Calculation of the free energy barriers associated with physical or chemical change of systems of large number of Degrees Of Freedom (such as a chemical reaction or structural phase transitions) is one of the major challenges in computational science. Atomistic computer simulation techniques such as molecular dynamics (MD) and Monte Carlo (MC) do allow for free energy estimates but for cases where the barrier heights are much smaller than kT.    
However when the barrier heights are comparable to kT these standard 
technique are of limited practical use. The ineffectiveness of MD and MC 
in such cases can be traced to the inefficiency in the sampling of the 
multi-dimensional potential energy surface (PES) of the system. 
The recently proposed metadynamics algorithm (by Liao and Parrinello in 2002) improves the sampling rates through incorporation of artificial time dependent potentials added to the real/natural potential energy surface of the system. The technique, for the simple case a particle moving in a double well potential, is demontrationstrated in the figures.  The major advantages are (1) elucidation of the mechanism of the process (in the cases of systems with many DOF), and (2) reliable estimates of the free energy barriers involed. 

or details see,
    • J. Chem. Phys., 126, 204315 (2007), "Dissociation of Carbonic Acid: Gas Phase Energetics and Mechanism from ab initio Metadynamics Simulation", Padma Kumar P., Andrey G. Kalinichev, and R. James Kirkpatrick.