Enhancement of sinterability and mechanical properties of B₄C ceramics using Ti₃AlC₂ as a sintering aid

Abstract

Boron carbide (B₄C) ceramics are currently the leading control rod materials in fast nuclear reactors and promising high temperature structural materials. However, several drawbacks such as poor sinterability and low toughness seriously limit their wide applications. In order to enhance the sinterability and mechanical properties of B₄C ceramics, titanium aluminum carbide (Ti₃AlC₂) was chosen as a new efficient sintering aid to densify B₄C ceramics using spark plasma sintering. Fully dense B₄C ceramics were obtained at a lower sintering temperature of 1500 °C by adding a small amount of Ti₃AlC₂. Meanwhile, the mechanical properties were enhanced remarkably. For B₄C ceramics sintered with 15 vol% Ti₃AlC₂, optimized mechanical properties were obtained with Vickers hardness of 40.2 GPa and indentation fracture toughness of 4.7 MPa m^1/2. These results indicate that Ti₃AlC₂ can be used as a novel sintering aid for the densification of B₄C ceramics.

---

Introduction

Boron carbide (B₄C) ceramics are excellent structural materials due to their low density (2.52 g cm⁻³), high Young's modulus (∼440 GPa) and high Vickers hardness (∼30 GPa),^1,2 which make B₄C ceramics excellent candidates for light-weight body armors. B₄C ceramics also have great potential for high temperature applications in non-oxidizing environments owing to their high melting point and good chemical stability. Besides, B₄C ceramics are important functional materials due to the neutron absorption ability of isotope ¹⁰B which owns a high neutron absorption cross-section over a wide range of neutron energies. Because of their high B content, good chemical inertness and good mechanical properties, B₄C ceramics are extensively used in the nuclear reactors as control rods, shielding materials and as neutron detectors. Currently, they are the leading control rod materials in liquid-metal-cooled fast breeder reactors (LMFBR) and boiling water reactors (BWR).³

However, there exist several major restrictions for the wide applications of B₄C ceramics as both structural and functional materials: (a) poor sinterability due to the highly covalent bonding and the associated low diffusion mobility; (b) low machinability due to high hardness and low electrical conductivity; (c) relatively low strength and toughness.¹ Further, as control rod materials in nuclear reactors, B₄C ceramics often experience serious cracking when subjected to thermal stresses and helium gas due to their intrinsic low thermal conductivity, low toughness and poor thermal shock resistance.⁴

The poor sinterability makes the preparation and processing of B₄C ceramics rather complicated and energy-consuming. The relatively low toughness and low thermal conductivity of B₄C ceramics could influence the lifetime of control rods and raise safety problems to the nuclear reactors.⁵ Thus it is essential to lower the sintering temperature and enhance the mechanical properties of B₄C ceramics for wider applications and the extension of lifetime of control rods in nuclear reactors. Many efforts have been paid to lower the sintering temperature of B₄C ceramics and improve their mechanical properties. For the reduction of sintering temperature, carbon (C),^6–8 metallic phase such as aluminium (Al),^9–11 silicon (Si),^12,13 and titanium (Ti),¹⁴ oxides such as alumina (Al₂O₃) and rare earth oxides,^15,16 as well as borides¹⁷ were commonly chosen as sintering aids. C is effective in reducing the sintering temperature of pressureless sintered B₄C because C can remove the oxide layer of B₄C particles and enhance the diffusion at the grain boundaries.^6–8 The other sintering aids mentioned above generally form liquid phase due to their lower melting points and thus promote densification of B₄C ceramics significantly.^9–17 Although these sintering aids can facilitate sintering of B₄C ceramics at a much lower temperature, in most cases, no remarkable improvements on mechanical properties of B₄C ceramics were achieved. Generally, the residual carbon or metallic phases at the grain boundaries generally deteriorate the unique properties of hard ceramics. Moreover, abnormal grain growth caused by liquid phase during sintering could lead to lower toughness of ceramics. In order to increase the strength and toughness of B₄C ceramics, several additives have been explored such as TiB₂, ZrB₂ and SiC,^11,18–22 among which TiB₂ is the most extensively studied. The B₄C–TiB₂ composites can be obtained by direct mixing of B₄C and TiB₂,¹¹ or by in situ reaction sintering between B₄C and TiC or TiO₂ and C.^18–20 Besides, carbon nanotube (CNT)⁹ and carbon nanofiber (CNF)²³ reinforced B₄C ceramics were also reported. The introduction of TiB₂, CNT or CNF could result in a remarkable increase of the mechanical properties of B₄C ceramics. Generally, B₄C–TiB₂ composites show an improved fracture toughness in the range of 3–6 MPa m^1/2 depending on the amount of TiB₂ and different sintering methods.^11,18–20 B₄C–CNF composite was reported to exhibit an extremely high fracture toughness of 7.5 MPa m^1/2 when the volume fraction of CNF was 15%.²³ However, the enhancement of sinterability for B₄C ceramics is quite limited. The sintering temperature of B₄C–TiB₂ composites still needs to be around 2000 °C and above by reaction hot-pressing of B₄C, TiO₂ and C powders.¹⁹

