Search

The potential energy surface of the ground state He+Cl2(1Σg) is calculated by using the perturbation theory of intermolecular forces and supermolecular Møller–Plesset perturbation theory approach. The potential energy surface of the first excited triplet He+Cl2(3Πu) was evaluated using the supermolecular unrestricted Møller–Plesset perturbation theory approach. In the ground state two stable isomers are found which correspond to the linear He–Cl–Cl structure (a primary minimum, De=45.1 cm−1, Re=4.25 Å) and to the T‐shaped structure with He perpendicular to the molecular axis (a secondary minimum, De=40.8 cm−1, Re=3.5 Å). The small difference between these geometries is mainly due to the induction effect which is larger for the linear form. The results obtained for the T‐shaped minimum are in good agreement with the excitation spectroscopy experiments which observed only the T‐shaped form [Beneventi et al., J. Chem. Phys. 98, 178 (1993)]. In the lowest triplet states correlating with Cl2(3Πu), 3A′ and 3A″, the same two isomers correspond to minima. Now, however, the T‐shaped form is lower in energy. The 3A′ and 3A″ states correspond to (De,Re) of (19.9 cm−1, 3.75 Å) and (30.3 cm−1, 3.50 Å), respectively, whereas the linear form is characterized by (19.8 cm−1, 5.0 Å). The binding energy for the T form in the lower 3A″ state is in good agreement with the experimental value of Beneventi et al.

Nonadditive contribution to the interaction energy in water trimer is analyzed in terms of Heitler–London exchange, SCF deformation, induction and dispersion nonadditivities. Nonadditivity originates mainly from the SCF deformation effect which is due to electric polarization. However, polarization does not serve as a universal mechanism for nonadditivity in water. In the double‐donor configuration, for example, the Heitler–London exchange contribution is the most important and polarization yields the wrong sign. Correlation effects do not contribute significantly to the nonadditivity. A detailed analysis of the pair potential is also provided. The present two‐body potential and its components are compared to the existing ab initio potentials (MCY) as well as to empirical ones (RWK2,TIP,SPC). The ways to improve these potentials are suggested.

The Heitler–London (HL) exchange energy is responsible for the anisotropy of the pair potential in methane. The equilibrium dimer structure is that which minimizes steric repulsion between hydrogens belonging to opposite subsystems. Dispersion energy, which represents a dominating attractive contribution, displays an orientation dependence which is the mirror image of that for HL exchange. The three‐body correction to the pair potential is a superposition of HL and second‐order exchange nonadditivities combined with the Axilrod–Teller dispersion nonadditivity. A great deal of cancellation between these terms results in near additivity of methane interactions in the long and intermediate regions.

Ab initio calculations are carried out using a doubly polarized basis set. Dispersion, evaluated by second‐order Møller–Plesset perturbation theory (MP2), is found to have a profound influence on the stabilities and structures of the H‐bonded complexes. The contribution of dispersion to the H‐bond energies of H2S‐‐HF and H2S‐‐HCl is 44% and 69%, respectively, placing this attractive term second in magnitude only to electrostatics. Reductions of the intermolecular distance of 0.17 and 0.34 Å result from inclusion of correlation effects. Nevertheless, the influence of dispersion upon the angular characteristics of the complexes is rather minor as the relative orientations of the subunits are controlled chiefly by electrostatic factors. The HF‐‐HSH geometry appears to be a true minimum on the potential energy surface but is much less stable than the H2S‐‐HF structure. Comparison of the above systems with previous results for H2O‐‐HF and H2O‐‐HCl reveals a number of regular patterns. Replacement of either first‐row atom of H2O‐‐HF with one from the second row equally diminishes the strength of the H bond; a further reduction to roughly half of the ΔE for H2O‐‐HF occurs when both O and F are exchanged. Comparison between the calculated and observed X‐‐Y distances suggests that the relative changes due to substitutions of O and F by S and Cl are predicted very well by MP2, indicating that this approach is capable of accurately reproducing relative (if not absolute) values of R(X‐‐Y) as well as ΔE. The contribution of dispersion to the interaction energy is magnified by each substitution by a second‐row atom; these exchanges also produce drastic increases in the correlation‐induced contraction of the H bond.

