Generation of B2 disordered structure for a binary alloy

Alloy Theoretic Automated Toolkit (ATAT)
1

Hi,

I prepared randstr.in file like this:

3.576685 0.000000 0.000000
0.000000 3.576685 0.000000
0.000000 0.000000 3.576685
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000
0.000000 0.000000 0.000000 Mg=0.680000,Li=0.320000
0.500000 0.500000 0.500000 Li

corrdump -l=randstr.in -ro -noe -nop -2=4 -3=4 -clus
mcsqs -n=100 -l=randstr.in -cf=clusters.out -ip=8
mcsqs -n=100 -l=randstr.in -cf=clusters.out -ip=8 -rc

Then bestsqs8.out was generated and Mg=34 and Li = 66.
In general, final Li = (N+ initial Li)/2; final Mg = N- final Li;

I appreciate if you let me know what is the logic behind this?
and how can I set the randstr.in so that I get the desired atomic % of Mg and Li in the disordered B2 binary alloy?

Thanks

2

You have a 2nd lattice filled with Li, so
#Li = 50 + 0.3250 = 66
#Mg= 0 + 0.6850= 34

3

Thanks very much Axel.
So if one needs to prepare a randstr.in file to be used in generating SQS of B2 structure
where the second lattice is filled with Mg and Li, then one should write:

makelat Mg,Li:Li,Mg B2

then randstr.in
3.637307 0.000000 0.000000
0.000000 3.637307 0.000000
0.000000 0.000000 3.637307
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000
0.000000 0.000000 0.000000 Mg=0.9,Li=0.1
0.500000 0.500000 0.500000 Li=0.1,Mg=0.9
and coredump and mcsqs commands.

As such, the resulting structure is still considered

  • B2
  • SQS (disordered)?

Thanks
Reza

4

B2 with both sublattices disordered at the same composition is just bcc!

5

Yes, you are quite right.
But if x = Li atomic percent
and we need to find the energies of Mg(1-x)Lix B2 alloys for

x=0.05, 0.10,…0.90, 0.95, 1.0

what would be the y, z and w in the 20 files of randstr.in to generate desired bestsqs.out (as a precursor to POSCAR files).

3.637307 0.000000 0.000000
0.000000 3.637307 0.000000
0.000000 0.000000 3.637307
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000
0.000000 0.000000 0.000000 Mg=(1-y),Li=y
0.500000 0.500000 0.500000 Li=z,Mg=w

I have seen several papers calculating phase diagrams of random wurzite structures of Ti(1-x)AlxN and the like [e.g. J. APPL. PHYS. 100 (2006) or J. APPL. PHYS. 113 (2013)].

Many thanks
Reza

6

If the formula is Ti_(1-x) Al_x N then one site is filled with Ti and/or Al (with fractional occupations) and N is on the other site, fully occupied. That I can see how to make a B2 structure out of.
In your case, I have no idea what is a Mg(1-x)Lix "B2" alloy.