Transmission question

[simCity]

Hi all,

I just wonder why the simulated transmission spectrum decreases in most of the structures in ATK when the applied voltage increases. I tried for nanotubes, graphene and molecular junctions. The transmission decreased for all these.

Thanks in advance.

[Anders Blom]

Basically because the more asymmetry you add to the system, the more potential scattering you introduce.

[simCity]

Thanks a lot. However does ATK model the scattering effects?

[Anders Blom]

Elastic potential scattering yes, sure, otherwise the transmission spectrum would just be 1 for every mode in the electrode (at zero bias, at least, and equal electrodes).

[simCity]

Thanks again, moreover in ATK is there any other way of visualizing potential scattering effects other than looking at T(E). I mean: can the change effective potential as the applied voltage increases, or some other parameter also give opinion, at least qualitatively, on the potential scattering mechanism? So that the variation of the transmission spectrum for different devices as the applied voltage increases can be interpreted using simulations.

Thanks a lot.

[Anders Blom]

You can at least compute the effective potential and compare it at zero and finite bias. You can subtract them to see the potential change induced by the bias.

[simCity]

I'll try, thanks.

[simCity]

Hi again,

I also thought to check the variation of the average electron density along the transport direction. I tried to modify the code you gave in one of the previous topics: http://quantumwise.com/forum/index.php?topic=2389.msg11132#msg11132 but unsuccessful to modify properly 

Any help would be appreciated. Thanks a lot.

[Anders Blom]

In ATK 13.8 there will be a special tool for very easily plotting 3D quantities, like the density, averaged in XY along Z (or other combinations).

[simCity]

You can at least compute the effective potential and compare it at zero and finite bias. You can subtract them to see the potential change induced by the bias.

Hi again,

I did calculate the potential profile along z direction as attached. Applied voltages are 0.25 and -0.25V at the leads. This is for a metallic CNT. It seems like the potential jumps to 0.3V near the left interface and to -0.27V at the right interface. Considering the applied voltages are 0.25V and -0.25V, is there an error becase the potential jumps higher than the applied voltages?

Thanks in advance.

[Anders Blom]

The ballistic transmission calculation for a perfect system without any scattering is not meaningful, and that's why you get this picture. If you introduce e.g. a Stone-Wales defect in the middle or something, you will see that the voltage drops at the defect, as expected since it acts as a scattering center.

The transmission spectrum for a perfect 1D conductor is just a step function anyway, and can be deduced trivially from the band structure. Cf. http://quantumwise.com/publications/tutorials/mini-tutorials/167.

[Dipankar Saha]

Sir,
I did not get what you tried to mean by saying that.......
"The ballistic transmission calculation for a perfect system without any scattering is not meaningful...."

[ams_nanolab]

Are phonon scattering taken into account? How, can it be implemented in ATK?

[Anders Blom]

Electron-phonon interactions are not included .