A PML for silver (drude model ) in FDTD

I'm trying to implement a PML for a 1d simulation domain of silver (1d code for testing purpose), which is modeled as a drude model. I've used the PML for dispersive media as introduced in Gedney's paper, where you introduce a P term and insert the sigma from silver as eps_r(w). And for PML, i just used the basic formula from Taflove's book with an order m=4 and my PML is not absorbing properly. I guess there's something wrong with my parameter for PML, i.e. i didn't vary the sigma_max for silver. I tried to search paper on this topic but it seems people don't give much detail on how they implement the pml in their papers. Could anyone give me some idea what i did wrong?

Thank you so much!

Hi,
You should grade conductivity (as well as non-unity parameter k if you use UPML) , either polynomially or geometrically. I attached matlab file for 2D TMz mode with TF/SF + UPML termination.Although it is 2D you can convert 1D easily. If you have any question, please feel free to contact with me.

Kind Regards,
Alper

Thanks a lot for helping me out and attach the code.

So far i figured my code has problem only with very fine mesh and shrinked simulation domain (i.e. no 1 lambda distance b/w source and pml, etc) but works for a more coarse mesh. I'm still working on it.