Regarding replacement of atoms of one element by another element

[nanomtj]

Dear Sir,

I am working with magnetic tunnel junction builder (Fe/MgO/Fe). As per my design requirement, I have to replace the oxygen atoms in the central region with Magnesium atoms manually to make the central region be made up of Magnesium alone. Can you please help me in identifying how to replace one atom by another? Appreciate your help.

Regards,
Mayank Chakraverty.

[Anders Blom]

Trivial in the Builder in VNL - hold down shift and double-click an oxygen atom, then use the periodic table tool to change it. Or build the structure using the Magnetic Tunnel Junction tool (install it via the AddOn Manager) and specify Mg from the beginning. Or search and replace in the Python script.

[nanomtj]

Dear Sir,

I tried holding down the shift button and double clicking the atom to be replaced. It is not working. Can you please suggest another means by which I can change the atom with the atom of another element? Your help in this regard would be really helpful.

Regards,
Mayank Chakraverty.

[Anders Blom]

VNL version? Need 12.2 at least, otherwise the search/replace approach will be the easiest.

[nanomtj]

Yes, I am trying to do the replacement of atoms using 12.8.2, the current version.

[Anders Blom]

Should definitely work in 12.8.2 - I use it myself every day. Double-click alone should select all atoms - Shift+double-click all atoms of the clicked element. Are you sure you don't double-click too slow?

You can also go to Select>By expression and enter "e=Mg".

[nanomtj]

Not able to make it happen Sir.

I do it this way :
1) Open VNL and open custom builder
2) Open the Magneto Tunnel Junction builder from Custom Builder
3) In the tunnel junction builder, the electrode specifications are set. The right and left electrodes are set to be Fe while the central oxide is set to be Mg to give a Fe/MgO/Fe structure. In this step, I tried going to the right hand side of the screen where I can see the figure of the tunnel junction and shift-double click on the individual atoms but it does not work.
4) After setting all the spacing and other dimensions right, I open the script generator. I set up the calculation. In the same manner, I go to the right hand side of the screen and try to change the atoms but it does not get changed.
5) I try to do it using the Initial State but does not work that way either.

Need your help in getting this fixed. I am trying to do this using the demo version. Is there any restriction in the demo version that I cannot carry out this kind of replacement? Please help.

Regards,
Mayank.

[Anders Blom]

And here's how you should do it

Send the MTJ structure from the Custom Builder to the Editor and from there send it to the Builder. Now you can do the operations I indicated.

[nanomtj]

Thank you Sir.   

I think I will be able to work it out now. I have two more queries:
1) While working with Script Generator, the New Calculator has an option of Device Algorithm Parameters. When we open Device Algorithm Parameters, there is an option to select the device algorithm (real) as Recursion, Krylov and Direct. Will there be any difference in the I-V characteristics of the device if we use Recursion or Krylov algorithm? Which algorithm would work for a P-doped Silicon/Intrinsic Silicon/N-doped Silicon system built using the magnetic tunnel junction builder?

2) While computing the Transmission Spectrum, there is an option to select the Self Energy Calculator from Recursion, Krylov and Direct. Which of the self energy calculator would be relevant for the same P-doped Silicon/Intrinsic Silicon/N-doped Silicon system built using the magnetic tunnel junction builder?

Appreciate your help in this regard.

Regards,
Mayank Chakraverty.

[Anders Blom]

1. Use Recursion, it's fast and accurate.
2. Same.

Normally all methods give the same result, they are just different way to compute the same quantity. But for numerical reasons the results can differ a bit; the Krylov method used to be faster, but after we optimized the Recursion method a bit, it's at least as fast now. so Krylov used to have a performance advantage - but also a disadvantage that it's a bit more approximate than Recursion. So now there is really no reason not to choose Recursion. Direct is basically the same but a bit slower.

[nanomtj]

Thanks a lot for your help Sir. 

Regards,
Mayank Chakraverty.