Author Topic: Is the system metallic?  (Read 6603 times)

0 Members and 1 Guest are viewing this topic.

Offline swap700

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: us
  • Reputation: 0
    • View Profile
Is the system metallic?
« on: September 17, 2011, 21:37 »
Hi,

I am attaching image of band stucture of a system I worked with for graphene, I believe the system is not metallic but I when I export it to other software for running some different calculations, I get an error saying the system is metallic, can anyone help me understand the band structure better. Will provide better zoomed images if needed.
Thank you

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
Re: Is the system metallic?
« Reply #1 on: September 17, 2011, 22:10 »
Without seeing your input structure it's a bit hard to say for sure. You cannot really say for certain if a system is metallic or not by just looking at the band structure between G and Z, unless of course the system is 1D (periodic in the Z direction and finite in X and Y).

The band structure looks a bit unusual, with a flat level so close to the Fermi level, it looks like a doping level. Numerically it might happen that the level wiggles a bit in the X/Y due to electrostatic interactions even if it's supposed to be finite in these directions, for instance if you have too little vacuum. A DOS plot (with relevant k-point sampling in all directions) would be able to determine the existence of a real band gap more convincingly. But as mentioned it's all speculative, not seeing the geometry.

Offline swap700

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: us
  • Reputation: 0
    • View Profile
Re: Is the system metallic?
« Reply #2 on: September 17, 2011, 22:15 »
Hey,

Thanks Anders, here is the input geometry.

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
Re: Is the system metallic?
« Reply #3 on: September 17, 2011, 22:28 »
So the system is effectively 2D, not 1D, hence the band structure also has a non-trivial dispersion in the Y direction. You should re-compute the band structure to also include this direction, say by doing G,Y,Z,G, although more properly one should try to use symmetry points of the underlying hexagonal lattice (G,K,M, in 2D), which can be a bit tricky because they are not defined for the square supercell you have created.

If you're only looking for a confirmation that this is a metallic system you can do that, there should be bands on G-Y that cross the Fermi level. On the other hand this fact almost doesn't need proving, since you basically just have pure graphene, which we know has a zero band gap at the K point, with a point defect, which is not enough by itself to open any real band gap.

If you need an accurate band structure, then I suggest creating the system a bit differently, namely by making a supercell by repeating the fundamental graphene hexagonal cell a few times, before introducing the defect. The conversion to supercell means the cell goes to UnitCell, and you lose the symmetry points, but in the Builder you can convert it back to Hexagonal, and then you can easily set up the band structure along the usual routes, like G,K,M,G.
« Last Edit: September 17, 2011, 22:29 by Anders Blom »

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
Re: Is the system metallic?
« Reply #4 on: September 17, 2011, 22:31 »
Btw, if you intended for this to be a semiconducting armchair edge nanoribbon, you need to add vacuum and hydrogen atoms in the Y-direction.

Offline swap700

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: us
  • Reputation: 0
    • View Profile
Re: Is the system metallic?
« Reply #5 on: September 17, 2011, 22:32 »
Hi,

Thanks for the prompt reply, I just want the broad sense if the system is metallic or not, so as you suggested I will modify the route G,Y,Z,G and visualize the bandstructure.

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Is the system metallic?
« Reply #6 on: September 18, 2011, 15:54 »
I guess that you are doing the band structure for the graphene with a single vacancy passivated by three hydrogen atoms. Please check the paper by Lehtinen et al, Phys. Rev. Lett.,93,187202(2004).

In your calculations, a supercell with  orthogonal shape was employed. However, in the given band structure, only the dispersion for the direction from G to Z was plotted. You had better check also the band dispersions for the other two important directions: G-->S, and X-S.
Here S point: (-0.5, 0.5, 0) given with respect to the reciprocal lattice vectors for an orthorhombic primitive Bravais lattice.

Maybe the high symmetry k points for  the Brillouine zones of all Bravais lattice are not predefined completely in the current version. The complete list for the high symmetry k points in the Brillouine zones of all Bravais lattice can be referred to:
Bradley C.J., Cracknell A.P. Mathematical Theory of Symmetry in Solids, Oxford, 1972,  pages 88-108.
 


« Last Edit: September 22, 2011, 03:39 by zh »

Offline swap700

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: us
  • Reputation: 0
    • View Profile
Re: Is the system metallic?
« Reply #7 on: September 19, 2011, 07:39 »
How do I define point M ? The only possible points I can define are G,X,Y,Z

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
Re: Is the system metallic?
« Reply #8 on: September 19, 2011, 09:51 »
See my post above which discusses this.

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Is the system metallic?
« Reply #9 on: September 22, 2011, 03:57 »
It is better to employ a hexagonal supercell with the size of (3n x 3n x1), here n is a positive integer. This will ensure that the Dirac point (K point) can be folded exactly into the Gamma point in the Brillouine zone of the supercell.