MICHAEL KASHA 37 



to the 77 ^ TT* transition at l/OOA of ethylene. Consequently, they 

 investigated the possibility that an ii -^ a* and n -^ 77* transition 

 could account for the weaker absorption bands occurring in for- 

 maldehyde (1900A, e < 2000; 2900A, e = 18). (The orbital pro- 

 motion designations tt -^ tt*, n -^ ir*, etc., were introduced in 1950 

 by Kasha, 17). Since McMurry was able to calculate approximately 

 the observed intensity of the 1900A band of formaldehyde consider- 

 ing an n -^ a* transition, this band was so assigned. (However, Simp- 

 son and Barnes (34a) have revised this assignment.) Although a calcu- 

 lation could not be made at the time to establish the nature of the 

 very weak 2900A absorption band, it was assigned as an n -> tt* transi- 

 tion essentially by the process of elimination of possibilities; however, 

 the assignment agreed with the fact that the unique n -^ tt* (singlet- 

 singlet) transition in formaldehyde should be formally forbidden 

 on symmetry grounds. In a subsequent section the excited state con- 

 figurations, selection rules, and polarizations will be developed. It 

 is of interest to recall that no direct physical evidence for the assign- 

 ment of the weak long wavelength band of carbonyl compounds 

 as an n -^ tt* transition was available at the time the assignment was 

 made. 



D. The n -^ tt* Transitions in N-Heterocyclics 



The low ionization potential of the lone-pair or /z-orbital electrons on 

 the O-atom was one of the plausibility arguments adduced by Mulliken 

 and McMurry (27, 22) for the assignment of the longest wavelength 

 ultraviolet absorption band of carbonyl compounds, as a transition 

 of essentially n -> tt* type. In the case of N-heterocyclic aromatic 

 molecules, it was considered generally that the expected higher ioniza- 

 tion potential of lone-pair electrons on the aza-nitrogen would pre- 

 clude the occurrence of lowest electronic transitions oi. n ^> tt* type. 

 There were several spectroscopic studies in the literature which seemed 

 to indicate that in N-heterocyclics the lowest singlet-singlet electronic 

 transitions involved only tt ^ tt* promotion. Thus, the spectrum 

 of pyridine as studied by Sponer and Stucklen (37) and later by 

 Sponer and Rush (35) , and the spectra of the diazines as studied 

 by MaccoU (20, cf. 32) and by Halverson and Hirt (11) were as- 

 sumed to fit in with such an interpretation. 



However, on the basis of studies of the emission characteristics 

 and solvent effects, in 1950 Kasha (17) proposed that the longest 

 wavelength singlet-singlet absorption bands or absorption shoulders 

 in several N-heterocyclics (especially pyridine, pyrazine, phenazine) 



