase disappear (see in Figure 2b). This completely transformation of α to β-Si 3 N 4 phase is believed to be essential to the enhancement of densification and mechanical performance [1,52].
Another important feature should be noted here is that the characteristic diffraction peaks of the raw Ti 3 SiC 2 powder nearly disappear completely after sintering. This could be ascribed to the fact that Ti 3 SiC 2 powder is thermal stable up to~800 • C, and the following reaction can be responsible for the decomposition of Ti 3 SiC 2 [53]: where the value of x ranges from 0.6 to 0.8 and y ≤ 1. Besides, the TiC x phase appears to result in more rapid deterioration of the Ti 3 SiC 2 phase. Also noted that the decomposition usually accomplished with decomposition of Ti 3 SiC 2 to form nonstoichiometric TiC x and gaseous Si, as demonstrated previously [54]: 1600 and 1700 °C, as well as 7 wt.% (Al2O3-Y2O3) densified Si3N4 ceramics are shown in Figure 2. As seen in Figure 2a, both α and β phase of Si3N4 could be detected when sintering temperature is 1600 °C, which suggests only a partial transformation of α phase to the more stable β phase. In contrast, when further improving sintering temperature to 1700 °C, all diffraction peaks of α-Si3N4 phase disappear (see in Figure 2b). This completely transformation of α to β-Si3N4 phase is believed to be essential to the enhancement of densification and mechanical performance [1,52]. Another important feature should be noted here is that the characteristic diffraction peaks of the raw Ti3SiC2 powder nearly disappear completely after sintering. This could be ascribed to the fact that Ti3SiC2 powder is thermal stable up to ~800 °C, and the following reaction can be responsible for the decomposition of Ti3SiC2 [53]: The Si is believed to be act as lubricating phase between Si 3 N 4 grains to promote densification of Si 3 N 4 ceramics through capillarity. Meanwhile, nitriding of TiC x which originated from interatomic diffusion of C and N [10] during high-temperature sintering process would lead to the formation of new phase C 0.3 N 0.7 Ti. In addition, the residual Si would further react with nitrogen to form Si 3 N 4 [55][56][57]. On the other hand, further heating during insulation stage will result in the likely loss of gaseous silicon. Therefore, it is reasonable to assume that the decomposition products of Ti 3 SiC 2 would further react with Si 3 N 4 through diffusion of C and N according to the following reactions: or It is reasonable to claim that the in situ reaction is responsible for the additional characteristic diffraction peaks of C 0.3 N 0.7 Ti and SiC in XRD patterns.
Furthermore, detailed refinement parameters by means of Rietveld method are summarized in Table 1. Detailed refinement parameters, including weight fraction and R factor, of Ti 3 SiC 2 doped Si 3 N 4 ceramics are illustrated in Figures S1-S6 (see in Supplementary Materials). Obviously, the weight fraction of both C 0.3 N 0.7 Ti (Fm-3m, PDF# 42-1448) and moissanite-3C SiC (F-43m, PDF# 29-1129) are inclined to increase with Ti 3 SiC 2 content, which in turn confirms the in situ densification sintering mechanism discussed above. In addition, theoretical density could be derived with help of Rietveld method. Results have shown that nearly full densification for Ti 3 SiC 2 doped Si 3 N 4 ceramics sintered at 1700 • C could be achieved.  Figure 3 shows the micromorphology of polished surface of Si 3 N 4 with different volume fractions of Ti 3 SiC 2 sintered at 1700 • C. Due to the lack of sufficient sintering aids, lots of pores could be observed, and grain growth of β-Si 3 N 4 is not complete for monolithic Si 3 N 4 ceramic (see Figure 3a). However, the microstructures of Ti 3 SiC 2 -Si 3 N 4 ceramics (see Figure 3b-g) exhibit much more close-grain structure and consist of randomly oriented elongated Si 3 N 4 grains which is accordant with XRD results in Figure 2. The average diameters of grains present slight increasing trend from 0.68 to 0.98 µm by quantitative image analysis as the amount of Ti 3 SiC 2 increased. Besides, the bright contrasted phase which uniformly embedded in Si 3 N 4 matrix could be observed and are inclined to aggregate especially when Ti 3 SiC 2 content exceeds 15 vol.%. Furthermore, as shown in Table 2, energy dispersive spectrometer (EDS) at spot A in Figure 3e suggests dominant phase of Si 3 N 4 and SiC, which is associated with reaction described by Equation (10). Additional O element may be originated from surface of raw α-Si 3 N 4 powders. Meanwhile, the bright region at spot B is proved to be enriched by Ti according to the EDS results in Table 2. Combined with the results of XRD analysis, it is reasonable to claim that the dispersive bright regions consist of C 0.3 N 0.7 Ti and SiC, which are believed to affect the mechanical performance of reaction bonded Si 3 N 4 ceramics.
