We recently developed the electron force field (eFF) method for practical nonadiabatic electron dynamics simulations of materials under extreme conditions and showed that it gave an excellent description of the shock thermodynamics of hydrogen from molecules to atoms to plasma, as well as the electron dynamics of the Auger decay in diamondoids following core electron ionization. Here we apply eFF to the shock thermodynamics of lithium metal, where we find two distinct consecutive phase changes that manifest themselves as a kink in the shock Hugoniot, previously observed experimentally, but not explained. Analyzing the atomic distribution functions, we establish that the first phase transition corresponds to (i) an fcc-to-cI16 phase transition that was observed previously in diamond anvil cell experiments at low temperature and (ii) a second phase transition that corresponds to the formation of a new amorphous phase (amor) of lithium that is distinct from normal molten lithium. The amorphous phase has enhanced valence electron-nucleus interactions due to localization of electrons into interstitial locations, along with a random connectivity distribution function. This indicates that eFF can characterize and compute the relative stability of states of matter under extreme conditions (e.g., warm dense matter).wavepacket dynamics | interstitial electron model | symmetry breaking T here are great uncertainties in the properties of matter at the high compression (several times ambient densities) and high temperatures (20,000-2,000,000 K) characteristic of deep interiors of giant planets (1), conditions of thermonuclear fusion, and phenomena generated by shocks from planetary impact (2). New methods for experimental study of these regimes are being developed (National Ignition Facility at Lawrence Livermore National Laboratory, Z-Pinch at Sandia National Laboratory) that generate data about materials under extreme conditions. However, theoretical and computational methods used to predict properties of warm dense matter [high temperatures (T), high pressures (P), and under rapidly changing conditions] have serious shortcomings leading to considerable uncertainties due to high degrees of electronic excitation, structural and electronic heterogeneity, and complex transient dynamics. This contrasts with the situation at room temperature (RT), where a wealth of structural and energetic data on compressed phases is available via experiments in diamond anvil cells and gas guns (3-5), and from theoretical studies such as density functional theory (DFT).To provide fresh insight into the dynamical properties of warm dense matter, we developed the eFF (electron force field) methodology that explicitly solves the time-dependent Schrödinger equation including all two-body interactions, with the restrictions that the electrons are described as Gaussian wave packets and that the wavefunction is described as a Hartree product (with exchange terms replaced by a spin dependent Hamiltonian). The eFF makes it practical to describe the non...