Skip to main content

SNAP (Simulated Nanoparticle Assembly with Proto-particles)

Posted by: Amanda Barnard

August 22, 2016

Simulating polydispersed mixtures of faceted nanoparticles just got a whole lot easier.
Simulating polydispersed mixtures of faceted nanoparticles just got a whole lot easier.

The understanding of nanoparticle self-assembly and aggregation has applications in fields as diverse as nanophotonics and drug delivery in medicine.  Therefore, a modelling methodology of the growth and stability of these large structures under different environmental conditions is vital in their understanding.

While nanoparticle interactions can be simulated numerically by Monte Carlo or Molecular Dynamics methods, these approaches are limited in practice, since they scale with the number of atoms in the system, and each nanoparticle may contain many thousands of atoms whose mutual interactions need to be calculated explicitly at each step.   We can model the structure and properties of individual particles very well, but simulating the global properties of large aggregates remains challenging, and the impact of size distributions or heterogeneity remains largely unknown.

Simulated Nanostructure Assembly with Proto-particles (SNAP) is a multi-scale modelling package under continued development which represents an atomistic nanoparticle as a coarse grained surface point mesh.   Individual points may represent regions on surfaces with similar features, enabling researchers to define complex combinations of sizes, shape and facet combinations, each with specifically defined interactions.

These interactions are entirely user-defined, and implemented with fast, two body potentials individually parameterized using pre-calculated data from higher order electronic structure simulations.

The unique feature of SNAP is the ability to replace large nanoparticles containing many thousands of atoms, intractable to convention molecular dynamics modelling, with fast and efficient surface meshes that interact via very simple interaction potentials. Why simulate the interaction of 10 nanoparticles containing 10,000 atoms each, when you could simulate the interactions of 10,000 proto-particles in a fraction of the time?

The code is parallelized using MPI/CUDA for both CPUs and GPUs, and is being continuously updated.  A generator for the nanoparticle surface meshes and initial configuration, along with post simulation analysis tools of the resulting aggregate are also included in the package.

For more information, contact the Principal Developer, Dr George Opletal.