The combination of supermolecular Møller–Plesset treatment with the perturbation theory of intermolecular forces is applied in the analysis of the potential energy surface of Ar–NH3. Anisotropy of the self‐consistent field (SCF) potential is determined by the first‐order exchange repulsion. Second‐order dispersion energy, the dominating attractive contribution, is anisotropic in the reciprocal sense to the first‐order exchange, i.e., minima in one nearly coincide with maxima in the other. The estimated second‐order correlation correction to the exchange effect is nearly as large as a half ΔESCF in the minimum and has a ‘‘smoothing’’ effect on the anisotropy of ϵ(20)disp. The model which combines ΔESCF with dispersion energy (SCF+D) is not accurate enough to quantitatively describe both radial and angular dependence of interaction energy. Comparison is also made between Ar–NH3 and Ar–PH3, as well as with the Ar dimer.

This paper presents the first ab initio attempt to reconstruct the observed band profile of the stretching fundamental vFX (X=H,D) in the FX⋅⋅⋅O(CH3)2 hydrogen‐bonded system. The two‐dimensional potential energy surface V(rFH,RF⋅⋅⋅O) is evaluated by means of large basis set SCF calculations. The related force constants up to the fourth order are obtained via the analytical fit to a polynomial expansion. The vibrational problem is solved by means of a variational treatment which includes the effects of mechanical anharmonicity. The side bands of the stretching fundamental vFX are described in terms of the vFX ±nvFX⋅⋅⋅O combination bands in excellent agreement with experiment.

The effects of electron correlation on the calculated properties of the (HOHOH)− anion are studied using Møller–Plesset (MP) perturbation theory. With this technique, inclusion of corrections up to third order are shown to provide results quite similar to those obtained with an extensive CI approach when equivalent basis sets are used. Barriers to proton transfer between the two oxygen atoms at a fixed R(OO) distance are computed with a number of basis sets ranging from split‐valence 4–31G to triple‐valence with polarization functions on all atoms, 6–311G∗∗. Each successive enlargement of the basis set leads to a greater barrier. The second‐order correction to the energy reduces the Hartree–Fock barrier dramatically while subsequent inclusion of the third‐order energy results in an increase over the MP2 barriers. MP3 formalism is found capable of accurately reproducing CI results for both the barrier height and functional dependence of the correlation energy upon the proton position. The potential energy surface is calculated as a function of both the R(OO) distance and the position of the central proton. At the Hartree–Fock level, all basis sets yield a surface with two minima separated by a saddle point, representing the transition state for adiabatic proton transfer. The surface is flattened a great deal by inclusion of second‐ and third‐order corrections such that the barrier to proton transfer is considerably below the estimated zero vibrational level for protonic motion. Electron correlation effects are also responsible for an increase of about 3 kcal/mol in the hydrogen‐bond energy of the (HOHOH)− complex.

The H bonds in H2O–HF and H2O–HCl are studied and compared using ab initio molecular orbital methods and the results compared to experimental data. Basis sets used are: (i) triple valence 6‐311G∗∗ and (ii) double ζ with two sets of polarization functions. Electron correlation, included via second‐ and third‐order Møller–Plesset perturbation theory, is found to have profound effects on both systems, particularly H2O–HCl. Both H bonds are strengthened substantially with a concomitant reduction in length. H‐bond energies and geometries calculated at correlated levels are in excellent accord with available experimental information. In both systems, all levels of theory indicate the equilibrium geometry contains a pyramidal arrangement about the oxygen atom. However, the difference in energy between this structure and a C2v planar arrangement is found to be small enough that consideration of probability amplitudes in the ground vibrational level leads to nearly equal likelihood of observing either geometry. Agreement between experimental vibrational frequencies in H2O–HF and those calculated at correlated levels and involving quadratic, cubic, and quartic force constants is quite good. An explanation is offered for the increase in HX bond length which occurs at SCF and correlated levels upon H‐bond formation based upon nearly linear relationships between this length on one hand and subunit dipole moment and polarizability on the other. The dispersion energy is found to be a very sensitive, almost exactly linear function of the increase of H–X bond length. This energy contributes substantially to the weakening of the HX bond upon complexation.