In order to enhance the sinterability and mechanical properties of B₄C ceramics simultaneously, titanium aluminum carbide (Ti₃AlC₂), a typical member of the MAX phases, was chosen as a sintering aid to prepare dense B₄C ceramics in this study. The densification was carried out by spark plasma sintering which is an effective way to fabricate dense nanostructured ceramics. The influence of Ti₃AlC₂ content on the sinterability and mechanical properties of B₄C ceramics was systematically studied.

Experimental

B₄C powders (Grade HS, H. C. Starck GmbH, Germany) and Ti₃AlC₂ powders (>99%, Forsman, China) were used as starting materials. Fig. 1 shows the microstructures of these two raw materials. The B₄C powder has a uniform grain size distribution with an average particle size of 0.5 μm and the Ti₃AlC₂ powder used in this study shows a broad grain size distribution from 0.5 μm to about 10 μm, as shown in Fig. 1. B₄C and Ti₃AlC₂ powders were mixed according to the formula of B₄C-x vol% Ti₃AlC₂ with x = 0, 5, 10, 15 and 20. The mixed powders were ball-milled for 10 h with SiO₂ balls using ethanol as a ball-milling media. After ball-milling, the mixed powders were dried and then put into a carbon die with an inner diameter of 15 mm. B₄C-x vol% Ti₃AlC₂ mixtures were sintered in vacuum using an SPS furnace (Labox-325, Japan) for 3 minutes at different temperatures. The heating rate is 100 °C min⁻¹ and a uniaxial pressure of 80 MPa was applied during sintering.

Fig. 1 Microstructure of (a) B₄C and (b) Ti₃AlC₂ powders.

The density of the sintered samples was measured by the Archimedes's method. Crystalline phases were characterized by X-ray diffraction (XRD). The microstructure of the specimens was observed using scanning electron microscopy (SEM) with an energy-dispersive X-ray (EDX) analyser. In order to measure the Vickers hardness (H_V) and fracture toughness (K_IC), specimens were first polished with diamond slurry. Hardness was determined by the Vickers diamond indentation method using the following equation:

H_v = 1.854 × 10⁹P/d² (1)

where P is the indentation load (9.8 N in this study) on the polished surface and d is the average diagonal length of the Vickers indentation. Indentation fracture toughness values of the specimens were semiquantitatively estimated from observed corner cracks and calculated Vickers hardness using the Anstis equation:

K_IC = 0.016(E/H_v)^1/2(P/c^3/2) (2)

where E is the Young's modulus and c is the half-length of cracks formed by the indentation (also induced with a 9.8 N load in this study).²⁴

Results and discussion

Fig. 2 shows the density of B₄C ceramics sintered with different amount of Ti₃AlC₂ at 1500 °C. The density of monolithic B₄C ceramics spark plasma sintered at 1500 °C is only 2.02 g cm⁻³, which is about 80.1% of the theoretical density. When 5 vol% Ti₃AlC₂ was added, the density of B₄C increased dramatically to 2.69 g cm⁻³. Further increase in Ti₃AlC₂ content did not bring to any appreciable density increase. 5 vol% of Ti₃AlC₂ was therefore enough to obtain a fully dense B₄C ceramic at 1500 °C. When the sintering temperature was increased to 1600 °C, the relative density of monolithic B₄C ceramics increased to 98.0%. This indicates that the lowest sintering temperature for fully dense B₄C ceramics is at least 100 °C lower when spark plasma sintered using Ti₃AlC₂ as a sintering aid, which suggests that Ti₃AlC₂ is quite effective in lowering the sintering temperature of B₄C ceramics.

