Silicene nanoribbon configuration

[huckelbuckel]

Hi, i have some basic questions to ask (Please forgive if questions look repetitive  )

Considering decent accuracy of results

1)What value of MonkhorstPackGrid(?,?,?)  points should be used when i want to calculate 'density of states' of a graphene or silicene nanoribbon.

2)What value of MonkhorstPackGrid(?,?,?)  points should be used when i want to calculate 'Device density of states' of a graphene or silicene FET junction.

3)What value of MonkhorstPackGrid(?,?,?)  points should be used when i want to calculate 'transmission spectrum' of a graphene or silicene nanoribbon.

4)What value of MonkhorstPackGrid(?,?,?)  points should be used when i want to calculate 'transmission spectrum' of a graphene/silicene junction.

5)How a gate region must be designed?? , (Considering Z-A-Z...image attached) by default in graphene junction enabling the gate spreads the gate region only in middle of armchair region. i.e, Gate covers only the '-A-' portion of Z-A-Z junction.When i design the gate for silicene or graphene FET should i use this default spread of gate?? or the gate should ideally span all the three (Z-A-Z) regions???

[huckelbuckel]

Ok, one more question to the list 

6) In some papers i have seen people working on tunable band gaps i.e, by increasing the perpendicular GATE electric field (by varying gate voltage) the band gap of the material in question starts to open up or show some variation.When i create a graphene junction  and apply the required bias i didn't see any available 'Bandstructure' calculation module in the analysis tab of script generator.Is there a way (direct or indirect)in ATK-VNL by which i can calculate the bandgap at various gate voltages.

Thanks

[Nordland]

I can answer 1-4 in a single reply.

If your system is a 1 dimensional, you must use (1,1,X), where X is a parameter you will to check for convergence.


Create a small test calculation, and then I would try with x=51, and compare it to x=101. If these are almost identical I would go with 51. If they differ, try to compare 101 to 151, and if they are identical, use 101. If these also differ, continue adding 50 until convergence is satisfied.

[Nordland]

5)How a gate region must be designed?? , (Considering Z-A-Z...image attached) by default in graphene junction enabling the gate spreads the gate region only in middle of armchair region. i.e, Gate covers only the '-A-' portion of Z-A-Z junction.When i design the gate for silicene or graphene FET should i use this default spread of gate?? or the gate should ideally span all the three (Z-A-Z) regions???

It depends on what you want kind of experiment you want to simulate. If you want to simulate a gate under the armchair region only, you should the structure you have. If you want to simulate the entire device being layered on a gate, then you should extend the gate into the Z-A-Z region.

From a physical point of view the two system should be somewhat identical, since the Z regions are metallic and therefore the effect of the gate should be very limited, while the A region is semi-conducting and here the energy bands would be altered.

If you are aiming for showing the FET effect in this junction, I would personally choose the the first system you present.

[Nordland]

Ok, one more question to the list 

6) In some papers i have seen people working on tunable band gaps i.e, by increasing the perpendicular GATE electric field (by varying gate voltage) the band gap of the material in question starts to open up or show some variation.When i create a graphene junction  and apply the required bias i didn't see any available 'Bandstructure' calculation module in the analysis tab of script generator.Is there a way (direct or indirect)in ATK-VNL by which i can calculate the bandgap at various gate voltages.

Thanks

Calculating the bandstructure is, no offence intended, a poor man choice in this case, but it can be good for understanding the physics. If you want to model the FET effect of the systems, you should calculate an IV curve for the the system using different gate voltages, and you will have your final result.

If you are eager to get some bandstructures to show the physics of the system in a simple view, I would build an periodic armchair in the builder with a gate under it, and for this periodic structures you are able to calculate bandstructure in ATK directly.

[huckelbuckel]

Hey Nordland!...thanks for this useful info 

[huckelbuckel]

In some of the technical papers it is written that graphene nano ribbons with spin polarized edges have much lower energy than non-spin polarized.Also anti ferromagnetic spins have lower energy than parallel spins.But in ATK-VNL whenever we give initial spin the optimization routine goes to preferred intrinsic spins of atoms and the total energy of the ribbon always comes out same (with or without spin) 

Is there a way through which i can find the energy difference between non-spin and spin polarization of edge states of ribbon???

Thanks in advance!

[kstokbro]

There is a tutorial on how to perform a spin polarized graphene nanoribbon calculation.
http://www.quantumwise.com/documents/tutorials/latest/GrapheneBloch/index.html/

(The anti-ferromagnetic edge state has the lowest energy, as shown in the tutorial)

[huckelbuckel]

Hi I have gone through this and replicated the same for silicene zigzag nanoribbons but i want to calculate 'TotalEnergy' for both spin polarised (Antispin case) and without spin and calculate the energy difference that exists in two cases as i vary the width of ribbon . but my problem is i am getting same total energy for both schemes??

[kstokbro]

I attach a script with no spin on the graphene and antiferromagnetic spin,
the total energies are:
no spin: -1293.28226 eV
anti ferromagnetic -1293.30601 eV
Thus, spinpolarization lower the energy by 0.024 eV

[huckelbuckel]

Hi, Thanks i have done the energy analysis of 'no spin','anti' and 'parallel' spins
yes, the energy difference is there although very small.
I have given spin to the edges i.e, one edge  has parallel spin while the other has opposite spin.
Thanks for the help.

[huckelbuckel]

I was trying to calculate the band gap variation with electric field of silicene armchair ribbon with gates.
Here are the two metallic regions defined in the script:

metallic_region_0 = BoxRegion(
      0*Volt,
      xmin = 8*Angstrom, xmax = 23*Angstrom,
      ymin = 4*Angstrom, ymax = 4.5*Angstrom,
      zmin = 0*Angstrom, zmax = 26.2*Angstrom
   )

metallic_region_1 = BoxRegion(
      0*Volt,
      xmin = 8*Angstrom, xmax = 23*Angstrom,
      ymin = 21.5*Angstrom, ymax = 22*Angstrom,
      zmin = 0*Angstrom, zmax = 26.2*Angstrom
   )

After giving the gate voltage the perpendicular electric field is defined as the ratio of:
 
             (Gate voltage)
     -- -----------------------------
(distance between two metallic plates)

Looking at the dimensions of metallic region in code what value should i choose for denominator??

is it (ymax_of_1 - ymin_of_0)=(22-4)=18A
or
is it (ymin_of_1 - ymax_of_0)=(21.5-4.5)=17A

[huckelbuckel]

Hi,

Can someone please answer me about the distance between metallic plates so that i can know the value of electric fields.

[Anders Blom]

The potential is constant within the metal, so it's the dimensions towards the structure which matter. From this you can compute the field - provided you have vacuum between the electrodes. You don't, I guess, you have some structure, and this will partly screen the field. So the field will be position-dependent and can be computed via the gradient of the electrostatic potential.

[huckelbuckel]

thanks for clarification.