materialsvirtuallab / data-driven-first-principles-methods-for-the-study-and-design-of-alkali-superionic-conductors Goto Github PK

Email (Optional)

[email protected]

Version

1 - Structure Generation

Which OS(es) are you using?

• MacOS
• Windows
• Linux

What happened?

Dear Authors,

Firstly, I would like to extend my appreciation for your significant contributions to the field through your insightful work and the development of the "1 - Structure Generation" notebook. Your efforts in making these resources available are highly commendable and greatly beneficial to the community.

I recently utilized your notebook to generate structures and successfully obtained 48 unique configurations. However, upon closer inspection, I noticed an inconsistency that I hope you can help clarify.

The structures I generated contain only 13 atoms and exhibit non-cubic geometries. This observation seems to contrast with various resources such as the Isosurface_800K_0.png, CHGCAR.vasp, vasprun.xml.relax2 files, and Figure 1 in your publication, which all suggest the generation of structures that are approximately cubic and contain 52 atoms.

I am keen to understand the reason behind this discrepancy and would be grateful for any insights or guidance you could provide to resolve this issue.

Thank you once again for your valuable work and for taking the time to address my query.

Best regards,

Wenshuo Liang

Code snippet

from pymatgen.core.structure import Structure
from pymatgen.symmetry.analyzer import SpacegroupAnalyzer
from pymatgen.transformations.advanced_transformations import EnumerateStructureTransformation
from pymatgen.io.vasp.sets import batch_write_input, MPRelaxSet

structure = Structure.from_file("EntryWithCollCode418490.cif")
#print(structure)

for i, site in enumerate(structure1):
    if not site.is_ordered:
        structure[i] = {"Li+": 0.5}
print(f"The composition after adjustments is {structure.composition.reduced_formula}")
#print(structure)

analyzer = SpacegroupAnalyzer(structure)
prim_cell = analyzer.find_primitive()
print(prim_cell)

enum = EnumerateStructureTransformation()
enumerated = enum.apply_transformation(prim_cell, 100)  # return no more than 100 structures
structures = [d["structure"] for d in enumerated]  
print(f"{len(structures)} structures returned")

batch_write_input(structures, vasp_input_set=MPRelaxSet, output_dir="Li6PS5Cl_orderings")

Log output

Below is an example of one of the 48 structures I generated:

Li6 P1 S5 Cl1
1.0
  -4.9295000000000000    0.0000000000000000   -4.9295000000000000
   4.9295000000000000   -4.9295000000000000    0.0000000000000000
  -4.9295000000000000   -4.9295000000000000    0.0000000000000000
Li P S Cl
6 1 5 1
direct
   0.3704000400000000    0.0180000000000000    0.0180000000000000 Li+
   0.6295999600000000    0.6115999600000001    0.0180000000000000 Li+
   0.9640000000000000    0.3524000400000000    0.0180000000000000 Li+
   0.9640000000000000    0.6115999600000001    0.0180000000000000 Li+
   0.3704000400000000    0.3524000400000000    0.6115999600000001 Li+
   0.6295999600000000    0.0180000000000000    0.3524000400000000 Li+
   0.9999999999999999    0.5000000000000001    0.5000000000000002 P5+
   0.4999999999999999    0.2500000000000001    0.2500000000000002 S2-
   0.7610600000000003    0.1415900000000008    0.6194700000000001 S2-
   0.7610600000000003    0.6194699999999996    0.6194700000000001 S2-
   0.2389399999999995    0.6194699999999999    0.6194699999999996 S2-
   0.2389399999999995    0.6194699999999997    0.1415900000000008 S2-
   0.0000000000000000    0.0000000000000002    0.9999999999999998 Cl-

Code of Conduct

• I agree to follow this project's Code of Conduct