2-(4-Azidophenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazole-3-oxide-1-oxyl (14), 2-(4-azidophenyl)benzimidazole-1-oxide-3-oxyl (16), 2-(4-azidophenyl)-1,2,6-triphenylverdazyl (19), 2-(3-azidophenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazole-3-oxide-1-oxyl (21), and (3-azidophenyl)-N-tert-butyl-N-aminoxyl (25) were photolyzed in frozen solution to give S = 3/2 state ESR spectra of the corresponding nitrenophenyl radicals with the following zero-field splitting parameters: |D/hc| = 0.277 cm(-1), |E/hc| < or = 0.002 cm(-1) (7 from 14); |D/hc| = 0.256 cm(-1), |E/hc| < or = 0.002 cm(-1) (8 from 16); |D/hc| = 0.288 cm(-1), |E/hc| < or = 0.002 cm(-1) (9 from 19); |D/hc| = 0.352 cm(-1), |E/hc| = 0.006 cm(-1) (10 from 21); |D/hc| = 0.336 cm(-1), |E/hc| = 0.004 cm(-1) (11 from 25). UB3LYP/6-31G computations and ESR spectroscopic analyses suggest that these are nitreno radicals, even para-linked systems with possible quinonoidal resonance forms. Neat samples of azidophenyl radicals 14 and 21 showed bulk paramagnetic behavior, consistent with the lack of close contacts in their crystal structures. Efforts to make photolabile coordination complexes of 14 and 21 with paramagnetic transition metal ions were unsuccessful: Cu(ClO4)2 x 6H2O instead oxidized them to the corresponding diamagnetic nitrosonium perchlorate salts.

[reaction: see text] Dinitrenes 4 and 5 both can form quinonoidal structures by conjugative bond formation. However, ESR spectroscopy detects a thermally populated, excited-state, triplet quinonoidal structure only for 4, with a zero-field splitting of |D/hc| = 0.0822 cm(-)(1), |E/hc| congruent with 0.0 cm(-)(1). The tendency to maintain aromaticity in the additional ring of 5 favors a dinitrene structure (with one less formal pi-bond) over a quinonoidal structure. The thermally populated quintet state of 5 has a zero-field splitting of |D/hc| = 0.287 cm(-)(1), |E/hc|

A number of 2,4-and 2,6-diazidopyridine derivatives were photolyzed under frozen matrix conditions (77 K), and their X-band electron spin resonance spectra (ESR) investigated. Typically, both mononitrene and dinitrene spectral features were observed. Cyano substituents reduced or eliminated spectral intensities, especially of the dinitrenes. The 2,4-dinitrenes were readily distinguished from 2,6-dinitrenes, since the major ESR resonance of the former typically occurred at about 3000 G, while that of the latter typically occurred at about 3300 G. Spectral line shape simulations suggest that the 2,6-dinitrenes have consistently larger zerofield splitting (zfs) than the related 2,4-dinitrenes. The 2,4-dinitrenes are estimated to have zfs of |D/hc| ∼ 0.21-0.24 cm -1 and |E/hc| ∼ 0.03-0.04 cm -1 , and the 2,6-dinitrenes to have zfs of |D/hc| Q ∼ 0.24-0.27 cm -1 and |E/hc| Q ∼ 0.040-0.05 cm -1 . The difference in spectral behavior is attributed to perturbation of spin density distributions and geometry-influenced interactions between the nitrene units in the 2,4-vs the 2,6-connectivities. Other substitutions on the pyridine had only small effects on the ESR spectra, although the photoefficiency of nitrene and dinitrene production was affected.

No abstract