In the present work, we sought to improve our sparkle model for the calculation of lanthanide complexes
, SMLC
,in various ways: (i) inclusion of the europium atomic mass, (ii) reparametrization of the model within AM1 from anew response function including all distances of the coordination polyhedron for tris(acetylacetonate)(1,10-phenanthroline) europium(III), (iii) implementation of the model in the software package MOPAC93r2, and (iv) inclusionof spherical Gaussian functions in the expression which computes the core-core repulsion energy. The parametrizationresults indicate that SMLC II is superior to the previous version of the model
because Gaussian functions provedessential if one requires a
better description of the geometries of the complexes. In order to validate ourparametrization, we carried out calculations on 96 europium(III) complexes, selected from
Cambridge StructuralDatabase 2003, and compared our predicted ground state geometries with the experimental ones. Our resultsshow that this new parametrization of the SMLC model, with the inclusion of spherical Gaussian functions in thecore-core repulsion energy, is
better capa
ble of predicting the Eu-ligand distances than the previous version. Theunsigned mean error for all interatomic distances Eu-L, in all 96 complexes, which, for the original SMLC is0.3564 Å, is lowered to 0.1993 Å when the model was parametrized with the inclusion of two Gaussian functions.Our results also indicate that this model is more applica
ble to europium complexes with
![](/images/gifchars/<font color=)
beta2.gif" BORDER=0 ALIGN="middle">-diketone ligands. Assuch, we conclude that this improved model can
be considered a powerful tool for the study of lanthanide complexesand their applications, such as the modeling of light conversion molecular devices.