The fully relaxed single-bond torsional potentials in typical conjugated systems were evaluated with the aid of ab initio self-consistent-field and Møller−Plesset second-order calculations and, additionally, with several recently developed variants of the density functional theory. For this systematic investigation, 1,3-butadiene, styrene, biphenyl, 2,2‘-bithiophene, 2,2‘-bipyrrole and 2,2‘-bifuran have been selected as model molecules. As representative examples for nonconjugated systems, the molecules n-butane and 1-butene have been treated at the very same calculational levels. For all conjugated molecules, the electron correlation corrections to the self-consistent-field torsional potentials, as obtained with the density functional methods, are dramatically different from those resulting from the more conventional Møller−Plesset second-order approximation. For those cases where experimental data for torsional barriers are available, the self-consistent-field and the Møller−Plesset second-order results agree reasonably, whereas the density functional results consistently predict too large barriers. This behavior is most probably caused by an overestimation of the stability of the planar π-systems by the density functional theory variants in question.

Full geometry optimizations on oligoenes have been performed with Hartree–Fock and density functional theory in combination with double zeta and triple zeta quality basis sets with primary focus on the degree of bond length alternation and on the energy gap. Monitoring the dependence of the computed properties on the oligomer size provides new insights into the reliability of the calculations, which are analyzed in terms of dynamical and nondynamical electron correlation. Our theoretical bond length alternation values for the oligomers of polyacetylene extrapolate to significantly smaller values than what has been established by experiments and earlier theoretical predictions. The exact exchange mixing to the exchange-correlation functional not only improves the agreement of the theoretical gap of oligoenes with experimental excitation energies but also increases the computed bond length alternations. Based on a newly proposed one parameter functional of Becke, the effect of the exact exchange mixing has been further identified, and a practical measure of nondynamical correlation energy has been suggested. The popular Becke–Lee–Yang–Parr density functional produces bond alternation values that are unacceptably small.

Geometric and vibrational spectroscopic data (rotational constants, bond distances and angles, vibrational frequencies, IR intensities, and OH/OD isotope effects) of phenol, benzaldehyde, and salicylaldehyde as calculated at various levels of theory (HF/6-31G(d,p), HF/6-311++G(d,p), MP2/6-31G(d,p), B3P86/6-31G(d,p), BLYP/6-31G(d,p), B3LYP/6-31G(d,p), and B3LYP/6-311++G(d,p)) are reported. The theoretical results are discussed mainly in terms of comparisons with available experimental data. For geometric data (rotational constants and bond distances) the best agreement between theory and experiment is obtained at the MP2 and B3LYP levels. B3P86 calculated data are slightly worse, while HF and BLYP calculations yield distinctly too small and too large bond distances, respectively. B3LYP calculated vibrational spectroscopic data excellently agree with experimental IR spectra for phenol, phenol-OD, and benzaldehyde, and with minor restrictions, also for salicylaldehyde and salicylaldehyde-OD. Considering frequency sequences, IR intensities, and OH/OD isotope effects, reliable and consistent assignments are given. BLYP and B3P86 calculated vibrational spectroscopic data are slightly worse, whereas MP2 and HF calculations suffer from several shortcomings that are already known from calculations of smaller molecules, such as benzene.

No abstract

IR gas-phase and quantum chemical data that characterize the intramolecular hydrogen bonds in a series of 2-hydroxybenzoyl compounds (Ph(OH)(COY):  Y = Cl, OH, SH, OCH3, SCH3, H, CH3, Ph, N(CH(CH3)CH2)2CH2, N(CH2CH2)2CH2, N(CH3)2, NHCH3, NH2, NHNH2) are reported and discussed. The sequences of ν(OH) and of ν(CO) frequencies calculated at HF/3-21G, HF/6-31G**, and B3LYP/6-31G** computational levels agree reasonably well with the corresponding experimental sequences. The consistency of the calculated data is demonstrated by exploiting several relations between calculated frequencies, force constants, and bond distances. It is shown that the sequence of hydrogen bond strengths, as defined either by ν(OH) frequency shifts or by R(O−H) bond distances, may qualitatively be interpreted in terms of two quantities, the R(H···O) hydrogen bond distances and the Q(O) oxygen partial charges, which characterize the intrinsic differences between the carbonyl groups.

The ground-state properties of the charge tranfer complexes formed between NH 3 and the halogens F 2 , ClF, and Cl 2 were investigated systematically with the aid of ab initio calculations. Extended basis sets were applied and electron correlation was included at different levels of Møller-Plesset perturbation theory and coupled cluster expansions. Not only the level of electron correlation and the basis set chosen are important, but also the use of the counterpoise correction to the basis set superposition error directly in the course of the geometry optimization has a significant influence on the calculated equilibrium structures. The trends in other ground-state properties of these complexes, such as vibrational spectra, infrared intensities, selected harmonic force constants, dipole moments, and nuclear quadrupole coupling constants are discussed and compared to experimental data. The strength of the intermolecular interaction in this series increases in the order NH 3 -F 2 , NH 3 -Cl 2 , and NH 3 -ClF.

The present work demonstrates the existence of cyclic structures in complexes of the CF3H molecule with hydrogen fluoride clusters, (HF)1 ≤ n ≤ 3, due to the formation of two types of hydrogen bonds, C−F···H−F and C−H···F. For the case of n = 3, a significant blue shift of the ν(C−H) stretching vibrational mode of fluoroform of about 60 cm-1 is predicted. This is indeed the largest theoretical estimate of a blue shift in a C−H···X system ever reported in the literature. The present analysis of the geometrical, energetical, and spectroscopic features of the cyclic CF3H−(HF)1 ≤ n ≤ 3 structures shows that the C−F···H−F bond is stronger than the C−H···F interaction. Therefore, they may be referred to as blue-shifted rather than blue-shifting complexes.

No abstract