Microscopic structure of NiCl₂ in methanol

A.K. Adya (Univ. Abertay Dundee), O.N. Kalugin (Univ. Abertay Dundee / Kharkiv Univ.), H.Fischer, D. Sullivan (ILL).

Fundamental research on non-aqueous electrolyte solutions have catalysed their wide technical applications: in many fields non-aqueous electrolyte solutions are actually competing with other ion conductors, especially at ambient and at low temperatures, due to the high flexibility based on the choice of numerous solvents, additives, and electrolytes with widely varying properties. High-energy primary and secondary batteries, wet double-layer capacitors and supercapacitors, electrodeposition and electroplating are some devices and processes for which the use of non-aqueous electrolyte solutions has brought the biggest success. Other fields where non-aqueous electrolyte solutions are broadly used include electrochromic displays and smart windows, photoelectrochemical cells, electromachining, etching, polishing, and electrosynthesis. In spite of wide technical applications, our understanding of these systems at a quantitative level is still lacking. For instance, the application of statistical mechanics theories to ion-molecular systems containing multicharged ions, especially d-block metal cations, still remains an unsolved problem. The main reason for this is the absence of detailed information about the nature and strength of ion-molecular interactions and their influence on structural and dynamic properties of non-aqueous electrolyte solutions. Such information at a microscopic level can be obtained by using direct experimental methods i.e., those methods for which the space and time dimensions are comparable to the molecular dimensions and characteristic time of molecular processes, respectively. The isotopic substitution technique allows us to sit on a specific atom and view its atomic surroundings in terms of pair-distribution, number of co-ordinating solvent molecules and , their orientations towards the ion. These details can then be used to construct molecular models of ion solvation shells. The ILL has played a vital role in providing high neutron flux and stable instrumentation, the two factors crucial for the success of these experiments.

Neutron scattering measurements were performed on 1.4 molal solutions of nickel chloride in methanol at room temperature using the D4 diffractometer, employing isotopic substitution on both Ni and Cl, and with H/D substitution on the solvent.

Figure 1: The experimental pair-distribution functions representing the Ni2+ solvation shell obtained by using first-order a) and second-order b, c) difference techniques of neutron diffraction.

Figures 1 and 2 show pair-distribution functions around Ni²⁺ and Cl-ions obtained by using first-order, and second-order neutron diffraction difference techniques. The peaks in the total nickel pair-distribution function, G_Ni(r) (see Fig. 1a), obtained by isotopic substitution on Ni (nat/62) in CD₃OH and CD₃OD have been assigned to ion-solvent interactions; the H/D contrast gives negative and positive peaks for Ni-H(hydroxyl) interactions in the two solvents, respectively. The second-order difference technique has been used to isolate the Ni-H and Ni-Cl partial pair-distribution functions. The G_NiCl(r) partial (see Fig. 1c) shows that chloride ion penetrates significantly into the first solvation shell of Ni²⁺ and contributes to the first (Ni-O) peaks in G_Ni(r) (see Fig. 1a). The co-ordination numbers calculated from the pair-distribution functions show that the solvation shell of Ni²⁺ comprises one chloride ion and four methanol molecules (see Fig. 3a).

Figure 2: The experimental pair-distribution functions representing the Cl&endash; solvation shell obtained by using first-order a) and second-order b) difference techniques of neutron diffraction.

The first two peaks in the first order difference function or total chlorine pair-distribution function, G_Cl (r) (see Fig. 2a), obtained by isotopic substitution of Cl (nat/35/37) in CD₃OH and CD₃OD have been assigned, as in the case of the cation, to interactions of the anion with hydrogen and oxygen of the OH-group. The H/D contrast can clearly be seen in terms of negative and positive peaks for Cl-H(hydroxyl) interactions in the two solvents. The second order difference technique has been used to isolate the Cl-H partial pair-distribution function (see Fig. 2b).

In order to assign those distances in the G_Ni(r) and G_Cl(r) total pair-distribution functions that could not be identified purely by using the isotopic substitution, and to clarify the spatial arrangement of the first solvation shells of the Ni²⁺ and Cl-ions, molecular modelling was performed. First, geometry and charge distribution of a single methanol molecule were obtained by using ab initio quantum chemical calculations with the 6-31G** basis set. Then, geometry of the Ni[(CH₃OH)₄ Cl]⁺ and Cl(CH₃OH)^-₄ complexes was optimised by using the molecular mechanics MM+ method. The resulting three dimensional configurations of the ion-solvation shells are shown in Fig. 3 for Ni ²⁺ in a) and Cl^- in b).

Figure 3: Three-dimensional configurations of Ni2+ a) and Cl- b) solvation shells obtained by using molecular modelling and confirmed by neutron-diffraction isotopic substitution.

The above results have demonstrated the power of using complementary techniques in resolving structural details for non-aqueous electrolyte solutions at a microscopic level. Our ultimate aim is to refine the model potentials used in molecular dynamic simulations so as to reproduce not only static structure but also dynamics.

These potentials will then be used for prediction of macroscopic properties of non-aqueous electrolyte solutions in regimes not currently accessible by experiments but having important practical significance.

Acknowledgements

We are grateful to Dr. R. A. Howe (University of Leicester) for providing us with the isotopes and Dr. A. C. Barnes for help with the NDIS experiments.