Hi. I am calculating reaction pathway based on AFM/STM experiment. The experimental setup includes I=0.5pA and V=0.2V and the bias voltage is increased to 2V. So here how will I define the electric field value in terms of eV?
More details concerning your calculation might be needed to be sure about the answer, but usually the conversion is simply one to one (1eV corresponds to 1V). That is because 1eV is the amount of energy by which the energy of en electron changes when the electron moves between two places which differ by 1V in their electrostatic potential. Be careful about the sign here (electron carries a negative charge and also, various DFT codes may differ in their respective sign conventions used when including en electric field in the calculation).
I am calculating the reaction pathway here. The experiment in STM is carried out with the bias voltage to 2 V (the voltage varies from 0.2 to 2) for up to 5 s. I am replicating this in DFT using vasp code. I am confused how to set this EFIELD parameter in the incar file of vasp. As you said above, if the bias voltage is 2V, the the electric field value will be 2eV during the vasp calculation. Am I right?
Thank you for the reply. Martin Ondracek
Follow this link and remove your confusion
https://www.rapidtables.com/convert/electric/volts-to-ev.html
Shakeel Ahmad Sofi thank you
how to find the charge flow (Q) as given in the equation for my problem?
Dear Rahul,
first, the problem is that the electric field is not the same thing as the bias. The field is essentially a gradient of the electrostatic potential, while the bias corresponds to the difference of electrostatic potentials between two places. So the unit of the electric field is V/m (in SI) or volt per angstrom (in VASP). The electric field should be such that it integrates (as a gradient) to the requested bias when integrated across the whole vacuum gap between the STM tip and the surface under the tip.
Second, do I understand correctly that for the reaction you study, you need to include both the STM tip and the surface below in your calculation at once, as two indispensable parts of one big system? Like, for example, does the reaction path involve a transfer of some atom or molecule between the tip and the surface? If so, it might be really difficult to model this (at least in VASP) and I am not sure if I am able to give any useful advice. The trouble is, you need to force the electric filed to appear in between the tip and the surface, but when EFIELD is set in the INCAR file for VASP, it only guarantees the field at the boundary of the supercell slab. In case no one with more experience appears in the comments to help, I can try to cook up some suggestions for this situation, but no success guaranteed.
If, on the other hand, the reaction takes place on the surface only (or on the tip only) and you need the opposite electrode for the sole purpose of introducing the electric field, the modeling should be doable. You can include only one of the electrodes in your calculated structure and estimate the EFIELD you need as the bias divided by the distance between the two electrodes (tip and surface). This way, you may miss some inhomogeneity of the field caused by a very sharp tip (especially if the electrode you are including explicitly is a flat surface opposite to the tip), which may or may not be a serious problem.
Anyway, the sign problem: I think if your bias is defined so that the positive electrode is placed at the bottom of the supercell (smaller /more negative/ coordinate along the z axis) and the negative electrode further up (larger /more positive/ coordinate along the z axis), EFIELD should be set negative. But please better check this independently of me if you really need to know.
Martin Ondracek Thanks for very deep insights. Again it’s my fault I haven’t given enough information. For now I am not modeling the stm setup in vasp. i am considering only the implicit field. For example, I am trying to study the bond breaking barrier in the presence of a efield. So as you said earlier, if the unit in vasp is eV/m, what will be the the EField parameter in incar file for the given experimental condition?
once again I thank you for giving me a detailed answer.
Rahul Suresh : Well, V/m is the base unit in the international system of units. In VASP, the unit is V per angstrom. The field specified by EFIELD should be the bias divided by the distance of the tip from the sample. So, for example, when the bias voltage is 0.2 V and the tip distance is 6 angstrom (those 6 angstrom seem to be a reasonable estimate to me, given the I/V values you quote, but you should definitely try to get a more accurate number if you have a way to do so), then the corresponding electric field (specified as EFIELD in VASP) will be 0.033.
Once again, be careful about the sign. One way to check it can be this: Compare the electron density (output in CHGCAR) for EFIELD=0 and for whatever non-zero EFIELD you have. The electrons should be pulled out of surface a bit compared to the zero-field case if the sign of the field corresponds to the negative surface polarity, while the electrons should be pushed more back inside the surface if the field corresponds to the positive surface polarity.
Also, when specifying EFIELD in INCAR for VASP, do not forget to also specify the direction of the field vector with IDIPOL=3 (if the field is along the z axis), LDIPOL=.TRUE. (for so called dipole correction) and perhaps also the center of your supercell by DIPOL (the last parameter can be important because, as the electrostatic potential in VASP is always periodic, the electric field needs to have a discontinuity somewhere, and you definitely want this discontinuity to appear somewhere safe into the vacuum, at the boundary of the supercell, rather than somewhere inside the sample material).
Martin Ondracek Again, but in stm the distance is normally in nm. here, the tip is positioned at a lateral distance 1-3 nm from the center of molecule and retracted 2-3 A from the stm set point for 0.2V. With this condition the voltage is increased to 2V. So, if I want to set the field for bias voltage 2V, then 2V/3A = 0.6666 eV/A
I will check with sign more carefully.
Rahul Suresh Yes, those 0.66 eV/A seem reasonable under your conditions. No need to be too precise with the number, because the true electric field in the experiment will be somewhat more complicated compared to the idealized field in VASP anyway. In particular, the lateral shift you mention probably introduces some lateral (parallel to the surface) component of the electric field. But trying to introduce such lateral field into the model when calculating with VASP is extremely troublesome, so I definitely recommend to settle just with the perpendicular-to-surface component to approximate the field.
