basic principle of two probe system

[sweta]

Hello,

I just want to know about the basic principle description of theoritical formalism (DFT-NEGF ) that ATK is using to calculate IV charecteristics of two probe system.

Thanx in advance!

[Anders Blom]

See section "More reading, background material, etc" at http://quantumwise.com/publications/scientific-publications. Also, the Reference Manual contains basic information (will be extended with a chapter on NEGF as part of the 12.2 release).

[sweta]

Thank you for your great response!!

Sir,
I want to study two probe system with different electrodes. I have done my study with gold electrode by taking position of anchoring group in fcc lattice (i.e. by setting alignment layer = 2) and all other parameters are taken as default parameters in custom builder molecular junction.

I want to do the same study with platinum electrode by taking same parameters as above but my problem is that when I drop platinum electrode to molecular junction, it increases the total number of atoms above 100 so its difficult to optimize the system.

Kindly help me how can I do the same study with platinum electrode (by taking all other parameters as that of gold electrodes) so that I can test effect of electrodes on our geometry.

[Anders Blom]

Pt has a smaller lattice constant, so if you set the "Minimal length" to 6 Å you should get the same number of atoms as for Au. Of course, 3 layers in the electrode really is minimal, and you are at the edge where there is a risk to have too many matrix elements cut off.

[sweta]

Please see the attached image and tell me is the both images will give same results??
In first image I have chosen alignment layer 2,2 and second image 1,2. I got confused because some article prefer first image and others use second image.
I had taken 2,2. why they are taking different alignment layer for symmetric gold electrode on both sides?

Thanx in advance!

[Anders Blom]

I think in particular older articles would use the first picture since you have the two electrodes identical, which reduces the calculation time (very slightly, though, to be honest). The effect on the results (or at least the conclusions you can draw from the results) should be marginal if it's just a matter of which layer ABC you have as surface. The really interesting part is the actual contact, if you place the gold on a hollow site or bridge site, or even on top.

It's a bit hard to see from the picture, but if you have a different coordination of the S atom to the gold surface, then the results will be different.

[sweta]

Thank you for your great response Dr. Blom!

I also want to work with other electrodes like Al, Li etc but the atoms of these electrodes are not forming bond when we drop it in molecular junction. How we can do our study with other electrodes

[Anders Blom]

Bonds are not relevant in the calculations, they are only used (in ATK) for the visual representation. Of course, sometimes not having bonds means the atoms are too far apart, either by mistake or design, but as long as the atoms are in their proper positions it's completely irrelevant if you see any bonds or not in the pictures.

[sweta]

Hi,
My 2 probe geometry showing the attached IV. Please help me how to interpret this graph. Is it showing NDR or oscillations or something else 

Thanx in advance!

[Anders Blom]

Probably you are crossing molecular levels as the bias increases, however it's impossible to say anything without all the supporting data, and intimate knowledge of the calculations that really only you have. You can look at many of the existing ATK publications regarding how to understand and analyze the I-V curve with the help of MPSH spectra, transmission spectra, LDOS (e.g. http://arxiv.org/abs/1201.6279) etc.

[sweta]

How to show the attached device image in covalent radius form in molecule configuration???
because to show this image properly in my article, when I do this it does not shows electrode region in molecule mode. So how to show full device configuration in covalent radius form???

[Anders Blom]

In the Viewer, right-click the 3D plot and choose Properties. Then you can set the atom size to covalent; you need to do it in each region (left, right, central) separately.

Please don't attach images as Word documents. Attach the image itself.

[sweta]

Thanx Dr. Blom. You are really great.
Is this possible to show scattering region + electrodes in molecule mode??

[Anders Blom]

Do you mean without the unit cell? The unit cell (as well as any other part of the system) can be turned on/off with the "Visible" checkbox under each item in the Properties dialog.

[sweta]

No, actually I want to see whole geometry i.e. electrode region as well as scattering region in the molecule mode of builder. but molecule mode show only scattering region.
So I want to know is this possible to show whole geometry in molecule mode??