A parallel algorithm to produce long polymer chains in molecular dynamics.

Area of Science:

• Computational chemistry
• Materials science
• Polymer physics

Background:

• Generating initial polymer melt configurations for simulations above entanglement molecular weight is computationally challenging.
• Existing methods often struggle with efficiency and scalability for large systems.

Purpose of the Study:

• To adapt a chemical polymerization-mimicking algorithm for all-atom force field simulations.
• To develop a parallel and efficient method for generating initial polymer melt configurations.

Main Methods:

• Sequential bond creation and relaxation from a monomer bath.
• Parallel implementation leveraging classical molecular dynamics code structure.
• Adaptation for all-atom force fields.

Main Results:

• Generation of large systems up to 7 million atoms.
• Linear scaling of simulation time with chain length for long chains.
• Production of polymer chains up to ~2000 carbons with accurate thermodynamic and structural properties.
• Near-equilibrium chain conformations achieved within picoseconds of simulation.

Conclusions:

• The proposed algorithm offers an efficient and scalable solution for generating initial polymer melt configurations.
• The method produces high-quality polymer structures suitable for further simulation and analysis.
• The algorithm's versatility allows for the creation of various polymer architectures, including copolymers and blends.