What is the meaning of the two value in the transmission coefficient?

[rosen]

when we do the calculation of the transmission coefficient in two probe system, the result will give two value, such as
 Transmission Coefficients: [ 4.43261237e-004  5.75744413e-014]
what is the meaning of them?
sometimes I do the voltage calculations with the same magnitude but different directions, the sign of the current change, and the value of the current is in the same order, but the transmission coefficients will changed with 2 order?
for example, the system with a (0.1 -0.1)Volt, the current is 2.6e-9A, the transmission coefficient is [2.12e-3  3.98e-17]
but the same system with a (-0.1 0.1)Volt, the current is -1.55e-9A, the transmission coefficient is [1.50e-5   1.66e-8]?
from the current value, it seems that the resistant change slightly when the bias is opposite, but from the transmission coefficient, it change dramatically, which one shoud we based to calculate the conductance of the two probe system?

[Anders Blom]

The two values might be because of spin up/down, or k-points. Hard to say without seeing the input script.

The fact that the transmission/current is different in two directions is the whole operation principle of a diode

[rosen]

In the system, the scattering is an isulator, and the two electrode is different metals, I think the difference of the work functions in the two electrode is the reason of the dipole behavior in the transmission coeficient. But I still don't know how to define the conductantce, form the I/V or from the transmission coefficient/spectrum?

[Anders Blom]

The differential conductance you get from the I-V curve, but it's quite costly to compute. At zero bias, you can define the conductance as being proportional to the Fermi-level transmission (this is exact in the limit of zero temperature).

[rosen]

then how to find the fermi level in the two probe system? using the "projected hamiltonian energy spectrum" with all the molecule atoms included?

[Anders Blom]

In a two-probe system, the average Fermi level is defined as 0 eV.

[rosen]

Ok, thank you for your answer