How can I get rid of noisy dips in the tranmission curve?

[jjhskang]

Hi!

I have calculated the zero-bias tranmission for (5,5) carbon nanotube with a molecule adsorbed on it. I find many dips in the transmission. For example, you will find a dip which shows that the tranmittance is zero at E=0 (Fermi energy). I do not know if it is a noise or not. Is there any way in which I can check if it's noise?

I used 80 k-points, 300K, DZP basis set, and default values for other parameters.

For your convenience, I  attach the script for the cvalculation and the tranmission curve resulting from the calculation.

Thanks!

[Nordland]

The best way to test this is to perform the exact same calculation without the molecule.

There you should get perfect integer transmission, and if this is not the case, there is a flaw in your setup, and
my guess would be that it is a too short electrode.

If this gives perfect integer transmission, then it is very likely the correct result that you see.