lcpp-org/RustBCA

[feature] Hofsäss and Stegmaier propose a BCA without surface binding energy

drobnyjt opened this issue · 2 comments

drobnyjt commented

Is your feature request related to a problem? Please describe.
Surface binding energies are problematic in the BCA because they are 1) difficult to measure 2) have to sometimes be adjusted seemingly arbitrarily

Proposed solution
Paper is here:
https://doi.org/10.1016/j.nimb.2022.02.012

Additional context
If this works, it would be a huge deal for BCA models. Removing the semi-arbitrary planar model of the surface binding energy would remove a very significant free parameter from the model

drobnyjt commented

Having read the paper more carefully, some thoughts.

  1. They don't quite demonstrate that the BBE model is superior to the SBE model. In fact, for their silicon sputtered angular distributions, it is significantly worse.

  2. The model requires feedback between distributions (local nuclear energy loss) and the local bulk binding energies. I'm not sure how this works in their code; I think you would have to iterate back and forth to reach some sort of equilibrium... It sounds like their code runs normally in a dynamic mode, with an implied fluence, which is not the goal of RustBCA.

  3. I wonder - if this were combined with attractive-repulsive potentials, would that get better results?

drobnyjt commented

In the end, I don't think this is worth pursuing unless a second paper improving upon the first comes out.