2025 年 133 巻 4 号 p. 131-138
P2O5 is an efficacious nucleating agent for promoting the crystallization of lithium disilicate glass, with numerous studies attributing its effectiveness to the formation of Li3PO4 crystals, which serve as heterogeneous nucleation sites for Li2Si2O5 precipitation. However, recent studies have not consistently observed Li3PO4 precipitation. To address this controversy, this study utilizes Li3PO4 as a nucleating agent to determine whether Li3PO4 crystallites induce the precipitation of Li2Si2O5. Additionally, Na3PO4, an analogue of Li3PO4, is employed to evaluate its potential to promote nucleation, while P2O5 serves as a comparative reference under identical conditions. Upon heating, all three agents fully melt in the glass. Subsequent heat treatment results in the precipitation of Li2SiO3 crystals, followed by Li2Si2O5 crystals. It remains unclear whether these phosphates precipitate prior to the Li2SiO3 crystals. Although all agents exhibit comparable effects, the Li2Si2O5 grain size in the Na3PO4 group is finer, leading to decreased strength and increased transparency in the samples. This may be due to the mixed alkali effect, where Na+ ions inhibit the growth of Li2Si2O5 crystals. Notably, the initially precipitated Li2SiO3 crystals exhibit a higher Li content than the original glass, indicating that the nucleating agents alter the local glass composition to resemble that of the Li2SiO3 crystals, thereby facilitating crystallization. Subsequently, Li2SiO3 crystals decompose to form Li2Si2O5 crystals, resulting in a redistribution of Li+ ions from localized enrichment back to a uniform distribution consistent with the original glass.