54 LIGHT AXD LIFE 



lar ring angular distortion frequency is excited, giving a very charac- 

 teristic vibrational pattern, in the n -^ tt* transitions. This has been 

 interpreted (10) as a consequence oi promotion of a hybridized 

 //-orbital electron, leading to changes in the ring lionding angles 

 during excitation, since the orbitals forming the ring are also hybrid- 

 ized. Mason (26) has also noted this point. 



/. Radical-Like Excited States 



The promotion of an (essentially atomic) //-orbital electron to a 

 77-antibonding-orbital leaves one electron in the //-orbital. This highly 

 localized single electron may impart radical-like behavior to the n,7T* 

 excited state. The promoted electron is not likely to be completely 

 localized, but may be found somewhat more probably at one atom 

 of the molecule in certain cases. It is interesting that the photo- 

 chemistry arising from the excited states of u,7t* type requires radical- 

 like, or diradical intermediates for its interpretation (16), whether 

 singlet or triplet excited states are involved. (Burawoy's assumption 

 of radical-like structures for the ground state of heteromolecules bears 

 no relation to the present discussion.) 



[Other characteristics of mainly spectroscopic interest have been 

 enumerated (17, M) but are omitted here.] 



VI. Spectra of N-heterocyclics 



A. Pyridine 



The pyridine molecule is the first N-heterocyclic which revealed 

 the presence of // — » tt* transitions (17) . Fig. 13 shows the dramatic 

 solvent effect observed. The curves shown were determined by a 

 continuous recording Gary spectrophotometer and show features ab- 

 sent from curves obtained by manual spectrophotometers (38) . The 

 region 2700 to 2900 Angstroms shows structure in a hydrocarbon 

 solvent in a region where the vapor spectrum shows very sharp lines. 

 This structure is certainly due to an n — > tt* transition since it 

 strongly blue-shifts in ethanol as solvent, and is absent in the absorp- 

 tion spectrum of benzene itself. However, a more remarkable effect 

 is the apparent sharpening of the main part of the band at 2500 A 

 upon changing from isopentane solvent to ethanol (Fig. 13), which 

 is of course the reverse of the usual blurring effect. This can only 

 mean that there is significant // -> tt* absorption even in the tt — > tt* 

 region at 2500 A. The increase of molar absorption coefficient from 

 1900 to 2500 with this solvent change facilitates the illusion of band 



