86 



LINSCHITZ: I think Pringsheim's book (14) appeared too early for this 

 theory to be included. But the shape of the quenching curves and concentrations 

 agree pretty well with Forster's quenching theory. I don't want to get into this 

 problem as such unless it is interesting. I want to talk in some more detail 

 about the process of resonative energy transfer, which Forster treats also in 

 detail. Kallmann's experiments show that the probability of such energy trans- 

 fer is so great that the energy may go through a thousand xylene molecules to 

 reach an anthracene before it is lost by thermal quenching in the xylene. This 

 is itself so strong that no xylene fluorescence can be observed in the pure sol- 

 vent. The actual times necessary for the transfer are really uncertain. I wish 

 I had some better estimates than the ones I want to give, but the data just are 

 not at hand. At any rate, one can make rough estimates of the jumping frequency 

 of the energy by saying that this frequency would be very roughly the resonance 

 energy arising from the coupling of two oscillators, divided by "h". If we take 

 the resonance energy to be even as small as 0.01 of a volt, then the frequency 

 comes out to be 10*2 jumps per second. 



ALLEN: How do you define this resonance energy? 



LINSCHITZ: This is just the energy of interaction of two dipoles at distance 

 R. The resonance energy will go as the product of the dipole moments times a 

 function involving the angle between the dipoles divided by R°. The point here to 

 be noted is that this energy is extremely sensitive to the separation of the mole- 

 cules. The dipole moments involved here are the total transition moments ob- 

 tained from„integrated absorption of the molecule. You get R values of the order 

 of 50 to 60 A, corresponding to, say, 50% transfer probability in 10" 8 seconds. 



KASHA: How does the collision frequency compare with the time for energy 

 transfer from the primary excited molecule? 



LINSCHITZ: Do you mean for these distances? 



KASHA: You calculated a time from the resonance frequency as you put it. 



LINSCHITZ: 10" 12 seconds for interaction energy of 0.01 volt. 



KASHA: That has to compete with collisions between solvent molecules, 

 doesn't it? 



LINSCHITZ: It has to compete with quenching of the energy in the solvent 

 molecule, that is true. The probability of this isn't known actually because of 

 the unknown increased lifetime for radiation in the solvent molecule due to the 

 coupling effect I spoke of in answer to Dr. Fano's question. Since you have no 

 way of knowing what the true radiative lifetime is in the solvent, the absence of 

 solvent luminescence doesn't tell you very much about the quenching probability. 

 The way you can estimate the net lifetime in the solvent alone is to measure the 

 yield of the competing process of energy transfer into the solute, and then esti- 

 mate the inherent probability of this latter transfer. 



KASHA: It may be a minor point, but the idea is then if you have liquid ben- 

 zene all by itself and you excite it with the light which benzene is capable of ab- 

 sorbing you say that no light is observed as emission because the lifetime is too 

 long. How long do you have to wait for that result? 



LINSCHITZ: What you are trying to explain is the existence of the effect of 

 self- quenching. This can be explained by assuming that the coupling of the 

 emitting molecules with each other increases the radiative lifetime, so that even 



