Computational Investigation of Ice-Nucleation and Ice-Nucleating Proteins

Due to climate change, the Department of Defense (DoD) is witnessing permafrost melting, putting critical coastal infrastructure at risk of degradation. This includes the warping of airstrips, sinking of buildings, and erosion of coastal structures. To combat this, the DoD has put forth multiple initiatives, one being to investigate different methods to nucleate ice. Ice-Nucleating Proteins (INPs) are proteins that bind water molecules to specific sites, encouraging a crystalline structure to form. Before specific proteins can be investigated, the mechanical properties of ice nucleation must be understood. Understanding the intricate dynamics of ice nucleation is essential for developing effective strategies to preserve vital infrastructure. This research delves into the factors influencing ice formation and the role of impurities in water. By elucidating the fundamental principles behind ice nucleation, we aim to identify potential methods for controlling ice formation in strategic locations, thereby mitigating the adverse effects of permafrost degradation. In order to achieve this task, Molecular Dynamics (MD) simulations were conducted to provide insights as to how ice crystals form. A Graph Convolutional Neural Network (GCNN) was also implemented to analyze atomic-level interactions during ice nucleation. These computational methods were implemented to gain an understanding of the ice nucleation process.

PRESENTER

Albrecht, Benjamin
benalbrecht6@gmail.com
Maimi University

CO-AUTHORS

Dr. Rajiv Berry
Dr. Dhananjai Rao

CATEGORY

Computational Chemistry & Materials

SYSTEMS USED

Not listed

SECRET

No