- #1
Useful nucleus
- 370
- 58
It is very common to see in the literature negative formation energies reported for point defects in solids using different simulation techniques ranging from Density Function Theory to Molecular Dynamics.
The authors rarely comment on what does that mean physically. I will mention here two possible explanations that I can think of and I'd appreciate it if you can share with us your understanding of this issue:
(1) Because of the small size of the systems that can be handled by simulation and because of the common practice (or in sometimes the need) to use periodic boundary conditions, defects interact with its images leading to unphysical lowering of their formation energies.
(2) The result is simply unphysical because the computed reference chemical potential of the species that caused the defect (for example sulfur gas reference for a sulfur defect in a sulfide ) is wrong.
The issue is very important especially when it comes to compute the equilibrium concentration of these defects using the exponential expression: n=n0exo(-E/kT)
The authors rarely comment on what does that mean physically. I will mention here two possible explanations that I can think of and I'd appreciate it if you can share with us your understanding of this issue:
(1) Because of the small size of the systems that can be handled by simulation and because of the common practice (or in sometimes the need) to use periodic boundary conditions, defects interact with its images leading to unphysical lowering of their formation energies.
(2) The result is simply unphysical because the computed reference chemical potential of the species that caused the defect (for example sulfur gas reference for a sulfur defect in a sulfide ) is wrong.
The issue is very important especially when it comes to compute the equilibrium concentration of these defects using the exponential expression: n=n0exo(-E/kT)