Figure 1 showed the XRD pattern of as-quenched glass and glass-ceramics. It could be seen that no crystalline phases had been precipitated in the prepared parent glass. However, LiAlSi2O6 and LiAlSi3O8 constituted as the major crystalline phases were precipitated in the prepared glass-ceramics. Besides, Li2SiO3 and BaAl2Si2O8 were identified as the minor crystalline phases. An overall analysis of Fig. 1 could reach the conclusion that the changing of crystallization temperature made no difference to the type of crystalline phase precipitated in the glass-ceramics.
Figure 2 displayed that the CTE of glass-ceramic crystallized at 550 °C was 2.91 × 10− 6 °C− 1, and it was 3.13 × 10− 6 °C− 1and 3.11 × 10− 6 °C− 1 when the crystallization temperature increased to 575 °C and 600 °C, respectively. However, the CTE of glass-ceramic had reduced to 2.58 × 10− 6 °C− 1 with a rising crystallization temperature to 625 °C, indicating a higher crystallization temperature was not conducive to the preparation of with CTE close to cBN (3.50 × 10− 6 °C− 1).
To explore the type and content of crystalline phases on the performance of CTE for glass-ceramics, a theoretical calculation method of CTE was adopted[11].
(1)
Where αgc denotes the calculated CTE of glass-ceramics, αg, αc1, αc2, αc3, and αc4 stands for the CTE of residual amorphous phase, LiAlSi2O6 (0.9 × 10− 6 °C− 1)[12], LiAlSi3O8 (3 × 10− 6 °C− 1)[13], Li2SiO3 (9.0 × 10− 6 °C− 1)[14], and BaAl2Si2O8 (2.2 × 10− 6 °C− 1)[15], respectively. While the xg, xc1, xc2, xc3, and xc4 represents the content of the above-mentioned phases.
In Eq. (1), the CTE of the residual amorphous phase could be determined as 7.69 × 10− 6 °C− 1 by the chemical compositions of glass shown in Table 1 according to the Appen method[16]. In addition, the phase content in glass-ceramics could be obtained by WPF refinement of XRD pattern, and the calculated and measured CTE of glass-ceramics, as well as the phase content in glass-ceramics, were depicted in Fig. 3.
As Fig. 3 pointed out that the calculated and measured CTEs of glass-ceramics were close to each other, indicating that the calculated method of CTEs for glass-ceramics in this paper were suitable and effective. The crystallinity of the glass-ceramic continued to increase with a rising crystallization temperature. As for the content of major crystalline phases in glass-ceramics, the phase content of LiAlSi2O6 reached its maximum in M1 among the four samples which was 63.25%, while it was 51.90%~53.55% for the rest crystallization temperatures, proving that the precipitation of LiAlSi2O6 was almost unaffected when the crystallization temperature was among 575 ~ 625 °C. Furthermore, the phase content of LiAlSi3O8 kept increasing with the rising crystallization temperature indicating that the precipitation of LiAlSi3O8 was inspired at a higher crystallization temperature. Actually, the transition temperature of nucleation from crystallization temperature for LiAlSi3O8 was lower than that of LiAlSi2O6[17], demonstrating that the high crystallization temperature was more favorable for the precipitation of LiAlSi3O8. Thus, a higher content of crystalline phase and LiAlSi3O8 in M4 could be reasonably explained. While it could be seen that the CTEs of LiAlSi2O6 and LiAlSi3O8 were obviously lower than the amorphous phase. Hence, a higher content of LiAlSi2O6 and LiAlSi3O8, as well as a lower content of the amorphous phase, could be beneficial for a lower CTE of M4 glass-ceramic according to Eq. (1). However, since the CTE of cBN was 3.50 × 10− 6 °C− 1, the difference between M4 and cBN in CTE was also too large to the formation of microcracks aroused by the mismatch of cBN and glass-ceramic bond. In brief, rational control of the crystallization temperature for the two-step heat treatment method can adjust the relative content of LiAlSi2O6 and LiAlSi3O8, thereby preparing a glass-ceramic with CTE close to cBN.