About the gate calculation

[tomato151]

Hi， @zerothi, I try to calculate the electronic transport properties of a graphene nanojunction with a charge gate using the Geometry.Charge block and I have two questions about the calculation details.

• Should the option SlabDipoleCorrection be chosen? In the Method part of your paper published in PCCP the slab dipole correction was used, however I did not find this option in the fdf file.

• If I put a charge -e in the gate plane, how to induce a charge +e in the graphene nanojunction to keep the whole system (the graphene nanojunction + gate) neutral?

Thanks for your kind help.   Regards, Haoyang

[zerothi]

1. Yes, you should use the dipole correction. They were also used for the calculation in mention.
2. The system is charge-neutral in total when using Geometry.Charge, as is shown in the paper in question, you add a charge g on the gate, and remove the same charge from the system.

[tomato151]

I notice that if the neutral system has 1148 electrons, when using

%block Geometry.Charge
 plane -2.0 charge
   gauss 1. 2. Ang
    0. 20.0 0. Ang
    0. 1. 0.
%endblock Geometry.Charge

There is a line in output transiesta: Charge distribution, target = 1150.00000 How do the program realize this? Is the Fermi level of the electrodes shifted?

Thank you!

[zerothi]

2. The system is charge-neutral in total when using Geometry.Charge, as is shown in the paper in question, you add a charge g on the gate, and remove the same charge from the system.

In the above case you put -2 electrons on the gate, and then add 2 electrons to the system. Effectively charging the system, while the total simulation cell is charge neutral (due to the gate).

[tomato151]

Thanks, I have understood.