Semiexperimental and mass-dependent structures by the mixed regression method: Accurate equilibrium structure and failure of the Kraitchman method for ethynylcyclohexane.

Abstract

The mixed regression method for determination of molecular structures is reviewed and applied to the investigation of ethynylcyclohexane, using both semiexperimental and mass-dependent methods. This methodology provides an efficient and computationally affordable route to obtain accurate molecular reference data, preventing ill-conditioning in the structural least-squares determinations from experimental rotational constants. New supersonic-jet microwave measurements are reported to obtain inertial data for the axial and equatorial species of ethynylcyclohexane, together with all 13C isotopologues of the equatorial form. The semiexperimental equilibrium (reSE) and mass-dependent (rm(2)) structures of the molecule are compared with high-level ab initio optimizations, showing that both methods deliver compatible structures with accuracies of about 0.002 Å for bond lengths and 0.2° for bond angles. We confirm that dependable predicates can be obtained for a large variety of bonds. Finally, we verify that the substitution method completely fails to determine a reliable structure for the title compound.