Initial Model Construction of Organic Ligands on Nanocrystal Surfaces:

GPUMD
1

I intend to study the configuration categories and their proportions of organic ligands on the surface of nanocrystals. When building initial models for the first time, what structures can I construct, and what requirements should these models meet?

2

I think you can try to construct the initial training structures using multiple approaches:

  1. Sampling with ab initio molecular dynamics (MD)

  2. MD sampling with an appropriate empirical potential

  3. Perhaps also MD sampling with the NEP89 model ([2504.21286] NEP89: Universal neuroevolution potential for inorganic and organic materials across 89 elements)

After obtaining a set of structures, you can perform single-point DFT calculations to label them and form a training dataset.

Then perhaps you can start the so-called active-learning loop to move on.

1 Like