Alkali metal binding energies of dibenzo-18-crown-6: experimental and computational results

We have experimentally bracketed the gas phase Li+, Na+, K+, Rb+, and Cs+ affinities of dibenzo-18-crown-6 in the order 15-crown-5 (15C5) < dibenzo-18-crown-6 (DB18C6) < 18-crown-6 (18C6). Comparison with published experimental threshold collision-induced dissociation measurements yields enthalpies of binding at 298 K (in kJ mol(-1)) as follows: DB18C6(.)Na(+), -299 +/- 26; DB18C6(.)K(+), -221 +/- 19; DB18C6(.)Rb(+), -154 +/- 38; DB18C6(.)Cs(+), -136 +/- 35 (experimental values for the bracketing ligands binding Li+ are not available). Thus, dibenzo substitution of the crown makes it a weaker ligand toward the alkali metal ions, in contrast to dicyclohexano substitution, which was previously shown to make the crown a stronger ligand toward all the alkali metal ions. We have also calculated the enthalpies of binding between these metal ions and DB18C6 using both the LANL2DZ basis set and a modified version of the 6-31+G* basis set wherein diffuse functions are removed from carbon atoms and the K+, Rb+, and Cs+ basis functions are those of the Hay-Wadt (n + 1) effective core potential set. Full geometry optimizations were carried out at the HF/6-31+G* and HF/LANL2DZ levels of theory, and energies were also calculated for the HF/6-31+G* geometries at the MP2(full)/6-31+G* and B3LYP/6-31+G* levels. The computed enthalpies of binding for the K+, Rb+, and Cs+ complexes of DB18C6 fall between computed values in the literature for 15C5 and 18C6 complexes at both HF/6-31+G* and MP2(full)/6-31+G* levels, in good agreement with our bracketing experiments. The values at both levels of theory for the DB18C6 complexes of Li+ and Na+ are larger than those in the literature for both 15C5 and 18C6; the reasons for this discrepancy are not clear. (C) 2003 Elsevier Science B.V. All rights reserved.