Materials 2020, 13,1428 6 of 13  Figure 3 shows the micromorphology of polished surface of Si3N4 with different volume fractions of Ti3SiC2 sintered at 1700 °C. Due to the lack of sufficient sintering aids, lots of pores could be observed, and grain growth of β-Si3N4 is not complete for monolithic Si3N4 ceramic (see Figure  3a). However, the microstructures of Ti3SiC2-Si3N4 ceramics (see Figure 3b-g) exhibit much more close-grain structure and consist of randomly oriented elongated Si3N4 grains which is accordant with XRD results in Figure 2. The average diameters of grains present slight increasing trend from 0.68 to 0.98 μm by quantitative image analysis as the amount of Ti3SiC2 increased. Besides, the bright contrasted phase which uniformly embedded in Si3N4 matrix could be observed and are inclined to aggregate especially when Ti3SiC2 content exceeds 15 vol.%. Furthermore, as shown in Table 2, energy dispersive spectrometer (EDS) at spot A in Figure 3e suggests dominant phase of Si3N4 and SiC, which is associated with reaction described by Equation (10). Additional O element may be originated from surface of raw α-Si3N4 powders. Meanwhile, the bright region at spot B is proved to be enriched by Ti according to the EDS results in Table 2. Combined with the results of XRD analysis, it is reasonable to claim that the dispersive bright regions consist of C0.3N0.7Ti and SiC, which are believed to affect the mechanical performance of reaction bonded Si3N4 ceramics.    The mechanical properties, including Vickers hardness, flexural strength, and fracture toughness, of dense Si 3 N 4 ceramics with different Ti 3 SiC 2 content sintered at 1700 • C are illustrated in Figure 4.
Clearly, the Vickers hardness of Si 3 N 4 ceramics has been upgraded after modification of Ti 3 SiC 2 , and presents slight increase compared with that of Si 3 N 4 ceramics containing conventional oxides aids. Besides, an obvious enhancement of flexural strength and fracture toughness could be observed. A maximum flexural strength of 795 MPa could be achieved for 5 vol.% Ti 3 SiC 2 doped Si 3 N 4 composites, which is almost twice that of 7 wt.% (Y 2 O 3 -Al 2 O 3 )-Si 3 N 4 ceramics prepared at the same condition. This enhancement of flexural strength could be attributed to the C 0.3 N 0.7 Ti and SiC which originated from reaction bonding between Ti 3 SiC 2 and Si 3 N 4 [10]. However, further increment of Ti 3 SiC 2 content reduces the flexural strength of Si 3 N 4 ceramics which may be ascribed to the enhanced residual stresses around grain boundary [58,59]. Please note that this residual stress is believed to result in microcracks and intergranular fracture mode which will be discussed later. Moreover, the fracture toughness of Si 3 N 4 composites is also effectively boosted after Ti 3 SiC 2 decoration, and reaches maximum value of 6.97 MPa·m 1/2 for 20 vol.% Ti 3 SiC 2 -Si 3 N 4 ceramics which is 37% higher than that of 7 wt.% (Y 2 O 3 -Al 2 O 3 )-Si 3 N 4 ceramics.