Fig. 2 Density of B₄C ceramics with different amount of Ti₃AlC₂ content sintered at 1500 °C.

The phase composition of B₄C ceramics sintered with different amount of Ti₃AlC₂ was examined by XRD analysis and the results are shown in Fig. 3. For monolithic B₄C ceramics, only diffraction peaks of single phase B₄C were detected. B₄C ceramics sintered with Ti₃AlC₂ show several additional diffraction peaks besides those corresponding to B₄C. However, no Ti₃AlC₂ diffraction peaks were observed, demonstrating that Ti₃AlC₂ decomposed completely at high temperature. The existence of TiB₂ can be identified by the diffraction peaks labelled in Fig. 4, and it can be seen that the intensity of the main TiB₂ diffraction peak around 45° increases gradually with the increase of Ti₃AlC₂ content. A broad peak around 26° indicates the existence of C and its intensity also increases with the increase of Ti₃AlC₂ amount. Further, a small peak around 33° which belongs to Al₂OC phase was observed for all B₄C ceramics sintered with Ti₃AlC₂. For B₄C ceramics sintered with 20 vol% Ti₃AlC₂, one more set of additional diffraction peaks was detected. An accurate peak matching revealed that it corresponds to the Al–B–C phase Al₈B₄C₇, which was also observed in previous B₄C ceramics sintered using Al as sintering aid.^7,13 Based on the XRD results and previous studies, it can be summarized that the Ti₃AlC₂ first decomposed according to the following reaction:²⁵

Ti₃AlC₂ → 3TiC_2/3 + Al (3)

Fig. 3 XRD profile of B₄C ceramics sintered at 1500 °C with different amount of Ti₃AlC₂ additive.

Fig. 4 Polished surfaces of B₄C ceramics sintered with different amount of Ti₃AlC₂ content and the related EDS spectra. (a) Monolithic B₄C sintered at 1500 °C; (b) monolithic B₄C sintered at 1600 °C; (c) B₄C-5 vol% Ti₃AlC₂; (d) B₄C-10 vol% Ti₃AlC₂; (e) B₄C-15 vol% Ti₃AlC₂; (f) B₄C-20 vol% Ti₃AlC₂; (g) B₄C-15 vol% Ti₃AlC₂ at higher magnification; (h) EDS spectra of areas shown in (g).

Then TiC_2/3 and Al reacted with B₄C respectively during sintering according to the following reactions:²⁶

3B₄C + 6TiC_2/3 → 6TiB₂ + 7C (4)

8Al + B₄C + 6C → Al₈B₄C₇ (5)

As to the Al₂OC phase, it is suggested that the oxygen may come from the oxide layer of B₄C particles. Generally, the outer surface of B₄C fine particle is covered by an oxide layer of B₂O₃, which is one reason for the low sinterability of B₄C ceramics.^1,7 In this study, it is suggested that Al decomposed from Ti₃AlC₂ and the residual C from the above reactions can remove this oxidized layer and form Al₂OC phase according to the following reactions:

2Al + B₂O₃ → Al₂O₃ + 2B (6)

Al₂O₃ + 3C → Al₂OC + 2CO (7)

which were also reported in previous studies.²⁷ The reaction between B₄C and Ti₃AlC₂ during sintering could also give rise to fine grains of B₄C ceramics because of the lower temperature.

Fig. 4 shows the microstructures observed in the polished surface of B₄C ceramics sintered with different amount of Ti₃AlC₂ and the typical EDS spectra of B₄C ceramics sintered with 15 vol% Ti₃AlC₂. Fig. 4(a) and (b) are polished surfaces of monolithic B₄C ceramics sintered at 1500 °C and 1600 °C respectively; Fig. 4(c)–(f) correspond to B₄C ceramics sintered at 1500 °C with 5, 10, 15 and 20 vol% Ti₃AlC₂ respectively. Monolithic B₄C ceramic sintered at 1500 °C shows a porous microstructure, which is consistent with its low density. When sintered at 1600 °C, B₄C ceramic becomes much denser as shown in Fig. 4(b). All the B₄C ceramics sintered at 1500 °C with Ti₃AlC₂ as sintering aid show quite dense microstructures. With the addition of Ti₃AlC₂, a white contrasted phase uniformly dispersed in the B₄C matrix starts to appear, indicating the presence of second phases. The size of these white spots varies from 0.5 μm to about 10 μm, which is consistent with the size distribution of raw Ti₃AlC₂ powders. The volume fraction of these white spots was estimated from the polished surfaces using the ImageJ software and it is calculated to be 6.2%, 13.8%, 17.4% and 24.9% for B₄C ceramics sintered with 5, 10, 15 and 20 vol% Ti₃AlC₂ respectively. These values are roughly consistent with the volume fractions of the secondary phases proposed by eqn (3)–(5) for B₄C ceramics sintered with different amount of Ti₃AlC₂.

