Error of SQS (mcsqs) with low %atom concentration

Alloy Theoretic Automated Toolkit (ATAT)
1

Hello everyone,
I got the error when I tried to generate SQS structure of Co-doped Ag2S system. The unitcell of Ag2S includes 12 atoms (8 Ag and 4 S) with a monoclinic phase. Now I want to replace 2 Ag atoms by 2 Co, randomly. -> Ag6Co2S4.
In rndstr.in

1 1 1 90 124.5369 90
4.1837906126625510 0.0000000000000000 0.0195524170958123
0.0000000000000000 6.9622533301403058 0.0000000000000000
-5.4342891319584234 0.0000000000000000 7.8168507265841969
0.9994522179189715 0.9998432464813263 0.2516562928886401 Ag=.75,Cr=.25
0.0005476280810299 0.0001567635186746 0.7483436861113617 Ag=.75,Cr=.25
0.0005477410810286 0.4998432364813254 0.2483436811113648 Ag=.75,Cr=.25
0.9994522289189689 0.5001567635186746 0.7516562208886342 Ag=.75,Cr=.25
0.7403914961354587 0.2503551736290319 0.4353043272305825 Ag=.75,Cr=.25
0.2596084388645394 0.7496447923709653 0.5646955897694141 Ag=.75,Cr=.25
0.2596084368645464 0.7503551736290319 0.0646956157694163 Ag=.75,Cr=.25
0.7403914751354606 0.2496448263709681 0.9353043162305852 Ag=.75,Cr=.25
0.5151289250517337 0.2497560593848505 0.1291786017763954 S
0.4848709409482694 0.7502439236151446 0.8708213922236041 S
0.4848709939482703 0.7497561103848511 0.3708213412236034 S
0.5151289400517314 0.2502439406151495 0.6291785757764004 S

Next, I used the command:
corrdump -l=rndstr.in -ro -noe -nop -clus -2=1.1
Because I want to keep the shape of the cell so I created sqscell.out like this:
sqscell.out

1

4.1837906126625510 0.0000000000000000 0.0195524170958123
0.0000000000000000 6.9622533301403058 0.0000000000000000
-5.4342891319584234 0.0000000000000000 7.8168507265841969

After that, I used the command below to keep the shape and generate a random structure:
mcsqs -rc
Unfortunately, I got this error:
"Impossible to match point correlations due to incompatible supercell size."
When I tried to change the percent of all Ag/Co atoms to Ag=0.5,Co=0.5 => It work (Now I have 4 Ag-4Co-4S in 1 cell!! and it’s not the structure as I expected.
So I don’t know exactly how can I choose the different doping concentrations like Ag=0.75,Co=0.25?
Do I make some mistakes in any part? I’m really stuck with this.
Please help me with your experience, Thank you very much!!!

2

You need to increase the size of your cell. I’ve explained this elsewhere in this forum, but a common problem is that, when a structure has symmetrically inequivalent sites, the notion of SQS requires that each sublattice is random. Look at the file rndstrgrp.out to see if you have multiple groups of equivalent sites (sublattices).

BTW, your input file is probably not correct. Your atomic coordinate are fractional but your lines 2-4 do are not a simple matrix of fractions. If you want to use fractional, line 1 should give the cell and line 2-4 should be the identify matrix or perhaps with some 0.5 entries for a centered Bravais lattice. Read mmaps -h