Comment on ``Phase diagram of MgO from density-functional theory and molecular-dynamics simulations''
The paper by Strachan et al. [Phys. Rev. B 60, 15 084 (1999)] presents a theoretical phase diagram for MgO. I argue that the B1-liquid phase boundary and the triple point B1-B2-liquid, if calculated correctly using the same model for MgO, might be significantly different from the presented ones. I argue further that the chosen method for calculating the B1-liquid boundary does not provide melting temperatures with a reasonable accuracy, is more time consuming than the two-phase simulation method, and less precise. The only molecular-dynamics simulation in which the MgO melting curve was determined is that by Belonoshko and Dubrovinsky.