Forum Replies Created
-
Posts
-
This seems to be due to ever increasing browser safety settings. I’ve now added a hint before the downloads how to circumvent this.
This will be dealt with in the feedback webinar (hint: losely bound = extending further into space = more distance needed to get uncoupled).
Correct answer, Jason. This topic is typically dealt with in introductory quantum physics courses, but it is extremely common that the value of this concept is not appreciated at that stage and gets forgotten ;-).
This will be summarized once more at the end of the feedback webinar.
relax ==> only internal positions are optimized, the cell shape and volume is frozen
vc-relax ==> internal positions as well as cell shape and size are optimized.The CELL block is for vc-relax only. It specifies the target pressure (often that will be zero, but you could want to find the cell shape and size at other pressures as well). In cases where the pressure at the end of the scf cycle does not agree with what you asked for, you’ll have to do another round (which may include setting the press variable to a new value, if you aim for the pressure that corresponds to a user-defined volume range).
Did this error message disappear — just as the previous ones did — when you reinstalled your environment last week? Or is it still there?
It may perhaps not have been said explicitly, but indeed, VESTA should better be run on your normal OS. It is not a part of Quantum Mobile, although of course it can be installed within Quantum Mobile.
Run Vesta locally, and transfer cif files between Vesta and your virtual machine via the shared folder. That’s the efficient way to do it.
When I open that cif file, I see lines like this:
Si1 0.14650 0.40820 0.05990
Si1a -0.14390 -0.40580 -0.05990This means: non-numerical labels, albeit with a number+character, not just a character. Can you see what happens if you replace ‘1a’ by ‘5’, ectc?
There is in this cif file a large block with anisotropy information first, and only then the coordinates. To be on the safe side, make the same (and consistent) changes in both blocks.
We’ll come back to this in the feedback webinar, but let me tell it already here:
It’s about the _atom_site_occupancy variable. That variable tells which fraction of a position coordinate is occupied by the given element. For DFT input, it must be 1: either there is an atom at that coordinate, or there is no atom at all. In experimental cifs, however, you can have partial or mixed occupancy: an element occupying only, say, 37% of all availabe sites. Or two elements appearing at the same site, in a 29/71% ratio.
For the neptunite cif by Cannillo et al, you see this happening:
Fe2+1 0.34020 0.32110 0.09830 0.50000 0.01925
Ti1 0.34020 0.32110 0.09830 0.50000 0.01925
Fe2+2 0.08830 0.05610 0.11280 0.50000 0.01925
Ti2 0.08830 0.05610 0.11280 0.50000 0.01925Fe and Ti both share the same positions, with a 50/50 chance for each. That’s fine for an experimental cell, but DFT cannot model this (we’ll see a way around this later, using supercells).
