72 LIGHT AND LIFE 



is necessary before triplet-state absorption can be observed. Both 

 oxygen and nitric oxide quench with nearly unit encounter efficiency 



(11). 



It was natural to enquire whether paramagnetic ions also quenched 

 the triplet state, and the findings here were rather strange (12, 13) . 

 Whilst all paramagnetic ions were quenchers, and diamagnetic ions 

 were not, the quenching efficiency of an ion showed no other correla- 

 tion with magnetic susceptibility or number of unpaired electrons. 

 Clearly the effect could not therefore be accounted for in terms of 

 perturbations of the triplet state by the magnetic field of the quencher. 



An explanation of this type of quenching has recently been given 

 by Porter and Wright (13). The process 



Triplet -f M ^ Singlet + M, 



where M is unchanged in multiplicity, is forbidden by spin selection 

 rules if M is a singlet but allowed if M has any other multiplicity. 

 Furthermore, from purely statistical considerations, the probability of 

 the process is independent of the multiplicity of M provided this is 

 greater than singlet. The observed differences between different 

 quenchers therefore arise from the varying degree of overlap of the 

 orbitals of unpaired electrons in the triplet and in M rather than 

 from the magnetic field interactions. 



Quenching by a Molecule with a Lower Triplet State 



This is an energy-transfer process in which the quenching molecule 

 becomes excited as follows: 



A,j, -\- Bs -^ A^ + Bt 



It is energetically possible whenever an encounter takes place with 

 a molecule with a lower triplet level and, since spin is conserved in 

 the overall process, it may occur with high probability. It has been 

 demonstrated in rigid media by Terenin (H), in solution for the 

 particidar case of diacetyl by Backstrom (1) , and in a range of sys- 

 tems in solution by Porter and Wilkinson (9) . The maximum 

 quenching constants observed in solution correspond to unit effi- 

 ciency at normal collision diameters, but quenching constants as low 

 as 10'' 1/mole/sec have also been measured. The present evidence 

 indicates that exchange transfer rather than Forster type resonance 

 transfer is the predominant process. 



