There are two ways to proceed with nanoscience: so-called top-down and bottom-up methods. Usually, the former methods are thought of as in the province of physicists and the latter in that of chemists. However, this is not entirely true because the physics of disordered matter, especially liquid metals, is well-developed bottom-up science and it has indeed provided nanoscience with basic ideas and theoretical tools such as ab initio molecular dynamics (MD) simulations. Here we wish to present experimental studies on such phenomena that originate from quantum mechanical properties and subsequently lead to classical non-equilibrium processes: among these are slow dynamics due to metal-nonmetal transitions in liquids, and wetting and dewetting transitions of liquid semiconductors. Since all these phenomena are related to a spatiotemporal range far wider than that treated by the present ab initio MD simulations, it is desirable that new progress in theoretical physics be stimulated, resulting in further developments in nanoscience.