Theoretical modeling of magnesium ion imprints in the Raman scattering of water.

Hydration envelopes of metallic ions significantly influence their chemical properties and biological functioning. Previous computational studies, nuclear magnetic resonance (NMR), and vibrational spectra indicated a strong affinity of the Mg(2+) cation to water. We find it interesting that, although monatomic ions do not vibrate themselves, they cause notable changes in the water Raman signal. Therefore, in this study, we used a combination of Raman spectroscopy and computer modeling to analyze the magnesium hydration shell and origin of the signal. In the measured spectra of several salts (LiCl, NaCl, KCl, MgCl(2), CaCl(2), MgBr(2), and MgI(2) water solutions), only the spectroscopic imprint of the hydrated Mg(2+) cation could clearly be identified as an exceptionally distinct peak at approximately 355 cm(-1). The assignment of this band to the Mg-O stretching motion could be confirmed on the basis of several models involving quantum chemical computations on metal/water clusters. Minor Raman spectral features could also be explained. Ab initio and Fourier transform (FT) techniques coupled with the Car-Parrinello molecular dynamics were adapted to provide the spectra from dynamical trajectories. The results suggest that even in concentrated solutions magnesium preferentially forms a Mg(H(2)O)(6) complex of a nearly octahedral symmetry; nevertheless, the Raman signal is primarily associated with the relatively strong metal-H(2)O bond. Partially covalent character of the Mg-O bond was confirmed by a natural bond orbital analysis. Computations on hydrated chlorine anion did not provide a specific signal. The FT techniques gave good spectral profiles in the high-frequency region, whereas the lowest-wavenumber vibrations were better reproduced by the cluster models. Both dynamical and cluster computational models provided a useful link between spectral shapes and specific ion-water interactions.