QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: markmick on March 15, 2016, 21:25

Title: BZ Routes: K and K-prime
Post by: markmick on March 15, 2016, 21:25

When using the VASP Scripter, how can you specify K-prime (as opposed to K).  I need a K-Gamma-K' because my k-space is strained, and won't be symmetric; so I get K'.  Do I do something like the following in the KPOINTS file?

0.333333 0.333333 0.000000 ! K
0.000000 0.000000 0.000000 ! G

0.000000 0.000000 0.000000 ! G
-0.333333 -0.333333 0.000000 ! K

Thanks,
Mark

Title: Re: BZ Routes: K and K-prime
Post by: Daniele Stradi on March 16, 2016, 10:24

Yes, that should work perfectly.

Title: Re: BZ Routes: K and K-prime
Post by: markmick on March 16, 2016, 21:52

Thanks.  That's what I'm doing.  When you're stretching a molecule, how much actual change can be expected from a symmetric hexagonal lattice to an asymmetric one? When I run vasp calculations, the difference is size between the two BZ routes is so small, it's almost imperceptible.   I have to subtract the differences between eigenval data before I see any change in size.  Even then it's quite small.  Is that to be expected?

Here's the way I run my bandstructure calculations.  I allow the poscar to relax for all 3 runs.  Should I fix them after optimization?  Or is this OK and my data is as expected?  I don't have any data to compare to as a reference, so I don't know if my data is right or wrong.  There is a clear trend of percent strain vs the ratio change of the two K-K' routes at least.

1) Optimization.  Unfixed poscar
2) Self-consistency.  Unfixed poscar
3) Bandstructure.  Unfixed poscar.

Mark

Title: Re: BZ Routes: K and K-prime
Post by: Daniele Stradi on March 17, 2016, 09:40

Your procedure is correct, but of course depends on the actual system, and on the program/method used.