MENU

Glass Formation Mechanisms: ML-Driven Analysis of 78,889 Entries

npj Computational Materials China
Overview
Researchers have developed a SHAP-integrated XGBoost machine learning framework that identifies universal and system-specific mechanisms for glass formation across covalent, ionic, and metallic networks. The model achieves 89.6% classification accuracy and reveals a “frustration inversion” principle that governs amorphous stability. This provides a data-driven foundation for the design of functional glasses across different bonding characters.
In Depth

Key Findings

The study establishes a unified physicochemical framework for glass formation by analyzing a success-bias-free dataset of 78,889 compositional entries. The resulting XGBoost framework attained 89.6% classification accuracy on the all-glass dataset and demonstrated robust generalization when applied to higher-dimensional spaces and extrapolative elemental substitutions.

Technical Details

The researchers encoded compositions as elemental-property matrices using the Magpie database. By applying Shapley analysis to both all-glass and system-specific models, the study determined that four glass families share the same set of frustration mechanisms, though the relative weights of these mechanisms redistribute based on bonding character.

A critical discovery is the “frustration inversion” effect, characterized by a bounded, non-monotonic dependence of governing descriptors. This principle indicates that compositional heterogeneity stabilizes the amorphous state only up to a family-specific threshold; this threshold narrows as network rigidity decreases, providing quantitative bounds on the confusion principle. Additionally, the analysis identified BCC-polymorph-referenced elemental descriptors as compositional-level predictors that encode structural preferences toward vitrification.

Background & Context

Existing models for glass formation have historically been limited because they were largely confined to a single glass family or biased toward compositions that had already been successfully vitrified. This research bridges that gap by utilizing a comprehensive dataset drawn from glass-forming region maps to eliminate success bias.

Significance & Outlook

By uncovering the universal mechanisms of amorphous stability across covalent, ionic, and metallic systems, this work provides a cross-system foundation for materials science. The integration of machine learning and SHAP analysis offers a scalable, data-driven route to the design of functional glasses, moving beyond trial-and-error compositional searches.

Source: https://www.nature.com/articles/s41524-026-02385-3

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.

  • Your email and selected fields are used only to deliver the newsletter.
  • We never share your information with third parties.
  • You can unsubscribe anytime via the link in each email.

See our Privacy Policy for details.

Takes about a minute · Unsubscribe anytime

Published by Troy-Technical, an independent site run by one engineer with a career in materials development.
About the author / Contact info@troy-technical.jp
Let's share this post !

Author of this article

TOC