Lithium Disilicate Glass-Ceramics Containing Lithium Tantalate As A Secondary Phase
Open Access
- Author:
- De Ceanne, Anthony
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 01, 2022
- Committee Members:
- Seong Kim, Outside Unit & Field Member
John Mauro, Chair & Dissertation Advisor
John Mauro, Program Head/Chair
Robert Kimel, Major Field Member
Allison Beese, Major Field Member - Keywords:
- Glass-ceramics
Lithium disilicate
Lithium tantalate
Machine learning - Abstract:
- Lithium disilicate (LD) glass-ceramics were discovered in the 1950s, and the evolution of the main target phase of these materials is well-known. However, the LD system is still an area of active research, especially when a secondary phase is introduced since the phase evolution of secondary phases is not currently understood. The first stage of the present work examines microstructural effects in a single parent composition of a LD system that contains lithium tantalate (LT) as a secondary phase. The phase assemblage evolution is studied via x-ray diffraction and the crystal morphology of the system is imaged via scanning electron microscopy. It is found that the nucleation of the precursor lithium metasilicate phase is coupled to both the nucleation of both LD and LT, and from this work, a reaction equation to describe these relationships is proposed: 2(Li2SiO3) + Ta2O5 Li2Si2O5 + 2(LiTaO3). A large data set is analyzed to investigate the relationship between heat treatments and several key properties, namely optical and mechanical properties, specifically the L*a*b* color coordinates and the Knoop hardness (HK). This is the first known systematic study of Knoop hardness for a LD glass-ceramic. It is found that the L* values increase once the growth rates of LD and LT increase and that HK is higher for microstructures that were produced via a high nucleation temperature and middle to high growth temperature. The experimental data are used to train several machine learning models to predict color coordinates and HK values. These first ever machine learning models for glass-ceramic materials successfully predict the color and hardness trends of the materials and also predict values for new microstructures that are not examined experimentally. Linear regression models provide physical insight into the HK results, showing that the hardest samples are the result of microstructures that contain a high amount of small crystals, especially LD and LT crystals. To build off this work, the second stage of the present study designed novel compositions to examine the effect of the ratio of [SiO2] to [Li2O] on the phase assemblage, microstructure, and corresponding properties of this LD system that contained LT as a secondary phase. This study was aimed at validating or disproving the proposed reaction equation from the first stage. The phase assemblage data are in support of the proposed reaction equation because LT content increased, and LD content decreased in the material as the [SiO2]/[Li2O] decreased. Lastly, the study provides a glimpse into the radiopacity values of each composition, finding that radiopacity increases as more Ta2O5 is introduced into the composition.