Dispersion energy is calculated in the systems H2O–HOH, H2O–HF, H3N–HF, and HF–HF as a function of the intermolecular separation using a variety of methods. M≂ller–Plesset perturbation theory to second and third orders is applied in conjunction with polarized basis sets of 6‐311G∗∗ type and with an extended basis set including a second set of polarization functions (DZ+2P). These results are compared to a multipole expansion of the dispersion energy, based on the Unsöld approximation, carried out to the inverse tenth power of the intermolecular distance. Pairwise evaluation is also carried out using both atom–atom and bond–bond formulations. The MP3/6‐311G∗∗ results are in generally excellent accord with the leading R−6 term of the multipole expansion. This expansion, if carried out to the R−10 term, reproduces extremely well previously reported dispersion energies calculated via variation‐perturbation theory. Little damping of the expansion is required for intermolecular distances equal to or greater than the equilibrium separation. Although the asymptotic behavior of the MP2 dispersion energy is somewhat different than that of the other methods, augmentation of the basis set by a second diffuse set of d functions leads to quite good agreement in the vicinity of the minima. Both the atom–atom and bond–bond parametrization schemes are in good qualitative agreement with the other methods tested. All approaches produce similar dependence of the dispersion energy upon the angular orientation between the two molecules involved in the H bond.

The geometries, energetics, and vibrational spectra are calculated for the two complexes at the SCF and correlated MP2 levels using the 6‐31G∗∗ basis set, augmented by a second set of d functions on Cl. While correlation represents an important factor in the binding of H3 N⋅⋅HCl, it contributes little to the stronger Li bond. Unlike the HCl stretch νs which decreases substantially in frequency and is greatly intensified in H3 N⋅⋅HCl, the frequency of the LiCl stretch undergoes an increase and little change is noted in its intensity, conforming to prior spectral measurements. The intensities of the intramolecular stretching modes of NH3 are greatly strengthened by formation of a H bond and even more so for a Li bond. These intensity patterns are analyzed via atomic polar tensors which reveal that formation of a H bond dramatically lessens the ability of the electron density to shift along with the proton. A stretch of H–Cl hence leads to a large increase in molecular dipole moment. This ‘‘freezing’’ of the electron cloud is much smaller in the Li bond and its effect on the νs intensity is counteracted by a much reduced Li atomic charge in the complex. Another distinction between the H and Li bonds relates to the destination of charge transferred from the NH3 subunit which accumulates on Cl in the former case but on Li in the latter.

Frequencies and intensities are calculated by ab initio methods for all vibrational modes of the 1:1 H3X–HF and 1:2 H3X–HF–HF complexes (X=N,P). The HF stretching frequencies are subject to red shifts, roughly proportional to the strength of the H bond, and to manyfold increases in intensity. Although the intramolecular frequency shifts within the proton acceptors are relatively modest, the intensities of the NH3 stretches are magnified by several orders of magnitude as a result of H bonding (in contrast to PH3 which exhibits little sensitivity in this regard). The frequencies and intensities corresponding to bending of the H3X–HF H‐bond rise with increasing H‐bond strength while the properties of the other intermolecular modes appear somewhat anomalous at first sight. The intensity patterns are analyzed by means of atomic polar tensors which reveal that intensification of the proton donor stretch is chiefly due to increasing charge flux associated with H‐bond formation. The different behavior of the N–H and P–H stretching intensities is attributed to the opposite sign of the hydrogen atomic charges in the two molecules. As a general rule, low intensities can be expected for intermolecular modes with the exception of those which involve motions of hydrogens that appreciably alter the magnitude or direction of a subunit’s dipole moment.