The conductor-like polarizable continuum model (CPCM) using several cavity models is applied to compute aqueous solvation free energies for a number of organic molecules (30 neutral molecules, 21 anions, and 19 cations). The calculated solvation free energies are compared to the available experimental data from the viewpoint of cavity models, computational methods, calculation time, and aqueous pKa values. The HF/6-31+G(d)//HF/6-31+G(d) and the HF/6-31+G(d)//B3LYP/6-31+G(d) with the UAKS cavities, in which radii are optimized with PBE0/6-31G(d), provide aqueous solvation effects in best agreement with available experimental data. The mean absolute deviations from experiment are 2.6 kcal/mol. The MP2/6-31++G(d,p)//HF/6-31+G(d) with the CPCM-UAKS(HF/6-31+G(d)) calculation is also performed for the base-catalyzed hydrolysis of methyl acetate in water.

No abstract

The selectivity of 5‘-TGGGT-3‘ and 5‘-CGGGC-3‘ sequences toward photoinduced one-electron oxidation was examined experimentally and by ab initio molecular orbital (MO) calculations. It was confirmed experimentally that G2 of 5‘-TG1G2G3T-3‘ is more reactive than G1, while for 5‘-CG1G2G3C-3‘ the selectivity is reversed, that is, G1 > G2. The ab initio MO analyses were performed to elucidate the difference of the selectivities between 5‘-TGGGT-3‘ and 5‘-CGGGC-3‘ sequences. For the 5‘-TGGG-3‘ sequence, the spin densities of G1 • and G2 • in neutral radical (5‘-TG1G2G3-3‘)• have a similar pattern, and the shapes of the corresponding radical orbitals are also very similar. It was concluded that the selectivity is due to the stability of the (5‘-TG1G2G3-3‘)• neutral radicals; that is, 5‘-TG1G2 •G3-3‘ is more stable in energy than 5‘-TG1 •G2G3-3‘. For the 5‘-CGGG-3‘ sequence, it was found that the spin density on N1 of G1 • in neutral radical (5‘-CG1G2G3-3‘)• is distinguishably different from the corresponding spin density of G2 •, which has a pattern similar to those of G1 • and G2 • in 5‘-TG1G2G3-3‘. The radical orbital (SOMO) of G1 • is delocalized on guanine base and up to the paired cytosine base, while the radical orbital of G2 • is essentially localized on guanine base. This drastic difference of the electron population in the radical orbitals, caused by the stacking interaction with the 5‘-side G of the opposite strand, can explain why G1 is more reactive than G2 in the 5‘-CG1G2G3-3‘ sequence.

We have investigated the decomposition pathway of dioxetanones 1c with a phenoxide anion group by the B3LYP/6-31+G(d) method together with the second-order multireference Møller-Plesset perturbation (MRMP) theory and propose charge-transfer-induced luminescence (CTIL) with polarization-induced branching excitation processes. In the gas phase, the thermal decomposition of 1c occurs by an asynchronous two-stage pathway without a discrete intermediate; that is, the initial O-O bond breaking to generate a charge-transfer (CT) diradical species is immediately followed by the subsequent C-C bond breaking with simultaneous back CT, which is responsible for the surface crossing at the avoided crossing. The activation energy is dramatically reduced from 19.4 to 3.8 kcal mol(-)(1) by the deprotonation of phenol meta-1d to its anion meta-1c, showing an important role of the endothermic CT. The odd/even selection rule for the chemiluminescence efficiency can be explained by the orbital interaction for the back CT between the carbonyl pi orbital and either a HOMO or a LUMO of the generated light emitters. To examine the accessibility of the chemically initiated electron exchange luminescence (CIEEL) route, we considered the solvent effects on the free-energy change of meta-1c by using continuum solvent models. The bending vibration mode of the CO(2) fragment is specifically considered. Borderline features emerges from the solution-phase CT reaction of meta-1c, which depends on the solvent polarity: one is a nonadiabatic or adiabatic back CT process (polarization-induced concerted CTIL), and the other is a radical dissociation, i.e., complete one-electron-transfer process (CIEEL).