Dear Prof. Axel van de Walle,
I am very happy to have access to your code ATAT, which works very well on our machines.
However, I don’t understand the output of the subroutines emc2 and phb whose outputs
consists of 21 ad 6 columns respectively. The manual doesn’t seem to give the description about
these outputs . Will you be kind enough to send us a description about the same.
Type
emc2 -h
and
phb -h
It’s at the end of the help.
For memc2, the file mcheader.out gives the meaning of the columns (it is dependent on the number of components, etc.)
Prof. Axel van de Walle,
I’m trying to use ATAT for C_x(BN)_{1-x}, a pseudo binary alloy. The maps code works fine with the input but when I try to run phb, it fails with the error, "Sorry, only binary alloys implemented". Is it possible to do such calculations?
My unit cell is hexagonal, with 2 atoms in it. The atom type used is C,B,N and C.
0.00000000 0.00000000 0.00000000 C,B,N
0.66666667 0.33333333 0.00000000 C.
Instead, I tried,
0.00000000 0.00000000 0.00000000 C,B
0.66666667 0.33333333 0.00000000 C,N.
In this case, phb works but only for concentrations at -1 and 0. The output of it has just two lines ,unlike the example files, at T=0 and T=dT.
emc2 fails as well with the following error, "Ground states must be in increasing order of concentration in gs_str.out."
We would really appreciate it if you could help us sort our difficulties.
Both of your examples are really multicomponent (because there is more than one independent composition variable), so you should use mmaps and memc2 (instead of maps and emc2).
(If maps worked, it may have given you a valid cluster expansion but the ground state search would not have been correct).
Unfortunately, there is no corresponding mphb (because it’s really hard to implement!) for multicomponent phb. But memc2 should work. And you should re-run mmaps on your existing data, just to be sure.