I am facing an issue with the MD simulations of Ga-Si. I have tried the simulations at 30kV with time-step 0.1-0.002fs, however the simulations show the Ga ion to travel more than 100nm Si depth (the Ga implantation depth in Si is 60-65 nm from literature and SRIM values). The simulations at low kV (1-5kV) are working fine and correspond to the SRIM/Literature values. I have tried hybrid ZBL/tersoff, and ZBL/Stillinger-Weber potentials. I have also tried different lattice orientation, ion incidence angles to account for channeling, in any, but still away from realistic values at 16-30kV. I am wondering if any lammps user ever came across such behavior during the high kV simulations and potential issue causing this problem. Any feedback or suggestion in this regard would be a great help.
Try calculating the resulting electric field when you apply such voltage across the macroscopic unit cell, and compare it to the electric field in the conditions that you are trying to reproduce.
in general, please keep in mind, that people on this list have no idea
about the details of your simulation, so the quality of responses
strongly depends on the completeness and accuracy of your description.
furthermore, it is impossible to assess the correctness of your
simulation.
One difference between SRIM and MD is that SRIM treats the substrate as amorphous
(as I recall) and MD treats it as a crystal lattice. So you can get channeling for incident ion motion in specific
directions, leading to deeper implantation. So you could try
alternate orientations of the substrate.
Or to debug you could calculate how close
a 30kV ion gets to a substrate atom in a head-on collision and verify that ZBL is giving you
a correct force at those separations (use the pair_write command). Or do
a 2 atom simulation with the substrate atom frozen, and shoot a 30kV incident
ion exactly at it and verify it rebounds with 30kV. If not, maybe your time stepping
is wrong for high-energy collisions.
Hi all
For a comparison between MD, SRIM and ACAT see and download the open access articles (using LAMMPS )
Molecular Dynamics Simulations of Platinum Plasma Sputtering: A Comparative Case Study -> Multiscale Molecular Dynamics Simulation of Plasma Processing: Application to Plasma Sputtering -> Enjoy reading Pascal
)