Transient carbonyl nitrenes RC(O)N, formed during thermal- or photoinduced decomposition of carbonyl azides RC(O)N3, are highly liable to the Curtius rearrangement, producing isocyanates RNCO in almost quantitative yield. Contrary to common belief, we found a thermally persistent triplet carbonyl nitrene, FC(O)N, that can be produced by flash pyrolysis of FC(O)N3 in 49% yield. The computed CBS-QB3 activation barrier for the thermal decomposition of FC(O)N3 to FC(O)N is 29 kJ mol–1 lower than that for a concerted pathway producing FNCO.

CASSCF and multireference MP2 calculations were carried out on thiophene-S-oxide (TO) and selenophene-Se-oxide (SeO), comparing the energies of the ground state to the first two electronically excited singlet and triplet states, using constrained optimizations and multiple fixed S–O or Se–O distances. For both molecules, one of the two triplet states smoothly dissociates to yield O(3P) with little or no barrier. Single point calculations are consistent with the same phenomenon occurring for dibenzothiophene-S-oxide (DBTO). This provides an explanation for the inefficient unimolecular photochemical dissociation of O(3P) from DBTO despite a phosphorescence energy below that of S–O dissociation, i.e., that S–O scission probably occurs from a spectroscopically unobserved triplet (T2) state.

Computational methods are used to investigate the mechanism by which fluorination of acetylnitrene reduces the stabilization of the singlet configuration. ΔEST is made more positive (favoring the triplet state) by 1.9, 1.3, and 0.7 kcal/mol by the addition of the first, second, and third fluorine, respectively, at the CR-CC(2,3)/6-311(3df,2p)//B3LYP/6-31G(d,p) level of theory. Smaller effects observed with substitution of β-fluorines in propanoylnitrene derivatives and examination of molecular geometries and orbitals demonstrate that the effect is due to inductive electron withdrawal by the fluorines, rather than hyperconjugation.

This work describes the preparation of a selenium-modified TiO2 photocatalyst and a preliminary evaluation of its photocatalytic activity. Se-TiO2 displayed greater visible absorption than undoped TiO2 and was still capable of degrading quinoline at a slightly faster rate than undoped TiO2 under UV light. Se-TiO2 was also able to degrade organic molecules under purely visible light by a single electron transfer pathway. Irradiation with >435 nm light showed no evidence of efficient production of HO•-like species. Se-TiO2 was also examined under hypoxic conditions, where the Se atoms were capable of trapping photogenerated electrons as evidenced by XPS.

13C-modified TiO2 was prepared to facilitate study of the dopant atoms and trace their chemical fate throughout the process. In the preannealed material, NMR showed strong evidence of many Ti−O−C bonds. After annealing, surface-bound coke is a major component. NMR also showed that a washing step before annealing led to the generation of orthocarbonate (C(OR)4) centers, observed at 126 ppm, which are located deep inside the TiO2 particles. Both NMR and XPS confirmed the presence of small amounts of regular sp2-hybridized carbonate species in all briefly annealed samples, while annealing for longer times led to a reduction removal of the COn centers. Quantitative NMR also shows the degree of carbon loss that accompanies annealing. Some variation in the chemical degradation of quinoline is noted among the catalysts, but coke-containing TiO2 catalysts are not qualitatively better catalysts for use with visible light with this substrate.

Tungsten-modified titanium dioxide catalysts prepared from sol–gel methods and obtained commercially were compared for their photocatalytic activity using mechanistic probes designed to examine chemical pathways of oxidation. No special visible absorbance was noted for the sol–gel catalysts. However, an increase in the single-electron transfer chemistry with the presence of WOx was noted, and a distinct wavelength dependence on the product ratios.

The initial step of TiO2-mediated photocatalytic degradation of dimethyl phenylphosphonate (DMPP), labeled with or deuteria in the methoxy groups, results in products due to ring hydroxylation and demethylation. The labeling experiments clearly demonstrate that the methyl group is lost, rather than a methoxy group, resulting in a labeled phosphonic mono-acid. Results from deuterium isotope experiments are more ambiguous.

The early degradation product distributions from TiO2-mediated photocatalytic degradations of a series of multiply hydroxylated benzenes and their methoxylated analogs is reported. The methoxylated compounds show a distinct trend away from ring-opening reactions that are attributed to electron transfer chemistry towards hydroxylations and demethylations that are attributed to hydroxyl-type chemistry. The initial rates of the reactions suggest that this is not simply due to the availability of the extra carbon sites. It is instead hypothesized that the compounds with hydroxyl groups are bound to the TiO2 in such a way as to facilitate electron transfer and its subsequent chemistry. Addition of isopropyl alcohol to the degradations slows the degradations of the methoxylated compounds more than it does the hydroxylated compounds and particularly inhibits the demethylation and hydroxylation reactions. It is suggested that the question of whether an organic substrate must be bound to the photocatalyst at the time of excitation may be dependent on the type of chemistry being observed.