Hybrid Monte Carlo technique for modeling of crystal nucleation and application to lithium disilicate glass-ceramics

Matthew E. McKenzie, John C. Mauro

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

We propose a computational method for studying crystal nucleation in glasses and supercooled liquids, combining the techniques of cluster formation via Monte Carlo, Steinhardt order parameter biasing, and an implicit solvation model. Each of these techniques calculates an important contribution to the overall nucleating free energy. This hybrid Monte Carlo technique is applied to the canonical example of a lithium disilicate glass-ceramic, where it is found that the cluster formation and cluster-to-crystal transition energies are approximately equal. The known crystal precursor, lithium metasilicate, has a smaller thermodynamic barrier to nucleation compared to that of lithium disilicate. The solvation energy is small compared to the formation and crystallization energies; yet the metasilicate solvation energy is much lower, indicating easier precipitation compared to the disilicate cluster. Our hybrid Monte Carlo approach is generally applicable to study the nucleation and crystallization processes in any arbitrary glass or supercooled liquid.

Original languageEnglish (US)
Pages (from-to)202-207
Number of pages6
JournalComputational Materials Science
Volume149
DOIs
StatePublished - Jun 15 2018

Fingerprint

Hybrid Monte Carlo
Glass-ceramics
Monte Carlo Techniques
Solvation
Glass ceramics
Nucleation
Lithium
Crystal
lithium
nucleation
ceramics
Crystallization
solvation
Cluster Formation
Supercooled Liquid
Crystals
glass
Energy
Modeling
crystals

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Chemistry(all)
  • Materials Science(all)
  • Mechanics of Materials
  • Physics and Astronomy(all)
  • Computational Mathematics

Cite this

@article{e5276949437d4fb58dd508bd72d1a2c0,
title = "Hybrid Monte Carlo technique for modeling of crystal nucleation and application to lithium disilicate glass-ceramics",
abstract = "We propose a computational method for studying crystal nucleation in glasses and supercooled liquids, combining the techniques of cluster formation via Monte Carlo, Steinhardt order parameter biasing, and an implicit solvation model. Each of these techniques calculates an important contribution to the overall nucleating free energy. This hybrid Monte Carlo technique is applied to the canonical example of a lithium disilicate glass-ceramic, where it is found that the cluster formation and cluster-to-crystal transition energies are approximately equal. The known crystal precursor, lithium metasilicate, has a smaller thermodynamic barrier to nucleation compared to that of lithium disilicate. The solvation energy is small compared to the formation and crystallization energies; yet the metasilicate solvation energy is much lower, indicating easier precipitation compared to the disilicate cluster. Our hybrid Monte Carlo approach is generally applicable to study the nucleation and crystallization processes in any arbitrary glass or supercooled liquid.",
author = "McKenzie, {Matthew E.} and Mauro, {John C.}",
year = "2018",
month = "6",
day = "15",
doi = "10.1016/j.commatsci.2018.03.034",
language = "English (US)",
volume = "149",
pages = "202--207",
journal = "Computational Materials Science",
issn = "0927-0256",
publisher = "Elsevier",

}

Hybrid Monte Carlo technique for modeling of crystal nucleation and application to lithium disilicate glass-ceramics. / McKenzie, Matthew E.; Mauro, John C.

In: Computational Materials Science, Vol. 149, 15.06.2018, p. 202-207.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Hybrid Monte Carlo technique for modeling of crystal nucleation and application to lithium disilicate glass-ceramics

AU - McKenzie, Matthew E.

AU - Mauro, John C.

PY - 2018/6/15

Y1 - 2018/6/15

N2 - We propose a computational method for studying crystal nucleation in glasses and supercooled liquids, combining the techniques of cluster formation via Monte Carlo, Steinhardt order parameter biasing, and an implicit solvation model. Each of these techniques calculates an important contribution to the overall nucleating free energy. This hybrid Monte Carlo technique is applied to the canonical example of a lithium disilicate glass-ceramic, where it is found that the cluster formation and cluster-to-crystal transition energies are approximately equal. The known crystal precursor, lithium metasilicate, has a smaller thermodynamic barrier to nucleation compared to that of lithium disilicate. The solvation energy is small compared to the formation and crystallization energies; yet the metasilicate solvation energy is much lower, indicating easier precipitation compared to the disilicate cluster. Our hybrid Monte Carlo approach is generally applicable to study the nucleation and crystallization processes in any arbitrary glass or supercooled liquid.

AB - We propose a computational method for studying crystal nucleation in glasses and supercooled liquids, combining the techniques of cluster formation via Monte Carlo, Steinhardt order parameter biasing, and an implicit solvation model. Each of these techniques calculates an important contribution to the overall nucleating free energy. This hybrid Monte Carlo technique is applied to the canonical example of a lithium disilicate glass-ceramic, where it is found that the cluster formation and cluster-to-crystal transition energies are approximately equal. The known crystal precursor, lithium metasilicate, has a smaller thermodynamic barrier to nucleation compared to that of lithium disilicate. The solvation energy is small compared to the formation and crystallization energies; yet the metasilicate solvation energy is much lower, indicating easier precipitation compared to the disilicate cluster. Our hybrid Monte Carlo approach is generally applicable to study the nucleation and crystallization processes in any arbitrary glass or supercooled liquid.

UR - http://www.scopus.com/inward/record.url?scp=85044164409&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85044164409&partnerID=8YFLogxK

U2 - 10.1016/j.commatsci.2018.03.034

DO - 10.1016/j.commatsci.2018.03.034

M3 - Article

AN - SCOPUS:85044164409

VL - 149

SP - 202

EP - 207

JO - Computational Materials Science

JF - Computational Materials Science

SN - 0927-0256

ER -