In order to present a better understanding of the distribution of different phases in the B₄C matrix, quantitative analysis of the second phases was done by SEM with EDS analyser on the polished surface of B₄C ceramics sintered with different amount of Ti₃AlC₂. As shown in Fig. 4, the grey area is the main B₄C phase, and the white spots indicated with white arrows are TiB₂ according to the EDS analysis (Fig. 4(h)). Besides the white spots, there also exist some white areas which have no clear boundaries as indicated in the blue circles of Fig. 4(c)–(f). The EDS spectrum demonstrates that they are Al–B–C phases (Fig. 4(h)), and referring to the XRD results they should be Al₈B₄C₇. It was supposed that Al originally existed at the brighter contrasted areas, and then the molten Al diffused into surroundings during sintering. The light grey spots indicated with red arrows in Fig. 4(e) and (f) and area D in Fig. 4(g) are Al–O–C phase, presumably Al₂OC (Fig. 4(h)).

From the above experimental results, it can be summarized that the mechanisms for the enhancement of sinterability of B₄C ceramics using Ti₃AlC₂ as a sintering aid are many-sided. Firstly, Ti₃AlC₂ can decompose into Al and TiC_2/3 at high temperature, which has been confirmed by previous study.²⁵ The Al decomposed from Ti₃AlC₂ can form liquid phase at high temperature and promote sintering effectively. Secondly, Al can remove the oxide layer of B₄C particles. The elimination of oxides layer on the surface of initial B₄C powders allows direct contact between B₄C particles, permitting sintering to initiate at significantly lower temperature,¹ thereby accelerating the densification of B₄C. Thirdly, the in situ reaction sintering between B₄C and TiC_2/3 facilitates mass transfer in the aggregate and the enhancement of mass transport could allow the densification to occur at lower temperatures where the grain growth is still limited. The residual C produced by the reaction between B₄C and TiC_2/3 could further removes the oxide layer. Besides, it has been confirmed that the presence of TiB₂ could result in lowering of activation energy for sintering of B₄C through various machnisms.²⁸ The increased sinterability has been reported in the B₄C–TiB₂ composites through reaction hot-pressing of B₄C and TiC.²⁹ Therefore, the improved sinterability of B₄C ceramics using Ti₃AlC₂ as sintering aid is a consequence of liquid phase sintering, removal of oxide layer and in situ reactive sintering. The reactive sintering and the presence of TiB₂ and C could also inhibit grain growth effectively, thereby giving rise to fine and uniform grain size distributions.

Fig. 5 shows the typical optical micrographs of the Vickers hardness indents and the induced cracks of B₄C ceramics sintered with different amounts of Ti₃AlC₂. Fig. 5(a) and (b) correspond to monolithic B₄C sintered at 1500 °C and 1600 °C respectively, and Fig. 5(b)–(d) correspond to B₄C ceramics sintered at 1500 °C with 5, 10, 15 and 20 vol% Ti₃AlC₂ respectively. It can be seen that all ceramics show clear indentations. Monolithic B₄C ceramics exhibit long and straight cracks induced by the indentation. In case of B₄C ceramics sintered with Ti₃AlC₂, the cracks become shorter and more curved and the indentation becomes smaller.

Fig. 5 Optical micrographs of the Vickers hardness indents and the induced cracks of (a) monolithic B₄C sintered at 1500 °C; (b) monolithic B₄C sintered at 1600 °C; (c) B₄C-5 vol% Ti₃AlC₂; (d) B₄C-10 vol% Ti₃AlC₂; (e) B₄C-15 vol% Ti₃AlC₂; (f) B₄C-20 vol% Ti₃AlC₂.

