What is the effect of the bias voltage on metal electrodes and screening region?

[zhangguangping]

Dear ATK staffs,

When I apply a bias voltage on the two electrodes of a sinlge-molecule junction, comprised of a single-molecule and two metal electrodes, what is the substantial change for the metal electrodes and the metal atoms in the screening regions?

As I understand, there is no change in the two metal electrodes except for a relative shift of the Fermi energies of the two metal electrodes. However, the electron populations, the DOS and anything else are the same with that at equilibrium. But for the metal atoms in the screening regions, they will suffer a electric field arising from the bias voltage. So, after the SCF for electronic structures, the electron populations, DOS and other quantities are also correspondingly changed. The aim of electron redistribution in the left and right parts of screening regions is to resist the electric field arising from the bias voltage. If so, the atoms in the screening region are no longer neutral. Taking the screening region which is adjacent to the electrode with a positive bias applied as an example, the atoms in the screening region close to the single molecule will be positively charged while the atoms in the screening region close to the electrode-screening region interface will be negatively charge. That is, the atoms at the two ends of the screening region are oppositely charged forming a double electrical layer.

So, if I want to use a cluster of metal atoms to approximately represent the bulk electrode, can I artificially set a double electrical layer by doping function (positively doping and negatively doping at the two ends of each atomic metal cluster) in ATK to mimic the effect of bias voltage on the metal atoms close to the core molecule (this will also promote the electron transfer between metal atoms and core molecule)? And by this way, I can even explore the spatial distributions of ions in the surrounding solvent since their spatial distributions certainly affect the energy level alignments between the core molecule and electrodes in the junction.

Please give some comments and guides. Thanks so much.


With best regards,

/Guang-Ping Zhang

[Anders Blom]

Hm, so effectively just creating a dipole in the metal? Not quite the same thing... Note that there is also non-conservation of the number of electrons in the central region so the charging of the molecule occurs not just from the neighboring metal atoms, but indeed from the infinite reservoirs.

Maybe it would be more efficient to consider a SurfaceConfiguration instead, which is a one-probe device with only one metal electrode, and you can run the potential of the metal to induce charge transfer to/from the molecule?

[zhangguangping]

Hm, so effectively just creating a dipole in the metal? Not quite the same thing... Note that there is also non-conservation of the number of electrons in the central region so the charging of the molecule occurs not just from the neighboring metal atoms, but indeed from the infinite reservoirs.

Maybe it would be more efficient to consider a SurfaceConfiguration instead, which is a one-probe device with only one metal electrode, and you can run the potential of the metal to induce charge transfer to/from the molecule?

Dear Anders Blom,

Thanks so much for your  kind reply. In fact, I want to optimize the geometry of the junction under the bias. However, for a two-probe system, this has not been realized and it also will time consuming.

With best regards,
/Guang-Ping Zhang