Aqueous Solvation of Ions
Location: CECAM-ETHZ, Zurich, Switzerland
Small inorganic ions in solution play a ubiquitous role in chemistry, physical chemistry and bio-chemistry. Solvated ions modify the viscosity and surface tension. They have a large role for forming electrical double layers around surfaces and thus they contribute to long range effective potentials between nano- and mesoscale particles. Ions are also involved in transport phenomena, for example across biological membranes. Chemical reactivity of ions encompass acid-base reaction, catalysis, especially transition metal cations, electrochemistry, etc. In all these examples, specific effects (dependence on the nature of the ion) of the interaction with water play a crucial role. For example, the ion selectivity upon adsorption in clays has been demonstrated to arise from the water-ion interaction rather than from the host-ion interaction. Other macroscopic quantities like viscosity, conductivity or surface tension depend drastically not only on the ion concentration but also of the nature of the solvated ions. It has also been demonstrated that the change in the hydration structure of transition metal cations during oxidation or reduction processes significantly influences the electrochemical properties. The hydration structure, at the microscopic level, of solvated ions has thus attracted a lot of attention and has progressed strongly during the last decade.
This progress has been made both on the experimental side and on the theoretical, modelling, side. From the theoretical side especially the emergence of realistic first-principles molecular dynamics has lead to a better understanding of these systems. Also on the experimental side several new results has has been published in the last 5 years. These include the x-ray and neutron diffraction results from Soper's group, nonlinear femtosecond mid-IR spectroscopy studied by Bakker, X-ray adsorption (EXAFS, XANES) measurements of D'Angelo. There are, however, still factors limiting the accuracy of the results. On the theoretical side they are obviously the models used. The pure DFT methods do not describe very well all properties of liquid water. For simple systems the empirical models can even work better but in the case of more complex systems, like ions or acids at higher concentrations, the predictive power of (non-polarisable) empirical models is not very good. On the experimental side, the raw data usually needs significant amount of processing and very often in this post-processing procedure some model of the system is used. It would be very useful for the simulation community to better understand the possibilities and limitations of the experimental techniques.
Away from the simple picture that ion solvation in water can be understood as the embedding of a charged sphere in a dielectric medium, a large range of situations and specific effects have been encountered. For example transition metal cations have been shown lately to have a variety of environments, even with a full shell d10 electronic configuration like coinage metals, Hg(II) etc. Pseudo-Jahn Teller effects were observed for Cu+ and Hg2+, with Cu+ being strongly bonded to two water molecules while the very similar Ag+ is bonded to four water molecules. Pb2+ has been also shown to expel one water molecule from its solvation shell in order to accomodate the 6s electron pair. Our knowledge and our current understanding of the solvation shells of these ions from 4th to 6th row is only limited. These ions play an important role for chemical reactivity as they may undergo oxydo-reduction reactions, be used as catalyst etc. They are also often poisonous through the replacement of inorganic ions in proteins. The first water hydration shell plays a crucial role in this reactivity as it has been demonstrated, for example, that the reorganisation of the hydration shell is strongly coupled with the oxidation of Cu+. One difficulty in order to characterise the hydration shell of these ions is that these structures are not rigid in liquid water. Even though the exchange time for water molecules between the first and second solvation shell may be long, the structure of the first hydration shell may be highly fluctuating as in the case of Cu2+, which undergoes frequent pseudo-rotations. Thus these structures can not simply be mapped to well defined gas phase or crystal structures.
The ability of water to dissolve ions stems from its very high dielectric constant that arises from its strong microscopic organisation due to hydrogen bonding. A crucial aspect of ion solvation is thus its interaction with the hydrogen bond network of water. For a long time, small inorganic ions were tentatively classified as "structure maker" or "structure breaker" as to within they would enhance, or disrupt respectively, the hydrogen bond network of water. Lately this simple picture has been challenged and demonstrated to be inadequate. Time-resolved infrared studies pointed out no influence of the ions on the rotational dynamics of water. Neutron diffraction data indicate that outside the first solvation shell the H-bond network is distorted with reorientation of water molecules but not disrupted in a manner that share common features with a pressure increase. These effects depend on the ion concentration and on the counter-ion. This brings new questions: What is the range of the reorientation induced by the ion? A similar question arises for the structure of water in porous media or in reverse micelles. What is the respective role of the ion and of the first hydration shell on the water arrangement further away? What is the effect of high concentrations? How is the H-bond network then disrupted? What is the role of ion pairing?
