Learning Forum

[tgp27]

I am trying to use Lumerical FDTD to simulate the reflection and scattering from particle with a gradient refractive index that is partially embedded into a medium of a refractive index n1,k1 , with the rest of the particle surrounded by electrolye of medium n2,k2. I am trying to simulate the widefield image at high N.A. objective.

[Guilin Sun]

Let me summarize your target to be simulated: random particles inside lossy medium, and use high NA to collect the scattering and reflection? is the scattering angular field-intensity distribution or the scattering cross section what you want? before any suggestions are given, I wish to know what exactly you want to get. You may also describe what quantity you get from experiment. Thank you!

[tgp27]

Hello. I would like to obtain the widefield image (intensity as a function of x and y spatial position) that would be observed using a high numerical-aperture microscope objective (at a single wavelength, e.g. lambda = 500nm, but ideally across the whole visible range). I believe this could be obtained using a planar monitor behind the source collecting the far-field back-scattering (i.e reflected) light. Is that correct? Experimentally we obtain microscope images.

[Guilin Sun]

If so, please refer this example https://support.lumerical.com/hc/en-us/articles/360042260213-Simplified-microscopic-imaging
TFSF is a differential technique that removes the direct illumination. If in your experiment you did not remove the direct illumination, a Gaussian source is recommended.
Please note that you are simulating random particles, so depending on the randomness, each time the result can be different if the random particles are different

[tgp27]

Thank you for your suggestion. I have tried that but it doesn't give the expected result. Even in a very simple simulation where I just set it up like the tfsf example I get very weird lines in the image, which don't make sense with the position of the sample and defocus (see image below for changing image with defocus). The particles are not random - the size and position are fixed. The variation is between different defocus values for a given simulation, not between different simulations. Is there a way of using the S-matrix approach (https://support.lumerical.com/hc/en-us/articles/360041768554-Optical-Defect-Metrology-S-Matrix) with widefield imaging, i.e. to do a coherent sum over the output plane waves? The example given there is written for scanning-spot microscopy and not general widefield microscopy.

[Guilin Sun]

Not, the result is not weird, it is true! Now it is smaller than wavelength, so the field distribution will not be uniform. Please refer this example: https://support.lumerical.com/hc/en-us/articles/360042256473
WhenÒÇÇto use S parameter, you will make sure that the result will not change with respect to the source size. I did not see this from your set up.
The methodology works for general microscopy and wide-field microscopy.
Before doing further simulation, take some time to make sure the result is physical and understand the physics.