The Vickers hardness and indentation fracture toughness of all these B₄C ceramics were calculated according to the length of the indentations and the induced cracks, and the results are shown in Fig. 6. The filled symbols represent B₄C ceramics sintered at 1500 °C with different amount of Ti₃AlC₂, while the open symbols represent monolithic B₄C ceramics sintered at 1600 °C. Monolithic B₄C ceramic sintered at 1500 °C shows a low Vickers hardness of 12 GPa due to its high porosity. The Vickers hardness of monolithic B₄C ceramic increases dramatically to 37.6 GPa when sintered at 1600 °C. In case of B₄C ceramics sintered with Ti₃AlC₂, only a slight increase of Vickers hardness is obtained compared to fully dense monolithic B₄C ceramics. With increasing the amount the Ti₃AlC₂, the Vickers harness of B₄C ceramics keeps almost the same around 40 GPa, which is higher than most of the B₄C ceramics and B₄C–TiB₂ composites reported^1,11 and close to the values of superhard B₄C–ZrB₂ ceramics.^30,31 Porosity and grain size are two most important factors that can influence the Vickers hardness of B₄C ceramics. In this study, except the monolithic ceramic sintered at 1500 °C, all the other B₄C ceramics have high density with only negligible porosity. Besides, SPS could suppress the grain growth effectively, especially when using Ti₃AlC₂ as a sintering aid. Thus the low porosity and fine grain size (around 0.5 μm) give rise to high Vickers hardness in this study. Different from the Vickers hardness, the fracture toughness of B₄C ceramics changes dramatically with the increase of Ti₃AlC₂ content. For dense monolithic B₄C ceramics, the indentation fracture toughness is only 2.32 MPa m^1/2, which is in good consistency with previous reports.^1,11 It increases gradually with increasing Ti₃AlC₂ content and a maximum value of 4.7 MPa m^1/2 is achieved for B₄C ceramics sintered with 10 and 15 vol% Ti₃AlC₂. The increase of indentation fracture toughness can be ascribed to following two main factors: a well sintered fine-grained B₄C matrix and uniformly distributed TiB₂ particles. The different thermal expansion coefficients of TiB₂ (α = 8.7 × 10⁻⁶ K⁻¹) and B₄C (α = 5.7 × 10⁻⁶ K⁻¹) generate residual stress around TiB₂ particles during cooling and this residual stress gives rise to the microcracks and thus the crack deflection.²⁰ In this study, the propagation of cracks was hindered by the dispersion of TiB₂ particles and the higher fracture toughness in B₄C ceramic sintered with 10 and 15 vol% Ti₃AlC₂ can be explained by the crack deflection by TiB₂ particles, as demonstrated in Fig. 7. With further increase of Ti₃AlC₂ content to 20 vol%, the fracture toughness of B₄C ceramics decreases slightly. The increase of Al-rich phases is one possible reason. For B₄C ceramics sintered with 20 vol% Ti₃AlC₂, some aggregation of TiB₂ particles are shown in Fig. 4(f), and the aggregation of TiB₂ phases could be responsible for the decrease of fracture toughness.

Fig. 6 Variations of Vickers hardness and fracture toughness of B₄C ceramics with the amount of Ti₃AlC₂ additive, hollow markers refer to monolithic B₄C sintered at 1600 °C.

Fig. 7 Demonstration of crack deflection in B₄C ceramics sintered with (a) 15 vol% and (b) 20 vol% Ti₃AlC₂, the arrows indicate the crack propagation directions.

Conclusions

Fully dense B₄C ceramics were obtained at a lower sintering temperature using Ti₃AlC₂ as a novel sintering aid in this study, meanwhile their mechanical properties were enhanced. The B₄C ceramics sintered with 15 vol% Ti₃AlC₂ exhibits the maximized mechanical properties with H_v = 40.2 GPa and K_IC = 4.7 MPa m^1/2. The mechanisms for the enhancement of sinterability include liquid phase of aluminium decomposed from Ti₃AlC₂ at high temperatures, removal of surface oxide layer B₂O₃ by C and Al and reaction sintering between B₄C and TiC_2/3. The enhancement of mechanical properties was mainly attributed to grain refinement and crack deflections owing to the presence of TiB₂ particles.

Acknowledgements

Haibin Zhang is grateful to the foundation by the Recruitment Program of Global Youth Experts and the Youth Hundred Talents Project of Sichuan Province. This work is supported by the National Natural Science Foundation of China (Grant no. 91326102), the Science and Technology Development Foundation of China Academy of Engineering Physics (Grant no. 2013A0301012), and Science and Technology Innovation Research Foundation of Institute of Nuclear Physics and Chemistry.

