d-SEAMS

d-SEAMS analyzes molecular dynamics trajectories to characterize nucleation processes in supercooled water and classify ice-like structures.

Key Features:

• Topology Analysis: Elucidates the structural evolution of ice-like formations within simulation trajectories for both strong-confinement and bulk systems.
• Innovative Algorithms: Implements novel algorithms for confined ice structure determination and topological network criteria tailored for bulk ice analysis.
• Order Parameter Development: Provides a new order parameter specifically designed to identify building blocks of quasi-one-dimensional ice.
• Implementation and Performance: Implemented in C++ as a High Performance Cluster-enabled postprocessing engine for molecular dynamics trajectories.
• Scripting and Build Pipeline: Extends functionality via a Lua scripting interface and uses a YAML-Lua scripting pipeline with nix for reproducible builds.
• External Integration: Supports integration with external tools and libraries, including R.

Scientific Applications:

• Heterogeneous ice nucleation on silver-exposed β-AgI: Applied to analyze structural time evolution and nucleation metrics on a silver-exposed β-AgI surface.
• Homogeneous ice nucleation: Used to characterize homogeneous ice nucleation in supercooled water trajectories.
• Flat monolayer square ice formation: Used to detect and analyze formation of flat monolayer square ice.
• Ice nanotube freezing: Applied to analyze the freezing process of an ice nanotube.

Methodology:

Performs postprocessing analysis of molecular dynamics trajectories using topology-based classification, confined-system algorithms and bulk topological network criteria, computes a dedicated order parameter for quasi-one-dimensional ice, and is implemented in C++ with Lua extensions and a YAML-Lua/nix build pipeline on HPC systems.