of 13
The mechanical properties, including Vickers hardness, flexural strength, and fracture toughness, of dense Si3N4 ceramics with different Ti3SiC2 content sintered at 1700 °C are illustrated in Figure 4. Clearly, the Vickers hardness of Si3N4 ceramics has been upgraded after modification of Ti3SiC2, and presents slight increase compared with that of Si3N4 ceramics containing conventional oxides aids. Besides, an obvious enhancement of flexural strength and fracture toughness could be observed. A maximum flexural strength of 795 MPa could be achieved for 5 vol.% Ti3SiC2 doped Si3N4 composites, which is almost twice that of 7 wt.% (Y2O3-Al2O3)-Si3N4 ceramics prepared at the same condition. This enhancement of flexural strength could be attributed to the C0.3N0.7Ti and SiC which originated from reaction bonding between Ti3SiC2 and Si3N4 [10]. However, further increment of Ti3SiC2 content reduces the flexural strength of Si3N4 ceramics which may be ascribed to the enhanced residual stresses around grain boundary [58,59]. Please note that this residual stress is believed to result in microcracks and intergranular fracture mode which will be discussed later. Moreover, the fracture toughness of Si3N4 composites is also effectively boosted after Ti3SiC2 decoration, and reaches maximum value of 6.97 MPa·m 1/2 for 20 vol.% Ti3SiC2-Si3N4 ceramics which is 37% higher than that of 7 wt.% (Y2O3-Al2O3)-Si3N4 ceramics.  Figure 5 illustrates the typical optical micrographs of the Vickers hardness indents and the induced cracks of Si3N4 ceramics with different Ti3SiC2 contents, as well as 7 wt.% (Y2O3-Al2O3). Clearly, the polished surfaces of Ti3SiC2 doped Si3N4 ceramics become much smoother than the monolithic Si3N4 ceramic which HP sintered at 1700 °C, corresponding to the enhancement of densification. Besides, it can be seen that the area of indentation presents no obvious change for Ti3SiC2 doped Si3N4 ceramics, which is consistent with the stable Vickers hardness. However, the cracks obviously become shorter especially when the Ti3SiC2 contents exceed 10 vol.%, which is responsible for the enhancement of fracture toughness.  . Clearly, the polished surfaces of Ti 3 SiC 2 doped Si 3 N 4 ceramics become much smoother than the monolithic Si 3 N 4 ceramic which HP sintered at 1700 • C, corresponding to the enhancement of densification. Besides, it can be seen that the area of indentation presents no obvious change for Ti 3 SiC 2 doped Si 3 N 4 ceramics, which is consistent with the stable Vickers hardness. However, the cracks obviously become shorter especially when the Ti 3 SiC 2 contents exceed 10 vol.%, which is responsible for the enhancement of fracture toughness.
To illustrate the fracture behaviors and activated toughening mechanisms, micromorphology and crack paths are investigated on cross-sectional fracture surfaces and polished surfaces, respectively. Comparison of typical fracture surfaces between Si 3 N 4 doped with Al 2 O 3 -Y 2 O 3 and Ti 3 SiC 2 is illustrated in Figure 6. As can be seen from Figure 6a, a small number of pores occur in the Si 3 N 4 -7 wt.% (Al 2 O 3 -Y 2 O 3 ) composites, which is harmful for the mechanical performance. In contrast, the Si 3 N 4 -Ti 3 SiC 2 specimen presents a much more close-grain fracture surface owning to the higher density. As marked by red arrows in Figure 6b, large amounts of dimples corresponding to the transgranular fracture could be observed. In addition, this fracture mode is considered to make a dominant contribution to the superior flexural strength of Si 3 N 4 -Ti 3 SiC 2 composites [10]. Besides, as marked by yellow arrows, lots of interface debonding between the Si 3 N 4 grains and grain boundary Materials 2020, 13, 1428 8 of 13 phase could be observed. This intergranular fracture mode may result from the pullout of elongated β-Si 3 N 4 grains, which is believed to make a contribution to the enhancement of overall fracture toughness [60][61][62].
