Some Basic problem from atk manual

[renren123123]

Hi, evergyone:
    Today, I studied carefully atk manual, and I have some problems as following: (1) from Fig1 and its notation as "By comparing to the MPSH eigenstates, we can find that the n=0 eigenstate arises from transmission through MPSH state 16. Since the state has a nodal plane it couples very weakly with the electrodes, and gives rise to a narrow peak in the transmission spectrum.
The n=1 eigenstate also has 2 nodal planes, however, does not have a direct resemblance with neither MPSH 15 or 16, which are the MPSH states with 2 nodal planes. Thus, the n=1 state must therefore be a linear combination of MPSH 15 and 16. This state has a stronger coupling with the electrodes, and gives rise to the broad peak in the transmission spectrum around 2.4 eV.
I want to ask: how to look a, two or more nodal planes?
(2) "border_atom = len(device_configuration.elements())/2" in the pathways is fit for other all cases, how to understand border_atom?  why it equals half of all atoms?
(3) how to understand "ElectronDifferenceDensity, returns the electron difference density, i.e. the difference between the self-consistent valence charge density and the superposition of atomic valence densities."  how to compare self-consistent valence charge density with superposition of atomic valence densities?  self-consistent valence charge density denotes self-charge for considered atom? but superposition of atomic valence densities donotes other atoms provide charge with considered atom? is it right? thank you very much!

[Anders Blom]

1) This should have been written "By comparing to the MPSH eigenstates, we can find that the n=0 eigenstate arises from transmission through MPSH state 16. Since the state has a nodal plane in the transport direction, it couples very weakly with the electrodes, and gives rise to a narrow peak in the transmission spectrum."

The nodal plane is simply the plane in the middle of the molecule where the eigenstate has a node, i.e. goes to zero.

2) This is from the reference manual (which took some time to understand...). I don't think you should bother about this at all.

3) For Huckel we have no immediate way to compute the full electron density, but there is a way to easily define the induced change in the density compared to the "neutral atom system", i.e. the case when the atoms are just put together but no interactions between them is turned on. For this "difference density" we can solve the Poisson equation, and obtain a Hartree potential, which we call the ElectrostaticDifferencePotential. It describes the electrostatic potential landscape that causes changes in the electron charge distribution compared to the neutral system, and thus describes the physics behind bonding, voltage drop, and other relevant effects.