Interface Relaxation

[AsifShah]

Dear Admin,
Are these methods same or different? (Kindly mention the difference if any) {Assume A & B are pre-relaxed before building interface}
Method 1: I make interface between two materials A & B using interface builder with a strain of 3%. Then I do interface relaxation using DFT keeping layers away from interface as rigid.

Method 2: I make interface between two materials A & B using interface builder with a strain of 0.8% (using large supercell). Then I do interface relaxation using DFT keeping layers away from interface as rigid.

Method 3: I just drag & drop two materials with (say) 4x4  supercell in builder & adjust interface distance to some value from experiments. Then perform a full DFT optimization.

Will all methods yield same end configuration? Also, how is method 3 different from 1 & 2 and if it is same what is then the purpose of interface builder?

[Anders Blom]

All methods are different and you are most likely obtaining different intrerfaces in each case.

The important functionality the interface builder offers is that by varying the orientation between A and B and their respective cell vectors, you can minimize the strain. So it's not just a matter of a larger cell (although generally smaller strain can only be obtained with larger cells), it's also about how the two interfaces are oriented relative each other.

[AsifShah]

Thanks that was helpuful.