Search

The potential energy curves for the manifold of molecular states dissociating to the lowest transition metal ion states derived from the 3d(sup n+1) and 3d(sup n)4s(sup 1) occupations have been determined for selected transition-metal ion- rare gas systems. These curves have been computed using large basis sets, and the state-averaged complete- active-space self-consistent-field/multireference configuration interaction level of electron correlation treatment. In general, the families of curves derived from the 3d(sup n+1) and 3d(sup n)4s(sup 1) metal occupations are disjoint; however, for Ti(+) there is a strong mixing of the 3d(sup 2)4s(sup 1) and 3d(sup 3) asymptotes, because of the small separation between the asymptotes. This mixing leads to a failure of single-reference-based techniques; this is discussed in the context of our previous single-reference-based treatments.

The heat of formation of CH2OH at 298 K is determined to be -15.2 +/- 3.5 kJ/mol using high levels of theory. This is in good agreement with some recent experimental determinations. The assignment of the error bars is discussed in detail and compared with that assigned in previous theoretical calculations. The largest uncertainty arises from the calculation of the zero-point energy.

On the basis of extensive ab initio calculations, we estimate the metal-O2 binding energies of NiO2 and CuO2 to be 48 +/- 7 and 18 +/- 4 kcal/mol, respectively. We feel that the experimental estimate of 57 +/- 10 kcal/mol for the binding energy of NiO2 is slightly too large, while we are in complete agreement with the experimental estimate of 15 +10/-5 kcal/mol for CuO2. While the 1A1 ground state of NiO2 definitely has a side-on C(2v) structure, matrix isolation studies suggest that CuO2 has an end-on C(s) structure. Calculations at the coupled-cluster singles plus doubles level with a perturbational estimate of triple excitations, CCSD(T), produce a 2A2 state with C(2v) as a global minimum. However, the entire 2A-double prime ground-state surface is exceedingly flat, precluding a reliable determination of the gas-phase equilibrium structure.

The calculated relative binding energies of 16 organic molecules to Mg(+) are compared with experimental results where available. The geometries of the ligands and the Mg(+)-ligand complexes arc optimized at the self-consistent field level using a 6-31G* basis set. The Mg(+) binding energies are evaluated using second-order perturbation theory and basis sets of triple-sigma quality augmented with two sets of polarization functions. This level of theory is calibrated against higher levels of theory for selected systems. The computed binding energies are accurate to about 2 kcal/mol.

Theoretical binding energies are reported for all first-row and selected second-row transition metal ions (M+) bound to benzene. The calculations employ basis sets of at least double-zeta plus polarization quality and account for electron correlation using the modified coupled-pair functional method. While the bending is predominantly electrostatic, the binding energies are significantly increased by electron correlation, because the donation from the metal d orbitals to the benzene pi* orbitals is not well described at the self-consistent-field level. The uncertainties in the computed binding energies are estimated to be about 5 kcal/mol. Although the calculated and experimental binding energies generally agree to within their combined uncertainties, it is likely that the true binding energies lie in the lower portion of the experimental range. This is supported by the very good agreement between the theoretical and recent experimental binding energies for AgC6H6(+).

Binding energies of Co(H2)n(+) (n = 1-3) and CrN2(+) computed using the modified coupled-pair functional approach agree well with measured values. The calculations show that both H2 and N2 bond much more strongly to Co(+) than to Cr(+). Theoretical studies also show that despite the similar polarizabilities of N2 and Ar, the CrN2(+) binding energy is twice as large as that of CrAr(+) due to the charge-quadrupole contribution. Finally, the bonding of H2 and N2 with Na(+) and Mg(+) is contrasted with that for Cr(+) and Co(+).

Ab initio calculations were performed to study the bonding of the first-row transition-metal ions with ethylene. While Sc(+) and Ti(+) insert into the pi bond of ethylene to form a three-membered ring, the ions V(+) through Cu(+) form an electrostatic complex with ethylene. The binding energies are compared with those from experiment and with those of comparable calculations performed previously for the metal-acetylene ion systems.

Three competing bonding mechanisms have been identified for the alkaline-earth metal superoxides: these result in a change in the optimal structure and ground state as the alkaline-earth metal becomes heavier. For example, BeO2 has a linear 3Sigma(-)g ground-state structure, whereas both CaO2 and SrO2 have C(2v)1A1 structures. For MgO2, the theoretical calculations are less definitive, as the 3A2 C(2v) structure is computed to lie only about 3 kcal/mol above the 3Sigma(-)g linear structure. The bond dissociation energies for the alkaline-earth metal superoxides have been computed using extensive Gaussian basis sets and treating electron correlation at the modified coupled-pair functional or coupled-cluster singles and doubles level with a perturbational estimate of the triple excitations.

