Abstract

Ice-nucleating proteins (INPs) are a unique class of biological macromolecules that catalyze the freezing of supercooled water far more efficiently than homogeneous nucleation. Their remarkable efficiency has motivated applications across diverse sectors, including agricultural frost protection, food processing and packaging, biomedical cryopreservation, and even strategies for mitigating glacier ice loss. The ice-nucleation performance of INPs and the mechanical behavior of the ice they produce depend strongly on their structural and biochemical characteristics. However, the links between INP properties, the resulting ice microstructure, and their mechanical behavior have yet to be systematically established. In this study, coarse-grained molecular dynamics (CGMD) simulations using the machine-learned ML-BOP potential are employed to investigate how varying INP densities influence the ice nucleation temperature, the resulting ice microstructure, and the mechanical behavior of the formed ice under creep tensile loading. We find that, depending on their density, INPs can significantly raise the ice nucleation rate while altering the grain structure of ice. Our simulations reveal that INP-assisted nucleation leads to faster stabilization of the resulting polycrystalline ice composed of hexagonal ice (ice Ih) and cubic ice (ice Ic) as compared to nucleation in pure water. Moreover, higher INP densities and smaller ice grain sizes reduce the overall yield stress, while promoting diffusion-accommodated grain boundary sliding creep. These findings provide molecular-level insight into how INPs influence both the nucleation process and the mechanical behavior of ice, highlighting a pathway to engineer ice with tailored stability for real-world settings, including human activities and infrastructure in polar and icy environments.

Abstract

Ice-nucleating proteins (INPs) are a unique class of biological macromolecules that catalyze the freezing of supercooled water far more efficiently than homogeneous nucleation. Their remarkable efficiency has motivated applications across diverse sectors, including agricultural frost protection, food processing and packaging, biomedical cryopreservation, and even strategies for mitigating glacier ice loss. The ice-nucleation performance of INPs and the mechanical behavior of the ice they produce depend strongly on their structural and biochemical characteristics. However, the links between INP properties, the resulting ice microstructure, and their mechanical behavior have yet to be systematically established. In this study, coarse-grained molecular dynamics (CGMD) simulations using the machine-learned ML-BOP potential are employed to investigate how varying INP densities influence the ice nucleation temperature, the resulting ice microstructure, and the mechanical behavior of the formed ice under creep tensile loading. We find that, depending on their density, INPs can significantly raise the ice nucleation rate while altering the grain structure of ice. Our simulations reveal that INP-assisted nucleation leads to faster stabilization of the resulting polycrystalline ice composed of hexagonal ice (ice Ih) and cubic ice (ice Ic) as compared to nucleation in pure water. Moreover, higher INP densities and smaller ice grain sizes reduce the overall yield stress, while promoting diffusion-accommodated grain boundary sliding creep. These findings provide molecular-level insight into how INPs influence both the nucleation process and the mechanical behavior of ice, highlighting a pathway to engineer ice with tailored stability for real-world settings, including human activities and infrastructure in polar and icy environments.

Abstract

Ice-nucleating proteins (INPs) are a unique class of biological macromolecules that catalyze the freezing of supercooled water far more efficiently than homogeneous nucleation. Their remarkable efficiency has motivated applications across diverse sectors, including agricultural frost protection, food processing and packaging, biomedical cryopreservation, and even strategies for mitigating glacier ice loss. The ice-nucleation performance of INPs and the mechanical behavior of the ice they produce depend strongly on their structural and biochemical characteristics. However, the links between INP properties, the resulting ice microstructure, and their mechanical behavior have yet to be systematically established. In this study, coarse-grained molecular dynamics (CGMD) simulations using the machine-learned ML-BOP potential are employed to investigate how varying INP densities influence the ice nucleation temperature, the resulting ice microstructure, and the mechanical behavior of the formed ice under creep tensile loading. We find that, depending on their density, INPs can significantly raise the ice nucleation rate while altering the grain structure of ice. Our simulations reveal that INP-assisted nucleation leads to faster stabilization of the resulting polycrystalline ice composed of hexagonal ice (ice Ih) and cubic ice (ice Ic) as compared to nucleation in pure water. Moreover, higher INP densities and smaller ice grain sizes reduce the overall yield stress, while promoting diffusion-accommodated grain boundary sliding creep. These findings provide molecular-level insight into how INPs influence both the nucleation process and the mechanical behavior of ice, highlighting a pathway to engineer ice with tailored stability for real-world settings, including human activities and infrastructure in polar and icy environments.

Abstract

Ice-nucleating proteins (INPs) are a unique class of biological macromolecules that catalyze the freezing of supercooled water far more efficiently than homogeneous nucleation. Their remarkable efficiency has motivated applications across diverse sectors, including agricultural frost protection, food processing and packaging, biomedical cryopreservation, and even strategies for mitigating glacier ice loss. The ice-nucleation performance of INPs and the mechanical behavior of the ice they produce depend strongly on their structural and biochemical characteristics. However, the links between INP properties, the resulting ice microstructure, and their mechanical behavior have yet to be systematically established. In this study, coarse-grained molecular dynamics (CGMD) simulations using the machine-learned ML-BOP potential are employed to investigate how varying INP densities influence the ice nucleation temperature, the resulting ice microstructure, and the mechanical behavior of the formed ice under creep tensile loading. We find that, depending on their density, INPs can significantly raise the ice nucleation rate while altering the grain structure of ice. Our simulations reveal that INP-assisted nucleation leads to faster stabilization of the resulting polycrystalline ice composed of hexagonal ice (ice Ih) and cubic ice (ice Ic) as compared to nucleation in pure water. Moreover, higher INP densities and smaller ice grain sizes reduce the overall yield stress, while promoting diffusion-accommodated grain boundary sliding creep. These findings provide molecular-level insight into how INPs influence both the nucleation process and the mechanical behavior of ice, highlighting a pathway to engineer ice with tailored stability for real-world settings, including human activities and infrastructure in polar and icy environments.