How to calculate the phase diagram by phb

Alloy Theoretic Automated Toolkit (ATAT)
1

Dear Axel:
I try to simulate the phase diagram of Si1-xGex alloy. The CE construction has implemented using maps code. The fitted ECIs were then used to predict the formation enthalpies of all symmetry inequivalent alloy configurations (~10000) up to 14 atoms per cell. CV scores is 0.00365942.
I try to compute the temperature-composition phase boundary:
phb -dT=10 -dmu=0.001 -ltep=1e-3 -er=9 -gs1=0 -gs2=1 -o=phb.out -k=8.617e-5 -dx=1e-4
I got output file phb.out. It stops at T=360K. I try to increase the value of er. It still could not get the whole phase diagram. How can I get complete phase diagram (concentration from 0 to 1)?
I am puzzled that the values of 3 column in phb.out are negative. I think it is reasonable that the concentration is from 0 to 1.
The output file of maps and phb in attachment.

Thank you.
Tieyu Lu
2

Don’t worry about the negative values: these are average spins ranging in [-1,1], which correspond to concentration in [0,1].
The code probably stops because the composition jumped by too much, which is interpreted (incorrectly) as a phase transition. try reducing dT or reducing -dx
BTW, you can restart phb just a few steps before is stopped (specifying the right -mu from column 2).

3

Thank you for your reply!