[QE-users] Boltztrap Calculation

[Davide Ceresoli]

Dear Elham,
... in short: use BoltzTrap2!

I have fixed qe2boltz.py but I don't recommend using Boltztrap 1...

I have found the following glitches:
- for spin polarized systems, it will compute the transport properties
   only of the spin up channel
- if you don't change 'TETRA' to 'HISTO' in the file generated by
   qe2boltz.py, the Fermi level is miscalculated, with the result
   that even insulators will have a finite conductivity
- when the Fermi level is inside the gap, the Seebeck coefficient
   is meaningless (don't plot it!)

I'm writing this, hoping to help whoever is doing thermoelectric
calculations. Do not blame "Reviewer #2" then... 😂

Best,
D.




On 5/3/24 14:50, Elham Rezaee wrote:
> Dear Quantum Espresso Users,
> I hope this email finds you well. I am reaching out to seek assistance regarding 
> a Boltztrap calculation I am attempting to perform on the ComputeCanada cluster. 
> I have successfully completed relax and nscf calculations in Quantum Espresso 
> (QE), and following the tutorial provided at 
> https://blog.levilentz.com/boltztrap-tutorial-for-quantum-espresso/ 
> <https://blog.levilentz.com/boltztrap-tutorial-for-quantum-espresso/>, I was 
> able to generate the files prefix.structure and prefix.energy. However, I 
> encountered difficulties obtaining the results for prefix.def and prefix.intract.
> I should mention that while I could obtain all these 4 files for the silicon 
> (Si) material, I am facing challenges with my compound, which lacks symmetry. 
> Despite employing the following command in Python to handle the 'No symmetry 
> found' error: /elif 'No symmetry found' in line:/
> /    nsym = 1/
> /    try:/
> /        print(nsym)/
> /    except:/
> /        nsym = 1/
> 
> I am unable to generate prefix.def and prefix.intract files.
> Does anyone have suggestions on what steps I should take next to address this 
> issue? Alternatively, does anyone possess a version of the Python file that 
> effectively handles this part of the process? The link provided in the tutorial 
> for qe2boltz.py seems to be inaccessible.
> Thank you,
> Best regards,
> Elham Rezaee, PhD
> University of New Brunswick, Canada