Mechanical Properties of Ice Ih from Deep Potential Molecular Simulations

Abstract The mechanical behavior of ice Ih under cryogenic conditions is of fundamental importance to glaciology, planetary geophysics, and materials science, yet a comprehensive atomistic understanding has remained elusive due to the limitations of both experimental probes and classical molecular simulations. Here, we present a deep potential molecular dynamics study of single-crystalline ice Ih across 50−273 K using a machine-learned potential trained on SCAN-DFT data. Our computed Young’s modulus and strength decrease monotonically with temperature and agree well with experimental results above ∼150 K, but they lack the low-temperature modulus plateau. We uncover a pronounced mechanical anomaly different from common brittle materials originating from fundamentally distinct deformation pathways: tension proceeds through uniform hydrogen-bond elongation and abrupt brittle deformation, whereas compression drives progressive tetrahedral angular distortion, cage breakdown, and ultimately localized shear amorphization. Notably, the small pre-yield hump observed in basal-plane compression at low temperatures is considered a structural signature of the ice Ih-to-ice II-like local ordering. Together with the temperature-dependent elastic moduli, this finding provides a mechanistic explanation for both the anomalous mechanical response and the stress-induced phase transformation observed in recent experiments on bent ice microfibers. Furthermore, point defects systematically reduce the failure strain while preserving the elastic stiffness and qualitative failure characteristics. Our results establish a mechanistic framework for the nano-mechanics of ice and demonstrate the predictive capability of machine-learned potentials for investigating hydrogen-bonded crystalline materials.

Authors

Institutions

Publication Details

Journal
Crystal Growth & Design
Published
2026-11-14
DOI
https://doi.org/10.1021/acs.cgd.6c01221
Primary Topic
Machine Learning in Materials Science
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Mechanical Properties of Ice Ih from Deep Potential Molecular Simulations

Guang‐Jun Guo, Zhengcai Zhang, Nengyou Wu, Yizhao Wan et al.
Crystal Growth & Design
Machine Learning in Materials Science
article

Mechanical Properties of Ice Ih from Deep Potential Molecular Simulations

Guang‐Jun Guo, Zhengcai Zhang, Nengyou Wu, Yizhao Wan, Yanlong Li, Yuan Zhou, Jiapeng Jin, Tiantian Sun, Li Huang, Minhui Qi
article en

Abstract

Abstract The mechanical behavior of ice Ih under cryogenic conditions is of fundamental importance to glaciology, planetary geophysics, and materials science, yet a comprehensive atomistic understanding has remained elusive due to the limitations of both experimental probes and classical molecular simulations. Here, we present a deep potential molecular dynamics study of single-crystalline ice Ih across 50−273 K using a machine-learned potential trained on SCAN-DFT data. Our computed Young’s modulus and strength decrease monotonically with temperature and agree well with experimental results above ∼150 K, but they lack the low-temperature modulus plateau. We uncover a pronounced mechanical anomaly different from common brittle materials originating from fundamentally distinct deformation pathways: tension proceeds through uniform hydrogen-bond elongation and abrupt brittle deformation, whereas compression drives progressive tetrahedral angular distortion, cage breakdown, and ultimately localized shear amorphization. Notably, the small pre-yield hump observed in basal-plane compression at low temperatures is considered a structural signature of the ice Ih-to-ice II-like local ordering. Together with the temperature-dependent elastic moduli, this finding provides a mechanistic explanation for both the anomalous mechanical response and the stress-induced phase transformation observed in recent experiments on bent ice microfibers. Furthermore, point defects systematically reduce the failure strain while preserving the elastic stiffness and qualitative failure characteristics. Our results establish a mechanistic framework for the nano-mechanics of ice and demonstrate the predictive capability of machine-learned potentials for investigating hydrogen-bonded crystalline materials.

Crystal Growth & Design
Chinese Academy of Sciences (CN), Qingdao Institute of Marine Geology (CN), University of Chinese Academy of Sciences (CN), Laoshan Laboratory (CN)
Machine Learning in Materials Science
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.

Laika access coming soon