PHOTOCHEMISTRY 9 



reduce its number of oscillational degrees of freedom and thereby to lower 

 the probability of internal conversion. It seems reasonable to assume 

 that this effect of binding the molecule to its surroundings has a relatively 

 small effect on the initial transition from F to M (or F' toF), since the 

 molecule and its surroundings momentarily have available an energy 

 excess equal to the difference between the energy of the photon and the 

 energy difference betw'een levels F and N. 



If the preceding simple explanation is correct, the difference between 

 the energ}^ levels F and .1/ can be determined in two independent ways, 

 which should yield identical results. The difference in potential energy 

 between levels F and .V is measured approximately by the quantity hp.i, 

 corresponding to the long-wave-length limit of the normal fluorescence. 

 Similarly, the energy difference between M and N is approximately equal 

 to hvs for the long-wave-length limit of the long-lived fluorescence. The 

 energy difference between F and AI is therefore equal to hv^ — hva. This 

 same energy difference should be obtainable from measurements, at two 

 or more temperatures, of the half life of the phosphorescence. In terms 

 of the simple Jablonski mechanism, e in the following equation should be 

 equal to Aj^s — hv^: 



For all cases for which the experimental evidence is available, this is 

 approximately true. Although there are minor and understandable dis- 

 crepancies (Pringsheim, 1949, p. 441), the available data support this 

 interpretation. The measured values (Kasha, 1947) of the M-F energy 

 difference lie in the range 5-40 kcal/mole. The lower values belong to 

 dyes, and the higher ones to simpler molecules, such as the aromatic 

 hydrocarbons. Since e~'^*^ will be a very small fraction for the higher 

 values of e, phosphorescence will be very inefficient and probably unde- 

 tectable in these cases. 



LONG-LIVED ENERGETIC STATES 



It Avas formerly maintained by many photochemists that fluorescence 

 and photochemical action are strictly complementary actions. How- 

 ever, a large amount of information, chiefly qualitative but in part quan- 

 titative, which is definitely incompatible with this belief, has accumulated. 

 Efficient photochemical reactions are known which involve compounds 

 whose fluorescence yield is small, even in dilute solutions in inert solvents. 

 Particularly in some dye-sensitized photooxidations involving weakly 

 fluorescent sensitizers (Shpol'skii and Sheremet'ev, 1936; Franck and 

 Livingston, 1941), the absorbed photon has too small an energy to dis- 

 sociate the absorbing molecule. At least in these cases a long-lived 

 excited state must be an intermediate in the photochemical reaction. 



To demonstrate this, let us consider a particular example: the auto- 



