Molecular dynamics of LiF melting

We performed molecular-dynamics simulations of the melting and/or freezing of LiF. The simulations were done using the Tosi-Fumi model and our own model of interatomic interactions. The latter was verified by ab initio calculations of the equation of state for LiF. We show that the recent molecular-dynamics calculations by Boehler and co-workers are not adequate and their model for the interactions is not capable of providing melting temperatures in agreement with experiment. Our calculated pressure dependence of the melting temperatures gives valuable information. We found that the B1-B2 transition in LiF at around 1 Mbar removes the discrepancy between the diamond-anvil cell and shockwave melting temperatures. An explanation of the controversy between ``low'' and ``high'' melting temperatures obtained from diamond-anvil cell experiments is suggested.