Materials 2020, 13,1428 8 of 13 To illustrate the fracture behaviors and activated toughening mechanisms, micromorphology and crack paths are investigated on cross-sectional fracture surfaces and polished surfaces, respectively. Comparison of typical fracture surfaces between Si3N4 doped with Al2O3-Y2O3 and Ti3SiC2 is illustrated in Figure 6. As can be seen from Figure 6a, a small number of pores occur in the Si3N4-7 wt.% (Al2O3-Y2O3) composites, which is harmful for the mechanical performance. In contrast, the Si3N4-Ti3SiC2 specimen presents a much more close-grain fracture surface owning to the higher density. As marked by red arrows in Figure 6b, large amounts of dimples corresponding to the transgranular fracture could be observed. In addition, this fracture mode is considered to make a dominant contribution to the superior flexural strength of Si3N4-Ti3SiC2 composites [10]. Besides, as marked by yellow arrows, lots of interface debonding between the Si3N4 grains and grain boundary phase could be observed. This intergranular fracture mode may result from the pullout of elongated β-Si3N4 grains, which is believed to make a contribution to the enhancement of overall fracture toughness [60][61][62]. Another mechanism of the enhanced fracture toughness of Si3N4-Ti3SiC2 composites could be ascribed to the crack branching, deflection, and grain bridging by in situ derived C0.3N0.7Ti and SiC grains embedded in Si3N4 matrix, which illustrated in Figure 7. Due to the superior hardness of C0.3N0.7Ti and the thermal mismatch between Si3N4 and C0.3N0.7Ti, there exists residual stress around C0.3N0.7Ti grains during cooling, giving rise to the microcracks inside composites. When subjected to the external mechanical stress, these microcracks tend to be activated and the propagation path of cracks tends to be split by C0.3N0.7Ti hard-phase and deflected along the interface. Such mechanisms consumed more fracture energy during the crack propagation which leads to crack arrest [63][64][65][66][67].
A comparison of mechanical properties of Si3N4-based ceramics obtained in the present work and selected previous works with conventional oxide aids is shown in Table 3. Clearly, the Vickers, flexural strength, and toughness of Ti3SiC2 doped Si3N4 ceramics present the same level or even better compared with Si3N4 ceramics sintered with oxide aids. Moreover, due to the superior  To illustrate the fracture behaviors and activated toughening mechanisms, micromorphology and crack paths are investigated on cross-sectional fracture surfaces and polished surfaces, respectively. Comparison of typical fracture surfaces between Si3N4 doped with Al2O3-Y2O3 and Ti3SiC2 is illustrated in Figure 6. As can be seen from Figure 6a, a small number of pores occur in the Si3N4-7 wt.% (Al2O3-Y2O3) composites, which is harmful for the mechanical performance. In contrast, the Si3N4-Ti3SiC2 specimen presents a much more close-grain fracture surface owning to the higher density. As marked by red arrows in Figure 6b, large amounts of dimples corresponding to the transgranular fracture could be observed. In addition, this fracture mode is considered to make a dominant contribution to the superior flexural strength of Si3N4-Ti3SiC2 composites [10]. Besides, as marked by yellow arrows, lots of interface debonding between the Si3N4 grains and grain boundary phase could be observed. This intergranular fracture mode may result from the pullout of elongated β-Si3N4 grains, which is believed to make a contribution to the enhancement of overall fracture toughness [60][61][62]. Another mechanism of the enhanced fracture toughness of Si3N4-Ti3SiC2 composites could be ascribed to the crack branching, deflection, and grain bridging by in situ derived C0.3N0.7Ti and SiC grains embedded in Si3N4 matrix, which illustrated in Figure 7. Due to the superior hardness of C0.3N0.7Ti and the thermal mismatch between Si3N4 and C0.3N0.7Ti, there exists residual stress around C0.3N0.7Ti grains during cooling, giving rise to the microcracks inside composites. When subjected to the external mechanical stress, these microcracks tend to be activated and the propagation path of cracks tends to be split by C0.3N0.7Ti hard-phase and deflected along the interface. Such mechanisms consumed more fracture energy during the crack propagation which leads to crack arrest [63][64][65][66][67].
