Fold2Bloch transformation matrix

Running Abinit Other tools and postprocessing in ABINIT
1

I am currently dealing with systems in which I have to consider supercells and therefore using the fold2Bloch tool will be very useful for my calculations. In the abinit tutorial it is stated that:

"The program fold2Bloch is executed by invoking the following command from the directory of the case file :

fold2Bloch case_WFK x:y:z

Where:

case is the name of the file that comes before _WFK identification
x:y:z are integer values( >0) that represent a multiplicity in the corresponding direction used when constructing the super-cell; separated by a colon (:)."

However, this input is only useful for supercells where the multiplicity is an integer, and not for sqrt(2) or sqrt(3) cells, which I am interested in. I was wondering if there is a way to unfold the bands associated with these types of supercells (e.g. by introducing the full transformation matrix). In the fold2Bloch implementation for VASP, it is possible to introduce the transformation matrix according to the repository (GitHub - rubel75/fold2Bloch-VASP: Unfolding the band structure of a supercell obtained with VASP), but I haven’t found any information about it regarding the abinit implementation.

Thanks

2

Did you simply try their general format with a word string of the full supercell matrix? Start with a trivial test you know should work with a diagonal matrix, low ecut and kpt, and compare to the standard input format

The question is really not for abinit but rather whether f2bloch can combine a random supercell with WFK file reading.

M.

3

Hello,

Yes, I tried, but I always got the following error:

— !ERROR
src_file: m_fold2block.F90
src_line: 336
mpi_rank: 0
message: |
Number of folds has to be a positive integer greater than 0

I have tried different input formats but none has worked so far.