I *believe* that I'm having an issue in Sentaurus TCAD, where enabling
Recombination (surfaceSRH)
in the physics definition of the material interface is not sufficient for true behaviour at that interface. What seems to be happening for the interface of Material1/Material2 is that:
*Holes and electrons on Material1 side recombine;
*Holes and electrons on Material2 side recombine;
*Holes from Material1 do not recombine with electrons from Material2;
*And vice versa.
Without carriers recombining across the interface, 'cliff like' band arrangements do not produce an increase in surface recombination with respect to 'spike like' band arrangements.
Does anyone know the correct manner to go about enabling heterojunction interface recombination in Sentaurus, so that it actually behaves physically?
I'm looking more deeply into this, and I can definitely see that the surfaceSRH is ignoring the higher electron population on the buffer layer side. This causes the surface recombination R(V) to *decrease* as the cliff increases, and the population of holes in the absorber at the interface, p(z=0) becomes larger than the population of electrons in the absorber at the interface, p(z=0)>n(z=0).
The net effect is that the inversion on the absorber side for large cliffs, p>n, z=0 causes an *increase* in Voc, because recombination is not being calculated correctly.
So I definitely need either a way to get surfaceSRH to respect the populations on both sides of the interface, or a way to define a custom recombination R(V)=p(V,z=0,absorber)*s(v), for s(v) the recombination velocity for holes.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators