Angew. Chem. Int. Ed.http://dx.doi.org/10.1002/anie.201302742 (2013)

Credit: © SCIENCE PHOTO LIBRARY

Elements of any given group in the periodic table generally have properties that are not too dissimilar, but notable exceptions exist; mercury is one such curiosity. Among other oddities, it is a liquid at room temperature — a particularly dense one — when all other metallic elements are solids. It has long been suspected that the peculiar characteristics of mercury arise from relativistic effects. In an international collaboration between France, New Zealand and Germany, Florent Calvo, Peter Schwerdtfeger and co-workers have now shown that this is indeed the case.

Including relativity in calculations means taking into account the fact that objects become heavier as they move faster. The resulting effects are usually negligible for light atoms, but for those with heavier nuclei — which exhibit larger electrostatic forces — inner-shell electrons reach relativistic speeds. These effects cause atomic orbitals to contract or expand, and this alters the bonding strength between atoms. Calvo, Schwerdtfeger and colleagues have now devised Monte Carlo simulations using a quantum diatomics-in-molecules method, derived from relativistic calculations for Hg2, to model the interactions between mercury atoms.

Heat-capacity plots were calculated for mercury clusters of different sizes (comprising 13, 19 or 55 atoms) and bulk mercury, with and without taking into account relativistic effects. Significant differences arose between the two scenarios, albeit not in a manner proportional to cluster size. On inclusion of relativity, the structures of the clusters were distorted and their melting points shifted — higher for Hg13 and Hg55, but lower for Hg19. These findings are consistent with complex many-body interactions that are known to occur in mercury: for clusters of increasing sizes, Hg–Hg bonding changes from van der Waals to covalent to metallic.

For the bulk material, accounting for relativity caused the calculated melting point to drop from 82 to −23 °C, much closer to the experimental value (−39 °C), and is consistent with mercury's liquidity at room temperature. The predicted density also shifted to a value very close to the experimental one. Further calculations revealed that the relativistic effects mostly arise from many-body contributions rather than spin–orbit coupling effects.