Hi,
I assume this issue (compare this post) should affect the simulations of most non-CMOS pixel and strip sensors, where the bulk is in direct contact with oxide. It should be especially prominent in simulations of strip sensors because of the large fraction of bulk-oxide interface. The geometry Peilin is using is very similar to a strip sensor where 660 nm laser light is injected between two strips.
It would be really great if it would be possible to define regions between the collecting implants where free charge carriers are not HALTED at the surface, but are reflected. For geometries where the lateral electric field (parallel to the interface) is weak there might be a large fraction of charge loss at the bulk-oxide interface. A rough modeling of charge loss at the interface could be achieved by giving the user the ability to adjust the reflection probability, so carriers which are not reflected are still HALTED (probability=0 would correspond to the current state as far as I understand).
A possible temporary workaround would be to turn off diffusion if this is possible in allpix?
Cheers
Christian