The paper cited in the OP is a reference in this paper, which I found in my files during an episode of cleaning up old downloads:
N. Chandrasekhar and A.-N. Unterreiner, Time-resolved polaron dynamics in molten solutions of cesium-doped cesium iodide J. Chem. Phys. 127, 184509 (2007). In salts having small ions, like say, lithium fluoride, it is possible to stabilize dipolarons, having two electrons in a molecular ion like F-Li-F -2. In Cs/CsI melts, only one electron can be localized on a polaron, dipolarons do not exist.
Instead there is something called "Drude type electrons" which are not formally attached to any atom, but are "delocalized" - they wander.
This has certain effects on the systems electronic and radiation absorption spectrum from the latter paper:
Ultrafast relaxation dynamics of excess electrons were investigated in cesium-doped cesium iodide melt at various temperatures. It is shown that the excess electrons are predominantly polarons in agreement with earlier conclusions from steady-state absorption studies that with increasing cation size, the tendency of formation of bipolarons decreases. Furthermore, temperature-dependent measurements show that ionic diffusion is an important process which determines the relaxation dynamics of excess electrons in M–MX solutions. The activation energy obtained from the temperature dependent relaxation rates of excess electrons agrees well with those calculated from ionic self-diffusion coefficients. This is a further confirmation of our earlier conclusions that the activation energy involved in the relaxation process is nothing but the energy required to form a hole. We have further shown that the relaxation rates of excess electrons in alkali-metal-doped alkali halide melts serve to estimate the size of polarons. Finally, the nature of the relaxation dynamics of polarons is shown to be independent of the excitation region within the stationary absorption spectrum of excess electrons. This is in contrast to the observations of hydrated electrons in aqueous solutions and adds further support to the thermodynamic defect model proposed to explain the electrical and optical properties of excess electrons in molten alkali-metal-doped alkali halide mixtures.
This system arises in nuclear fuels and substances that can be isolated from used nuclear fuels, and I have been very interested in convincing my son that the properties of this system merit serious consideration in addressing certain very serious environmental issues that do not currently admit solution.
= new reply since forum marked as read