Abstract: Seeking to better understand the origin of the different colors of emerald and ruby, both ab initio periodic and cluster calculations have been carried out. The calculations reproduce the interatomic distances measured for pure Be3Si6Al2O18 and Al2O3 as well as the Cr3+?O2? distances in emerald and ruby. The mean Cr3+?O2? distance for Be3Si6Al2O18:Cr3+ and Al2O3:Cr3+ is found to be practically equal to 1.97 Å, in agreement with recent experimental values. The present calculations confirm that the variations of optical properties due to Cr3+ impurities along the series of ionic oxides can be understood merely through the CrO69? unit but subject to the electric field due to the rest of the lattice ions. As a salient feature it is proved that changes in electronic density and covalency due to the internal field are not the cause of the color shift. Therefore, the red color of ruby is not due to the polarization of the electronic cloud around chromium as a result of the C3 local symmetry. The present study also demonstrates that the variation of the ligand field splitting parameter, 10Dq, induced by the internal electric field comes mainly from the contributions of first shells of ions around the CrO69? unit. As a consequence, 10Dq in emerald is not influenced by the internal field, as the contribution from Be2+ first neighbors is practically compensated by that of Si4+ second neighbors. In contrast, in ruby the t2g levels are shifted by the internal field 0.24 eV more than the eg ones, so explaining the color shift in this gemstone in comparison with emerald. This result is shown to arise partially from the asymmetric form of the internal electrostatic potential along the C3 axis in Al2O3.

Authorship

Download reference