Sparkle/PM3 for the modeling of europium(III), gadolinium(III), and terbium(III) complexes

The Sparkle/PM3 model is extended to europium(III), gadolinium(III), and terbium(III) complexes. The validation procedure was carried out using only high quality crystallographic structures, for a total of ninety-six Eu(III) complexes, seventy Gd(III) complexes, and forty-two Tb(III) complexes. The Sparkle/PM3 unsigned mean error, for all interatomic distances between the trivalent lanthanide ion and the ligand atoms of the first sphere of coordination, is: 0.080 A for Eu(III); 0.063 A for Gd(III); and 0.070 A for Tb(III). These figures are similar to the Sparkle/AM1 ones of 0.082 A, 0.061 A, and 0.068 A respectively, indicating they are all comparable parameterizations. Moreover, their accuracy is similar to what can be obtained by present-day ab initio effective core potential full geometry optimization calculations on such lanthanide complexes. Finally, we report a preliminary attempt to show that Sparkle/PM3 geometry predictions are reliable. For one of the Eu(III) complexes, BAFZEO, we created hundreds of different input geometries by randomly varying the distances and angles of the ligands to the central Eu(III) ion, which were all subsequently fully optimized. A significant trend was unveiled, indicating that more accurate local minima geometries cluster at lower total energies, thus reinforcing the validity of sparkle model calculations. (author)