Though i wrote the above as a general query, but i need to understand 'how is the E un-relax is calculated? in these references?' (Ref. DOI 10.1007/s10853-017-0803-5) and (Ref. X. Tian et al. / Applied Surface Science 427 (2018) 357–362)

I have many doubts of these are actually computed.

1. The Un-relaxed model could be 'any model' we start with before it is actually subjected to 'relaxation' calculation. Then if we take 'static run' of that 'any model' is the one referred to as 'E un-relax'?

1(a) If yes, then this 'any model' could be a manually arranged (or adjusted) atomic model which may be highly distorted from real model. Will this be OK, for calculations? and how much accurate is such method of computation starting from very distorted model?

1(b) if no, then 'how exactly this term to be calculated'?

2. Can anyone suggest any reference (or explain) which has very clear explanation of how such calculation is to be done? [Though the references i gave has explained the steps, but sorry, i could not decipher it fully]

Regards