A comparison of mechanical properties of Si3N4-based ceramics obtained in the present work and selected previous works with conventional oxide aids is shown in Table 3. Clearly, the Vickers, flexural strength, and toughness of Ti3SiC2 doped Si3N4 ceramics present the same level or even better compared with Si3N4 ceramics sintered with oxide aids. Moreover, due to the superior Another mechanism of the enhanced fracture toughness of Si 3 N 4 -Ti 3 SiC 2 composites could be ascribed to the crack branching, deflection, and grain bridging by in situ derived C 0.3 N 0.7 Ti and SiC grains embedded in Si 3 N 4 matrix, which illustrated in Figure 7. Due to the superior hardness of C 0.3 N 0.7 Ti and the thermal mismatch between Si 3 N 4 and C 0.3 N 0.7 Ti, there exists residual stress around C 0.3 N 0.7 Ti grains during cooling, giving rise to the microcracks inside composites. When subjected to the external mechanical stress, these microcracks tend to be activated and the propagation path of cracks tends to be split by C 0.3 N 0.7 Ti hard-phase and deflected along the interface. Such mechanisms consumed more fracture energy during the crack propagation which leads to crack arrest [63][64][65][66][67].
A comparison of mechanical properties of Si 3 N 4 -based ceramics obtained in the present work and selected previous works with conventional oxide aids is shown in Table 3. Clearly, the Vickers, flexural strength, and toughness of Ti 3 SiC 2 doped Si 3 N 4 ceramics present the same level or even better compared with Si 3 N 4 ceramics sintered with oxide aids. Moreover, due to the superior mechanical and thermal properties of in situ formed C 0.3 N 0.7 Ti and SiC, Si 3 N 4 ceramics obtained in this work are believed to have a significant competitive advantage and to promote the development of Si 3 N 4 -based ceramics at high temperatures. mechanical and thermal properties of in situ formed C0.3N0.7Ti and SiC, Si3N4 ceramics obtained in this work are believed to have a significant competitive advantage and to promote the development of Si3N4-based ceramics at high temperatures.

Conclusions
In summary, we proposed non-oxide Ti3SiC2 (one of typical MAX cermets) as a novel sintering aid to densify Si3N4 ceramics with enhanced mechanical properties. A remarkable relative density increment of 20.5% (from 2.58 to 3.11 g·cm −3 ) could be observed for 1600 °C hot-press sintered Si3N4 ceramics doped with only 5 vol.% Ti3SiC2 compared with Si3N4 ceramics without aids. Further increase in sintering temperature to 1700 °C brought appreciable consolidation of nearly full dense

Conclusions
In summary, we proposed non-oxide Ti 3 SiC 2 (one of typical MAX cermets) as a novel sintering aid to densify Si 3 N 4 ceramics with enhanced mechanical properties. A remarkable relative density increment of 20.5% (from 2.58 to 3.11 g·cm −3 ) could be observed for 1600 • C hot-press sintered Si 3 N 4 ceramics doped with only 5 vol.% Ti 3 SiC 2 compared with Si 3 N 4 ceramics without aids. Further increase in sintering temperature to 1700 • C brought appreciable consolidation of nearly full dense Ti 3 SiC 2 -Si 3 N 4 ceramics. XRD and EDS investigations demonstrated the formation of C 0.3 N 0.7 Ti and SiC which resulted from in situ reaction between Ti 3 SiC 2 and Si 3 N 4 through diffusion of C and N. The Vickers hardness of Ti 3 SiC 2 doped Si 3 N 4 ceramics increased slight compared with that of Si 3 N 4 ceramics containing conventional oxides aids. Nevertheless, an obvious enhancement of flexural strength and fracture toughness could be observed. A maximum flexural strength of 795 MPa could be obtained for 5 vol.% Ti 3 SiC 2 doped Si 3 N 4 composites. Moreover, the fracture toughness of Ti 3 SiC 2 densified Si 3 N 4 composites exhibited a remarkable increase with increasing in volume fraction, and reached maximum value of 6.97 MPa·m 1/2 for 20 vol.% Ti 3 SiC 2 -Si 3 N 4 ceramics. Pull out of elongated Si 3 N 4 grains, crack bridging and deflection were demonstrated to promote fracture toughness of Ti 3 SiC 2 densified Si 3 N 4 composites. With these successes, MAX phase densified Si 3 N 4 ceramics with enhanced strength and toughness will be necessary to meet demands of potential future markets for advanced ceramics. Further efforts are encouraged to be devoted to thermal properties investigations of MAX enabled Si 3 N 4 composites.