n-GaN: -2.5<z<0 electron mobility=200*(300/T)
undoped GaN: -4.5<z<-2.5 electron mobility=440*(300/T)
Can you describe the two epi-layers to n-Contact layer with function tab?
1. Sure, using Function tab one you can describe inside n-contact layer as many layers as you want. However, please aware of the vertical grid step specified in the configuration tab. The properties are calculated in the cell center and then used for the whole cell volume. So rough grid may smooth fine fetures described by the user function.
2. For your particular example, the difference in the doping level is much more important than the difference in mobility. My personal opinion is that one can simply ignore undoped n-GaN. However, if the user wants simulate exactly his own structure, he should do the following
- Make two script functions for doping and mobility of the n-contact layer.
- Make sure that the boundary z = -2.5 coinsides with boundary between cells. Otherwise, one will get changed the n-GaN layer thickness, and this effect will be much more pronounced that contribution of the undoped GaN to the conductivity.
Please note, that for planar LEDs the vertical grid step may differ from the (a/Na) value from the Configuration tab because of the mesa.
To check the vertical grid boundaries one need look into result *.cgs file and choose 'z-planes' mode. To check the grid, one need not wait for completed computations, one may stop computations just after few iterations.
