Decoding crystallization behavior of aluminoborosilicate glasses: From structural descriptors to Quantitative Structure - Property Relationship (QSPR) based predictive models - Université de Lille Accéder directement au contenu
Article Dans Une Revue Acta Materialia Année : 2024

Decoding crystallization behavior of aluminoborosilicate glasses: From structural descriptors to Quantitative Structure - Property Relationship (QSPR) based predictive models

Résumé

Successful decoding of structural descriptors controlling the crystallization in multicomponent functional glasses can pave the way for the transition from the trial-and-error approach and empirical modeling for glass/glass-ceramic composition design toward more rational and scientifically rigorous Quantitative Structure-Property Relationship (QSPR) based models. However, due to the compositional and structural complexity of multicomponent glasses and the longer time and length scales associated with nucleation, the development and validation of QSPR models are still in it's infancy. The work presented in the article is an attempt to leap forward in this pursuit by combining the strengths of experimental and computational materials science to decode the chemo-structural drivers that promote or suppress nucleation and crystal growth in alkali/alkaline-earth aluminoborosilicate glasses leading to the development of a QSPR-based model (powered by MD simulations). The results reveal the following two descriptors that govern the nucleation and crystallization of a particular aluminosilicate phase in the functional glasses: (1) degree of mixing between the SiO4 and AlO4 units, i.e., Si–O–Al linkages, and (2) difference/similarity between the short-to-intermediate range ordering in the glass structure to that of the structure of corresponding crystalline phase. Based on the established composition–structure–crystallization behavior relationships, a cluster analysis based QSPR model has been developed (and tested) to predict the propensity of nepheline (and anorthite) crystallization in the investigated glasses. The model has been tested on several compositions from the present and previous studies and has successfully predicted the crystallization propensity of all glass compositions, even in cases where previous empirical and semi-empirical models were unsuccessful.
Fichier principal
Vignette du fichier
1-s2.0-S1359645424001368-main.pdf (5.43 Mo) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-04583241 , version 1 (22-05-2024)

Licence

Identifiants

Citer

Yingcheng Zhang, Marco Bertani, Alfonso Pedone, Randall E Youngman, Grégory Tricot, et al.. Decoding crystallization behavior of aluminoborosilicate glasses: From structural descriptors to Quantitative Structure - Property Relationship (QSPR) based predictive models. Acta Materialia, 2024, Acta Mater., 268, ⟨10.1016/j.actamat.2024.119784⟩. ⟨hal-04583241⟩
3 Consultations
1 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More