Specific composition with mmaps

Alloy Theoretic Automated Toolkit (ATAT)
1

Hi all,

I’m trying to find the GS configuration for a specific alloy composition. MMAPS should be able to do that if I set narrow limits using crange.in, correct?

However, if I’m being very specific, mmaps doesn’t seem to go up to more atoms per cell, but instead gets stuck on an infinite loop on the "Finding best structure…" step.

Example: start with atat/examples/cuau.in. Let’s say we’re just interested in how CuAu looks (should find the Au "layer"), so I define crange.in as:

1.0*Cu >= 0.50
1.0*Au >= 0.50

mmaps calculates pure Cu, pure Au and then hangs when it should generate the first 2 atoms cell… this works when I remove one of the constraints, but that means a lot of compositions are calculated that I’m not interested in. Note: the same happens when I use ">= 0.49" or something, so it’s not just the inequalities becoming impossible because of rounding errors.

Did I misunderstand something here?

Regards,
Sebastian

2

First, you should probably relax the constraints (for numerical reasons), e.g.:
1.0Cu >= 0.48
1.0Au >= 0.48

It could take longer to enumerate because the code still enumerates all compositions and just outputs those that satisfy the constraint. Make sure you download the latest beta version (the enumeration routine is a lot faster).

Now, I should warn you that enumerating at a fixed composition is risky: very often the ground state (GS) at one composition is a phase-separated mixture of other phases that bracket that composition. This is why maps may very well output other compositions outside of your constraints. This is necessary to ensure that the GS at that composition is a true GS, not something masked by a common tangent between phases at other compositions.