Learning Forum

[HAOYU LIU]

I want  to test photocurrent pf photodiode in the charge but my struture has air holes in the absorption layer. Without holes, the simulation can succeed. But when I add the holes, "Refinement limit exceeded for optical generation rate constraint. The mesh is valid but may not be adequately refined", "APPLICATION TERMINATED WITH THE EXIT STRING: Hangup (signal 1)" errers appear. How to fix this?

[Guilin Sun]

This indicates that the simulation needs finer mesh to resolve the details. You may beed to increase the iteration limit. Please also do participation and check the mesh. If the mesh is too coarse, you may also need to use mesh constraint or make the min edge length smaller. 

[HAOYU LIU]

when I set the min mesh to 0.001μm, it still have the error of APPLICATION TERMINATED WITH THE EXIT STRING: Hangup (signal 1). Use about 6 hours for tessellating and go error. How to fix it? This structure has no meaning for 2D simulation. My design is adding some air holes (material: air) in the Ge absorption layer.

[HAOYU LIU]

Because air is not semiconductor, can this be simulated?

[Guilin Sun]

You can try to increase the number of iteration, and local mesh constraints. Do not use so small the edge length for the whol region.  If you have thin layers, please make sure the thickness is mandatory otherwise you can use a larger thickness.  Ait holes might be the main issue: air as insulator does not involve the simulation since no charge there. You can  diagnose the issues by first disabling all the holes and see if the Ge layer is working. Then gradually increase the number of holes.

 

[HAOYU LIU]

For the materials of nanoholes, I find I can choose none for this structure group, does it mean in this area there is no simulation for the chosen area?

My intention is cut off a part of material in the photodiode, in the simulation it means this part will not participate in the simulation.

[Guilin Sun]

No. Usually you can either input the index or the material from the material daabase. You can read the script inside.  I am not sure if "None" will work.

As mentioned before, dielectric material is insulator, it will not be meshed and simulated.

[HAOYU LIU]

For transient to test the bandwidth, the recombination parameters will not affect too much. But when using the ssac to test the bandwidth, the recombinations count a lot. For some situation, the bandwidth will decrease as the bias voltage increase and the UTC-PD bandwidth will decrease when using the graded doping in the absorption layer, it violate the principle. I want to know how to set the appropriate parameters in the ssac, the parameters in the example of vertical PD are not suitable for my model.

[Guilin Sun]

This is a different question. I suggest to write a new post. I will ask my colleague to reply you in the new post.