Pressure-dependent melting temperature of xeon

In the past few years, we developed several new transition state searching method (CBM,CBD, BP-CBD, DESW), and based on these methods, we designed a new global optimization method for potential energy surface (PES) search using the approximate normal mode (second derivatives), namely, stochastic surface walking (SSW) method (J. Chem. Theory Comput, 2013, 9, 1838). This method goes beyond the first derivatives, as most current algorithms do for PES search and at the same time does not compute accurate second derivatives, which are generally regarded to be essential for saddle point location. By adding bias potentials and performing local relaxation repeatedly, the SSW method can perturb smoothly the structure from one minimum to another following a random direction. The SSW method in combination with parallel replica exchange can be utilized for searching structure and predicting chemical reactions with a high efficiency. Using this method, we were able to identify the global minimum of short ranged Morse clusters, the carbon fullerene up to 100 atoms and Boron 40 cages based on classical potentials and first principles calculations. In this talk, I will give a few new examples of SSW method for structure searching, automated reaction prediction from solid phase transition to molecular reactions.