All these effects are strongly modified in the presence of interfaces or when the thermodynamics state is changed as for super-critical water. Ions at interface play a crucial role for the stability and reactivity of the interface, as is evidenced from the dependence of the water-vapour surface tension on ion concentration and ion nature and the high reactivity of sea water suspensions. In these situations, the polarisability of the ions, and in particular anions, is crucial. It is now proposed to lead to an excess concentration of anions like Cl- at the surface, unlike the expected depletion in continuum dielectric theory.
Finally, hydronium and hydroxide ions play a very special role in water solutions. Hydronium H3O+ is now regarded as an equilibrium between two structures, H9O4+ or Eigen cation, an H3O+ cation solvated by three water molecules, and H5O2+ or Zundel cation in which the excess proton is equally shared between two water molecules. The detailed mechanism of proton diffusion and its coupling with H-bond rearrangements in liquid water is still eluding, as well as the role of finite concentration on this equilibrium.
The experimental techniques to study the structure of the hydration shell of small inorganic ions at the microscopic level include diffraction measurements (neutron and x-ray), x-ray absorption spectroscopy, vibrational and time-resolved spectroscopies. Recently, terahertz spectroscopy has been also proposed to probe the range of the hydration shell. Much progress has been made in the last five-ten years in the field, still the interpretation of raw experimental data can benefit greatly from simulations. One goal of the proposed workshop is to facilitate this link between experiments and theory.
On the theoretical side, a lot of progress has been made in first-principle molecular dynamics simulations (FPMD). The solvation of many different types of ions have been now studied by FPMD: alcalines, alcaline-earth, transition metals, halogenures etc. As a result, density functionals and pseudo-potentials have been extensively tested and it is possible with FPMD to extend the range of studies on ion solvation. Different environments, chelation or reactivity (electrochemistry, hydrolysis) of ions are currently being studied. However, although FPMD is to a large extent able to capture specific effects, it suffers from not so good a description of the solvent. Indeed, DFT water (BLYP or PBE) is supercooled at 300 K and liquid (with a diffusion constant close to experimental water at standard conditions) only around 350 K. At 300 K, DFT water is indeed much overstructured. This stems mainly from the lack of long-range dispersion forces.
In comparison, the advent of classical polarisable force fields has opened the possibilty to have concurrently a good description of the solvent and of simple ions like halogenures and alcalines. With these methods it is then possible to study larger systems and reach finite concentrations or study interfaces. One open question that attracts a lot of attention currently is the parametrization of these forces fields. FPMD simulations are thought to be very valuable to this respect: they can either be used directly to calibrate the force fields or as a check of a given force field on equivalent systems at the same thermodynamical conditions (something that is impossible with gas phase quantum chemistry methods).
Finally, a third route for the study of solvation, adequate for infinite dilution, is the use of embedding. This represents an other compromise with the description of the ion and eventually a few solvent molecules at high level of sophistication while the rest of the solvent is represented at a lower level. The advent of continuum description of the solvent makes these approaches very cheap in computer time and in particular makes the evaluation of free energies quickly reachable.
Another goal of this workshop is thus to discuss the status and perspectives of these methods for the study of solvation. In particular we would like to discuss how their combination can be most efficient, along with experiments, to increase our knowledge of ion solvation in water.
Kari Laasonen (Aalto University School of Chemical Technology) - Organiser
Ari Paavo Seitsonen (École Normale Supérieure) - Organiser
Riccardo Spezia (Sorbonne Université & CNRS) - Organiser & speaker
Rodolphe Vuilleumier (Sorbonne Université - ENS-PSL) - Organiser