Accurate ab initio calculations for all of the first-row transition-metal ions with He, Ne, and Ar were performed at the modified coupled-pair functional level of electron correlation treatment. It is concluded that there is a strong correlation between the binding energy and bond length, since the bonding is predominantly electrostatic. The calculated binding energies are found to be about 20 pct less than the experimental values due to the theoretical limitations.

The geometries of the molecules formed by the interaction of the first- and second-row transition-metal cations with methylene are optimized at the modified coupled-pair functional (MCPF) level of theory using large Gaussian basis sets, and their dissociation energies are computed employing both the MCPF and internally contracted averaged coupled-pair functional (ICACPF) methods. The computed binding energies are generally in good agreement with the available experimental results, although the calculations indicate that the experimental values for ScCH2(+), TiCH2(+), and NbCH2(+) are probably too large. The nature of the bonding in each case and trends in the bonding patterns across the transition-metal rows are discussed.

The bonding of Nb(2+) with CH2, C2H2, and C2H4 is studied by using electronic structure calculations that include high levels of electron correlation. The binding energy for NbCH2(2+) is in good agreement with the lower bound determined from the reaction with CH4 but is significantly smaller than the value determined from the binding energy and ionization potential of NbCH2(+). The calculations and a new interpretation of the experiment indicate that the larger value is in error primarily because the ionization potential of NbCH2(+) determined from bracketing charge-exchange reactions is too small. The computed binding energy of NbC2H2(2+) is in good agreement with experiment. The calculations show that the bonding is predominantly covalent in character for both NbCH2(2+) and NbC2H2(2+), whereas for NbC2H4(2+) the electronic states that are predominantly ionic and covalent are nearly degenerate. The trend in binding energies, CH2 greater than C2H2 greater than C2H4, is consistent with the energy required to prepare the ligands for bonding.

Theoretical calculations employing large basis sets and including correlation are carried out for Mg(+) with methanol, water, and formaldehyde. For Mg(+) with ethanol and acetaldehyde, the trends in the binding energies are studied at the self-consistent-field level. The predictions for the binding energy of Mg(+) to methanol and water of 41 + or - 5 and 36 + or - 5 kcal/mol, respectively, are much less than the experimental upper bounds, of 61 + or - 5 and 60 + or - 5 kcal mol, determined by using photodissociation techniques. The theoretical results are inconsistent with the onset of Mg(+) production observed in the photodissociation experiments, as the smallest absorptions are calculated at about 80 kcal/mol for both Mg(+)-CH3OH and Mg(+)-H2O, and these transitions are to bound excited states. The binding energy for Mg(+) with formaldehyde is predicted to be similar to Mg(+)-H2O. The relative binding energies are in reasonable agreement with experiment. The binding energy of a second water molecule to Mg(+) is predicted to be similar to the first. This suggests that the reduced reaction rate observed for the second ligand is not a consequence of a significantly smaller binding energy, at least for the smaller ligards such as those considered in this work.

The bonding of one to four waterligands to Mg(+) and Al(+) is studied. The bonding is very different from that found for multiple ligands bound to Na(+) or Cu(+). All the waters bind on the same side of the Mg or Al because of the polarization of the 3s orbital away from the ligands. Two benzene and methanol molecules also have a ligand-Mg-ligand angle of less than 180 deg with the angle determined by ligand-ligand repulsion.

Equilibrium structures for the ground and the selected low-lying excited states of the first- and the second-row transition-metal dimethyls and their positive ions were determined using the basis sets and correlation approach described by Bauschlicher et al. (1989). In particular, the dimethyl insertion products were examined to elucidate the trends in bonding. Results show that the bonding in the dimethyl transition-metal positive ions arises from the sd hybridization, leading, for most dimethyls, to a C-M-C bond angle of less than 134 deg.

The low-lying 1Sigma(g)(+), 3Pi(u), and 3Sigma(g)(-) states of N2(2+) in the immediate region of the minima in their respective potential curves are investigated. Spectroscopic constants for the 1Sigma(u)(+) state are computed for comparison with experiment. Approaches to computing the entire potential curves of many states N2(2+) are calibrated.