Handling arbitrary crystal tilts with abTem's orthogonalize function

[tiarnan-doherty]

Hi,

I'd like to simulate some microED patterns of crystals oriented relative to the incoming beam at arbitrary tilt angles in abTEM. I've been trying to come up with the best workflow for generating orthogonal cells compatible with a multislice simulation, which is not particularly straightforward.

Am I understanding correctly that if I use ASE to rotate an atoms object with rotate_cell = True, then pass that rotated atoms object to orthogonalize_cell, that I should get an orthogonal cell (though possibly one to which strain has been applied) that preserves the periodicity of the rotated unit cell?

Looking at the atoms.py, this seems to be correct but am I either misunderstanding the functionality of orthogonalize_cell or missing some scenario where this approach might fail? I know previously there has been some discussion surrounding the limitations of orthogonalize_cell (#25) but looking at the source code now it seems like there have been some substantial additions to the orthogonalize_cell functions since 2021.

If orthogonalize_cell can't handle the described situation, I would be very happy to help contribute to developing it further if there was interest.

[TomaSusi]

Hi Tiarnan!

This is a highly non-trivial problem, but one that we have been working quite intensively on in the context of a collaboration. Orthogonalize_cell isn't really the main issue here, but doing the rotation in a way that the cell is kept orthogonal and repeated so that the rotated atoms can fit inside commensurately.

We do have a solution that is fairly general but we're not quite ready to share it yet.

[tiarnan-doherty]

Hi Toma,
Thanks so much for your reply. I’ll keep an eye on abTEM for future updates on the code that you’re developing! In the meantime I’ll see if I can figure out some suitable representations of the tilted crystals I’m interested in.