| Tóm tắt: | A most striking feature of concentrated metal-ammonia solutions (excluding cesium) is their separation into two liquid phases below a certain critical temperature (Tc; ca. -42°C for sodium-ammonia). Pitzer, over 20 years ago, suggested that this phenomenon in metal-ammonia solutions is the analogue within the liquid-ammonia matrix of the liquid-vapor separation that accompanies the cooling of a nonideal alkali metal vapor. In this communication we reexamine Pitzer's hypothesis in the light of the considerable advances made recently both in the theoretical and experimental study of supercritical fluid alkali metals, doped semiconductors, metal-ammonia, and metal-methylamine solutions. For the entire spectrum of gaseous- and matrix-bound systems, there exists a fundamental link between this type of thermodynamic condensation phenomenon and the major constitutional changes that must occur in the electronic properties of a system as it moves through a metal-nonmetal (M-NM) transition. Critical electron densities (nc) at the metallic onset for all systems are well described by the simple experimental relation nc1/3a*H = 0.26 ± 0.05, where a*H is a characteristic radius associated with the electron state wave function in the low-electron-density (nonmetallic) regime. This relation is therefore applicable to both gaseous- and matrix-bound systems, and its apparent universality (Edwards, P. P.; Sienko, M. J. Phys. Rev. B: Condens. Matter 1978, 17, 2575) now extends over almost 9 orders of magnitude in nc and approximately 600 Å in a*H. Our recent studies of lithium-methylamine solutions are used to highlight the matrix dependence of both phenomena in metal solutions; the liquid-liquid phase separation in this system occurs around 13 mol % metal, compared to approximately 4 mol % metal in sodium-ammonia solutions. In both solvent systems this critical composition also marks the onset of M-NM transitions for T > Tc. We report calculations based on the adiabatic cavity model of Jortner for the solvated electron which lead us to suggest that the observed differences in critical compositions for meta-ammonia and lithium-methylamine solutions may simply arise from the differences in the radial extension of the isolated, solvated electron wave function. © 1981 American Chemical Society.
|
|---|