Learning Forum

[hamid.h.javadi]

Perhaps you know the answer (to this question) already.

I have read a paper discussing coupling of a plane wave to 2D grating with a new caveat; two plane waves are incident (with different oblique angles to the normal direction) on a horizontal plane containing the 2D plasmonic grating (parallel lines of thin gold lines with the unit cell consisting of individual elements with difference phases providing a linear phase gradient along a specific direction). The plane waves interfer to generate diffraction pattern with spacing similar to the period of the plasmonic grating. The interference pattern (of the two plane waves) moves along the 2D plane (of the plasmonic grating) with time (due to slight differences in their frequencies). The k-vector component (on the horizental plane) is cancelled with an opposite phase gradient of the grating lines within the unit cell. This is called space-time coupling of the plane waves to 2D plasmonic grating. Can you provide insights on how to simulate this configuration in Ansys Lumerical FDTD:

1- two plane waves with two distinct frequencies (slight offset) imging on a surface in an angle to normal:

i- plane wave type: diffracting vs Bloch/periodic vs BFAST

ii- plane wave frequency: CW vs short pulse envelope

iii- boundary condition (Bloch vs periodic)

iv- grating phase gradient using different lengths of the grating lines vs resonating patterns

v- polarization

vi- transmission and reflection 

2- calculating bandstructure and identification of forbidden transitions (final state located below the light line in the energy-momentum space)

3- calculating coupling efficiency

Appreciate your help.

[Guilin Sun]

Two-plane-wave source simulation is doable, with some strategy.

since they have different spectrum, and the result is normalized to the source (Understanding frequency domain CW normalization), I would suggest that you do two separate simulations, and sum the result coherently. For angled incidence you can use BFAST and if the angle is large, say 70 deg or above you will need to use Bloch BCs in the plane of tilt to sweep wavelength. 

Lumerical FDTD always uses a pulse for the cause of simulation efficiency, except for rare cases where a CW signal is required.

Polarization is what you want to simulate. It is up to you to choose. Usually s and p polarization may have different responses.

Once you get sum of the coherent result, using gratingpolar script to get the amplitude quantities: https://optics.ansys.com/hc/en-us/articles/360034407034-gratingpolar-Script-command

you can get the summed power transmission/reflection for each order. Please make sure that you added them from the same diffraction order, which is in the farfield.

If you want the fields inside the grating, you can use script to get the E fields from a monitor covering the region you desire.

Band structure simulation is relatively easy. Please refer to the photonic crystal examples: https://optics.ansys.com/hc/en-us/sections/360006919074-Photonic-crystals

As long as you know how to quantify the coupling efficiency with a formula, eg, transmission=transmitted_power/source_power, you will be able to extract it from simulation.

Since your questions are quite comprehensive, I would suggest that you do it one by one, and write new post for each question so to isolate posisble issues. Otherwise it might lead to some confusions. Please try the easy one first.

 

 

 

[hamid.h.javadi]

Hi,

I need some guidance of how to correctly setup a nonlinear optical model involving both FDTD and DEVICE. The answer may be in plain sight but it is hidden for a novice.

In FDTD, Two (2) CW optical sources are incident on plasmonic grating on GaAs substrate thus generating electron and hole pairs. The applied voltage between electrodes accelerates the charges. At the same time, other parameters of the DEVICE model are also in play (mobility, diffusion, recombination,…)

Nonlinearity is approached by modelling cyclic perturbations (optical source à e-h pairs, that are moved under the influence of DC voltage à change of index of refraction à impacting optical response).

I need to know how to correctly setup FDTD and DEVICE models.

1. FDTD generates e-h pairs by using the analysis script “CW generation rate“ in “Optoelectronics - CW generation rate” in Object Library 
2. eh concentration is saved in “export file" name in (.mat) format. source intensity is set here.
3. DEVICE loads the e-h concentration under “CHARGE - Optical Generation Rate”. File name in (.mat) format.
4. DEVICE applies DC voltage specified under the "boundary conditions". a charge monitor saves n-p data in a new (.mat) file.
5. In FDTD, “NPDensityGridAttribute” is added as analysis structure to import n-p concentration in (.mat) format.
6. In FDTD, the new material is added to include "np density". The original material is replaced with a new one that uses "Drude model" to calculate change in the index of refraction.
7. FDTD --> DEVICE/CHARGE --> FDTD --> DEVICE/CHARGE is run in cycles till convergence is achieved.

Does this setup guarantee full account of the intended nonlinear model?

Thanks,

Hamid Javadi