Local atomic environments reveal structural controls on MgSiO3 liquid-crystal similarity

Abstract MgSiO3 is a fundamental silicate system in Earth and rocky planetary interiors, where its structures control crystallization, differentiation, and evolution of host planets. Yet, silicate structure has usually been analyzed through aggregate structural observables that do not place individual local atomic environments from crystals and liquids into a common similarity space. Here we compare local environments across eight MgSiO3 polymorphs and the liquid up to 700 GPa and 12,000 K generated via a machine-learning potential and first-principles calculations. We use the Smooth Overlap of Atomic Positions (SOAP) descriptor, principal component analysis (PCA), and k-nearest-neighbor distances in the full descriptor space. PCA shows that the dominant variation largely follows density-driven packing, whereas full-space distances reveal pronounced differences in structural similarity among MgSiO3 polymorphs and the liquid. Pyroxene-family phases, majorite, and akimotoite remain comparatively close to the liquid, but bridgmanite and post-perovskite are systematically less liquid-like. This separation is not explained by the SiO4-to-SiO6 coordination change alone, and instead stems from a coupled variation in multiple structural features. These results show that local atomic descriptors expose structural distinctions hidden by conventional averaged metrics and suggest that high-pressure liquid thermodynamics, including the non-monotonic pressure dependence of liquid Grüneisen parameters, need not be interpreted simply as progressive convergence toward the stable lower-mantle crystal structures.

Authors

Institutions

Publication Details

Journal
PNAS Nexus
Published
2026-10-06
DOI
https://doi.org/10.1093/pnasnexus/pgag347
Primary Topic
High-pressure geophysics and materials
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Local atomic environments reveal structural controls on MgSiO3 liquid-crystal similarity

Jie Deng, Jake Acquadro, Xiyuan Bao
PNAS Nexus
High-pressure geophysics and materials
article

Local atomic environments reveal structural controls on MgSiO3 liquid-crystal similarity

Jie Deng, Jake Acquadro, Xiyuan Bao
article en

Abstract

Abstract MgSiO3 is a fundamental silicate system in Earth and rocky planetary interiors, where its structures control crystallization, differentiation, and evolution of host planets. Yet, silicate structure has usually been analyzed through aggregate structural observables that do not place individual local atomic environments from crystals and liquids into a common similarity space. Here we compare local environments across eight MgSiO3 polymorphs and the liquid up to 700 GPa and 12,000 K generated via a machine-learning potential and first-principles calculations. We use the Smooth Overlap of Atomic Positions (SOAP) descriptor, principal component analysis (PCA), and k-nearest-neighbor distances in the full descriptor space. PCA shows that the dominant variation largely follows density-driven packing, whereas full-space distances reveal pronounced differences in structural similarity among MgSiO3 polymorphs and the liquid. Pyroxene-family phases, majorite, and akimotoite remain comparatively close to the liquid, but bridgmanite and post-perovskite are systematically less liquid-like. This separation is not explained by the SiO4-to-SiO6 coordination change alone, and instead stems from a coupled variation in multiple structural features. These results show that local atomic descriptors expose structural distinctions hidden by conventional averaged metrics and suggest that high-pressure liquid thermodynamics, including the non-monotonic pressure dependence of liquid Grüneisen parameters, need not be interpreted simply as progressive convergence toward the stable lower-mantle crystal structures.

PNAS Nexus
Planetary Science Institute (US)
Openalex Percentile: Top 15%
High-pressure geophysics and materials
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.