Determination of ionic hydration numbers, mechanisms and rates of exchange of coordinated water molecules and interaction energies between ions and water have been studied since Arrhenius. The observed exchange rates constants cover more than 18 orders of magnitude. The most inert (slow exchanges) are trivalent transition metal ions, while the most labile (fast exchanges) are alkali, alkaline earth and halide ions. Experimentally, exchange constants are determined mainly from nuclear magnetic resonance measurements. For labile complexes this information is not experimentally accesible and other techniques should be used. For example, specific information on solvation shells can be obtained, by exciting and detecting a dissolved probe molecule using ultrafast infrared nonlinear spectroscopy. In this type of experiment, the dynamics of the solvating water molecules are measured through the time-dependent response of the probe molecule. A disadvantage of this technique is that it probes the solvating water molecules indirectly, but, on the other hand, it is the only experimental technique so far that gives information on the dynamics of aqueous solvation shells. It provides information, for example, on the time scale on which the distance between a solvating water molecule and a dissolved ion/molecule changes.
Molecular Dynamics simulation prove to be an excellent way for obtaining a deeper insight on the structure and dynamics of aqueous solvation shells; it could complete the missing information hardly accessible due to experimental difficulties. For example, while the solvation free energy for a salt can be measured, it is impossible to separate by experiment cation and anion contributions. With computer simulations we can directly simulate a single cation or anion in any solvent. Moreover the microscopic nature of the exchange reaction can be understood by following directly the single molecules trajectories in a MD simulation.
The majority of earlier simulation studies on ion solvation has primarily been concerned with the structure and thermodynamics. Some studies were also concerned with solvent exchange processes, their kinetic and their dynamics. Most of them focused the discussion on residence times and exchange rates of solvent molecules in the first solvation shell of the ions. In this context Rey and Hynes introduced the use of the reactive flux method for the study of the exchange process (seen as an activated one), which could be applied also to the study of slow exchanges.
Part of my research is dedicated to the study of solvent exchange around ions. In particular we try to answer to the following three questions:
1) how do the static and dynamical properties of water exchange arond lithium ion vary with the change of thermodynamical conditions? One could expect that, varying the density and/or the temperature of the system, different behaviours should arise. We showed that the activated process show similar features for a large variety of thermodynamic conditions, while a change of the exchange rate is observed. The study was made over a wide range of points of the water PT diagram, spanning from liquid to supercritical conditions.
2) is there any connection between solvent shell exchange and ion mobility? In most studies the diffusion coefficient is usually calculated, but not much attention is payed to the interplay between ionic mobility and exchange processes. Given the different time scales over which the two phenomena relax, it should be immediate that they would not couple. Nevertheless, we showed that the interplay is two-fold: (i) the solvation exchange mechanism is affected by the ion diffusion and (ii) ionic diffusion is enhanced (hindered) by the disruption of local solvent structure.
3) what is the relation between ion and first shell molecules diffusion coefficient? For tightly bound ion-shell systems, it should be obvious that they are identical since the complex diffuses as a unit. Contrary to this expectation, this equality is not found in computer simulations for cases where it is manifest that no exchanges have taken place during the calculation. We explained which the cause of this inconsinstency is.