248 M. E. WiNFIELD 



transfer. The well-known difficulty of adding an electron to catalase without 

 the aid of a donor-type complexing agent (e.g. azide) tends to confirm that 

 catalase is not amenable to direct electron transfer, and that it is not until the 

 — OH group in the sixth co-ordination position is replaced by a 'promoter' 

 ligand that the passivity can be overcome. 



If we are to persist with our thesis that in Cat. HgOg I the sixth ligand is 

 not H2O2, we have to make a choice between H0~ and 0=, the only two 

 alternatives. Of these the oxygen ion is by far the more likely to stabilize a 

 higher formal oxidation state of iron. Amongst the iron compounds whose 

 structure is known, the states written formally as Fe^^ and Fe^^ are encountered 

 only when the metal is associated with the 0= ion. There can be little doubt 

 that the actual charge on the metal atom is small, and that this is made 

 possible by the contribution by 0= of more than two electrons to the bond 

 with iron. In those complexes in which only one oxygen ion is attached to 

 the metal, e.g. in (V^^0)++, it seems necessary to admit the presence of a 

 complete tt bond, rather than one of fractional order (i.e. resonating) as in 

 (Fe^^04)++. We are inclined to the view that in Cat. HoOa I there is one 

 0= ligand and that the 'co-ordinative' -n bond from oxygen to metal is non- 

 resonating. When we write Cat. H2O2 I as Fe^ :^ O or Fe^ : :0 we mean a 

 structure indistinguishable from that obtained by combining Fe^^^^ with the 

 oxygen atom : O to give Fe^^^ :^ O or Fe^^^^ : O, in which oxygen donates two 



electrons to form the a bond to Fe while the latter donates two dg electrons 

 to form a dative 77 bond to oxygen. 



None of the known inorganic complexes of iron contain Fe^, perhaps be- 

 cause the ions which we write formally as Fe^^ and Fe^^ can have one more tt 

 bond. In our previous paper (King and Winfield, 1959a) we therefore hesitated 

 to propose the presence of Fe^ in Cat. HgOg I. The alternatives of writing 



•C Fe^^, C Fe^^ (where P~ is the protein) and C Fe^ — O are rendered 



I I I ^1 1 I I 



P- 

 less likely by the results of recent experiments by Gibson et al. (1958) 

 on Mb^" oxidation, as well as by the lack of direct 1 -electron oxidation pro- 

 ducts of catalase. We are therefore of the opinion that Cat. H2O2 I is 



\ / 

 better written as C Fe^ = 0. 



I ^1 



The reason for the low extinction coefficient of the Cat. H2O2 1 Soret band 

 is obscure, but seems to be related to the presence of the 0== ion. A large 

 decline in the Soret band of Mb^" when oxidized to Mb^^ was noted by 

 Keilin and Hartree (1951). It is presumably the same Mb^^ complex which 

 has been shown by Gibson et al. (1958) to be lacking in free radical character, 



