Dear Dirk,
One should just be aware that going much below the micrometer scale into the nanometer scale (say below 100 nm), one soon meets conditions where several assumptions made when running a MC simulation (e.g. assuming a homogeneous and isotropic medium) break down. Just to name a couple:
-
Since your question was from a neutronics perspective, it is worthwhile pointing out that e.g. thermal neutron interactions are sensitive to the local chemical/molecular/crystallographic environment, so you can expect nanostructure specificities. More so if the nanostructure has a net magnetization (from possibly unpaired e- spin states in the nanostructure), especially if the incoming neutrons are spin-polarized: the scattering of the latter is sensitive to the local magnetic moment density. If one wishes to be terribly accurate, things complicate substantially beyond what one can reasonably do out of the box with a general-purpose Monte Carlo simulation, needing input from either experiment or density-functional-theory et al (ideally both). Neutronics experts may want to further comment/correct/elaborate, though.
-
The description of charged particles undergoing electronic stopping in the course of collisions with target electrons would need to be revisited: new excitation modes (e.g. surface plasmons, volume plasmons in the nanostructure, etc) other than those of an (infinite) electron gas appear and offer great variety depending on the shape and size of the nanostructure (even near surfaces). This gives rise to a position- and direction-dependent energy loss probability. Granted, these might be low-energy excitations, so swift projectiles might be fine, but better be aware.
The items above can be addressed, but one needs to go beyond a general-purpose Monte Carlo code framework.
Cheers,
Cesc