104 Discussion 



Griffith estimated the ionization potential for O2 in structure A to be 20-9 eV, but 

 that in structure B to be 16-6 eV. Consequently he concluded that B is the energetically 

 favoured structure (see Froc. Roy. Soc. A235, 23, 1956). 



However, since part of the bonding electrons in both structures A and B are from 

 the Fe^^, this problem may be treated by the usual molecular-orbital method. We 

 may recall that in forming the molecular-orbital CjV'i + c^y'z from the atomic orbitals 

 Vi and v'2. with energies Ei and E2 respectively, the energy E of the bonding molecular- 

 orbital is 



^~^^ E^-E ' 



where E < E^ < E^, ^ = iy)j*Htp2dT, S = Wi*^'zdr. 



In order to have effective bonding, £"2 — E^ must be small, /3 and S must be large, 

 and the overlap of this bonding molecular-orbital with occupied non-bonding 

 molecular-orbitals must also be small. Prediction of the relative stabilities of structures 

 A and B by considering E^ — E^ alone and neglecting all the other factors is hazardous. 

 The fact that O3 has an obtuse rather than acute isosceles-triangular structure indicates 

 that these other factors may indeed be more influential in determining the relative 

 stabihties. (See structures C and D.) 



•o o 



I . II 



:0: .0. 



(*f) 



C D 



For this reason one should not overlook the possibility that A may actually be the 

 energetically more favoured structure, and that consequently structure B can exist 

 only when A is impossible without drastic rearrangements in the rest of the molecule, 

 such as in the case of ethylene-metal complexes. 

 Orgel: This may be a good moment to draw attention to the work of Elvidge and Lever 

 (Proc. chem. Soc. 195, 1959) who have shown that manganous phthalocyanine is a 

 reversible oxygen carrier in pyridine. The relevant reaction is probably 



Py Py 



2 /Mn"/ -I- O2 ^ 2/ Mn^ 



O 



The Mniv compound is strictly analogous to the ferryl derivatives suggested by 

 George to be important in various haem reactions. This system could be thought of 

 as a model for oxidase action in which an oxygen molecule is dissociated and the 

 atoms become attached to two different iron atoms — of course there is no evidence 

 for this, but it may be worth considering along with other models. 



