HomeScienceMax-Flow Algorithm Predicts Solid-Stat
SCIENCE

Max-Flow Algorithm Predicts Solid-State Glass Conductivity

Researchers used graph-based network flow calculations on static atomic structures to estimate ionic conductivity in sulfide glass battery electrolytes.

WHAT YOU NEED TO KNOW
  • Nature Communications published the peer-reviewed findings on August 31, 2026.
  • The method models static atomic structures as periodic weighted graphs to calculate maximum flow values.
  • Simulations verified positive correlations between calculated flow and ionic conductivity for lithium and sodium sulfide glasses.
  • Authors from Aalborg University, IST Austria, and the University of Sydney utilized Sophia and LUMI-C supercomputing resources.

Researchers developed a computational method to predict the ionic conductivity of sulfide glass solid-state electrolytes by applying a max-flow algorithm to static atomic structures, according to a study published in Nature Communications.

Disordered glassy materials offer potential as electrolytes for solid-state batteries, though existing methods to evaluate their ionic conductivity remain inefficient. The team built a framework that summarizes the topological properties of an electrolyte's atomic structure into a weighted graph using periodic boundary conditions.

The researchers adapted a variation of the max-flow algorithm, commonly used in continuum models for transportation and fluid flow, to run on the graph representations. They validated the method through molecular dynamics simulations across multiple glassy solid-state electrolyte families containing mobile lithium and sodium ions.

Simulations showed a positive correlation between ionic conductivity and the area-normalized maximum flow calculated directly from static glass structures, particularly in high-conductivity sulfide glass electrolytes. Because the approach relies entirely on static atomic data, the authors noted it can accelerate the discovery of conductive glasses for all-solid-state batteries.

Matteo Pegoraro, Rasmus Christensen, and Søren S. Sørensen contributed equally to the study alongside co-authors at Aalborg University, the Institute of Science and Technology Austria, and the University of Sydney. The project used computational resources at Sophia and LUMI-C provided by the Danish e-Infrastructure Cooperation, with funding from the Independent Research Fund Denmark, the European Research Council, the Austrian Science Fund, and the Danish Data Science Academy.

Xentir Media
Xentir Media NewsroomSource-backed AI and technology coverage, drafted by Xentir's automated editorial system under fixed human-set rules. See our editorial policy and AI usage policy.
J
Jomon · Founder & EditorFounder and editor of Xentir Media. Sets the editorial rules the newsroom system runs under, and is accountable for its corrections. About Jomon · [email protected]
The Xentir Brief
The developments worth knowing — one useful email.
Get the Brief →