Learning Forum

[jiglesi1]

Greetings,

I am comparing effects of using constant real index vs a detailed material index for lnP on optical properties in Ansys FDTD. I am still new to Lumerical so I am hoping to get some some guidance for this. What are the steps to simulate the lnP layer using both a constant index and dispersive index from the material database.I attached pics of my design and current plot. Any advice in doing this is very welcome. 

[jiglesi1]

I ran the simulation and got the following plots shown below. I ended up using germanium as a dispersive index. According to my plots, the germanium, shown in red here, has less reflection than the lnP with a constant real value since it is closer to the 0 on the y-axis. Is my logic correct,and this is also why the germanium has more transmission, shown in purple. Is my logical deduction correct?

[Guilin Sun]

When you use a fixed refractive index across the source spectrum, it means all the wavelengths will not experience any dispersion. eg, for the same incident angle the exit angle is the same, if we use geometric optics to explain it.

Dispersive materials, on the other hand, will lead to different interaction.

As for the peak-valley, it depends on the phase change due to the slab. for the same optical path length, eg, fixed refractive index, the phase is different since phase is 2pi*OPL/wavelength. 

If you know a material is dispersive, you should use the dispersive model.

For thin film type devices, as suggested in another post, you can use stackrt to get accurate reflection and transmission, as long as you use the correct material model.