References

1. A. K. Suri, C. Subramanian, J. K. Sonber and T. S. R. C. Murthy, Int. Mater. Rev., 2010, 55, 1.
2. V. Domnich, S. Reynaud, R. A. Haber and M. Chhowalla, J. Am. Ceram. Soc., 2011, 94, 3605.
3. F. Thevenot, J. Eur. Ceram. Soc., 1990, 6, 205.
4. T. Stoto, N. Housseau, L. Zuppiroli and B. Kryger, J. Appl. Phys., 1990, 68, 3198.
5. T. Maruyama, J. Tech. Assoc. Refract., Jpn., 2010, 30, 80.
6. K. A. Schwetz and W. Grellner, J. Less-Common Met., 1981, 82, 37.
7. B. Y. Yin, L. S. Wang and Y. C. Fang, J. Chin. Ceram. Soc., 2001, 29, 68.
8. H. Lee and R. F. Speyer, J. Am. Ceram. Soc., 2003, 86, 1468.
9. T. Kobayashi, K. Yoshida and T. Yano, J. Nucl. Mater., 2013, 440, 524.
10. C. H. Danny, J. P. Aleksander, A. A. Ilhan and E. S. William, J. Am. Ceram. Soc., 1989, 72, 775.
11. M. Mashhadi, E. Taheri-Nassaj, M. Mashhadi and V. M. Sglavo, Ceram. Interfaces, 2011, 37, 3229.
12. F. Ye, Z. Hou, H. Zhang and L. Liu, J. Am. Ceram. Soc., 2010, 93, 2956.
13. C. Xu, H. Zeng and G. Zhang, Int. J. Refract. Met. Hard Mater., 2013, 41, 2.
14. R. Telle and G. Petzow, Powder Metall., 1986, 2, 1155.
15. C. H. Lee and C. H. Kim, J. Mater. Sci., 1992, 27, 6335.
16. K. Y. Xie, M. F. Toksoy, K. Kuwelkar, B. Zhang, J. A. Krogstad, R. A. Haber and K. J. Hemker, J. Am. Ceram. Soc., 2014, 97, 3710.
17. S. Yamada, K. Hirao, Y. Yamachi and S. Knzaki, J. Mater. Sci. Lett., 2002, 21, 1445.
18. A. Goldstein, Y. Yeshurun and A. Goldenberg, J. Eur. Ceram. Soc., 2007, 27, 695.
19. S. Yamadaa, K. Hiraob, Y. Yamauchib and S. Kanzakib, J. Eur. Ceram. Soc., 2003, 23, 1123.
20. V. Skorokhod and V. D. Krstic, J. Mater. Sci. Lett., 2000, 19, 237.
21. A. Goldstein, Y. Geffen and A. Goldenberg, J. Am. Ceram. Soc., 2001, 84, 642.
22. Z. Zhang, X. Du, Z. Li, W. Wang, J. Zhang and Z. Fu, J. Eur. Ceram. Soc., 2014, 34, 2153.
23. K. Hirota, Y. Nakayama and S. Nakane, US Pat., US 2010/0311561 A1, 2010.
24. G. R. Anstis, P. Chantikul, B. R. Lawn and D. B. Marshall, J. Am. Ceram. Soc., 1981, 64, 533.
25. L. Y. Zheng, F. Z. Li and Y. C. Zhou, J. Am. Ceram. Soc., 2012, 95, 2028.
26. T. Wang and A. Yamaguchi, J. Mater. Sci. Lett., 2000, 19, 1245.
27. Y. Gao, Z. Huang, M. Fang, Y. Liu, S. Huang and X. Ouyang, Powder Technol., 2012, 226, 269.
28. V. V. Skorokhod, M. D. Vlajic and V. D. Krstic, Mater. Sci. Forum, 1998, 282–283, 219.
29. J. Deng and J. Sun, Ceram. Interfaces, 2009, 35, 771.
30. W. G. Fahrenholtz, E. W. Neuman, H. J. Brown-Shaklee and G. E. Hilmas, J. Am. Ceram. Soc., 2010, 93, 3580.
31. J. Zou, S. Huang, K. Vanmeensel, G. Zhang, J. Vleugels and O. van der Biest, J. Am. Ceram. Soc., 2013, 96, 1055.

---