p. C. WILHELMSEN, R. LUMRY AND H. EYRING 



95 



tially all the hydrogen peroxide could be titrated with potassium 

 iodide and dilute sulfuric acid with ammonium molybdate as a 

 catalyst. This would also be characteristic of the compound with 

 hydrogen peroxide replacing water of crystallization. The compound 

 containing the hydrogen peroxide in this form would have the same 

 percentage composition as the one isolated. The evidence therefore 

 seems to indicate that the compound isolated by Drew is not an 

 intermediate in the reaction but a coordination complex with hydrogen 

 peroxide. 



Further information is needed concerning intermediates and end 

 products before the mechanism can be fully established. However, a 

 knowledge of the general behavior of molecules makes possible a 

 rather plausible supposition as to the mechanism for this reaction. 



Proposed Mechanism 



The DPD's probably exist in solution as a mixture of the diketo, 

 mono keto, and dienol forms: 







^\ 



NH 



OH 



I 



OH 



.NH 



II 

 O 



.NH 



X 



/ 



N 



I 



N 



O 



c 



OH 



However, in basic solution, the enol forms probably ionize. This 

 would serve to shift the equilibrium in favor of the enol forms. Stross 

 and Branch (1939) have reported a first ionization constant of 10 ~'^ 

 for luminol. They have also reported that there was no second ioniza- 

 tion constant greater than lO"^^. It seems reasonable to suppose that 

 in basic solutions such as are required for chemiluminescence, the 

 major portion of the luminol is in the form of the singly ionized enolic 

 form. In this connection it might be noted that the mono sodium salt 

 is the one normally isolated from sodium hydroxide solutions. 



As has been explained earlier, there is evidence that indicates 

 that preHminary oxidation takes place before the reaction with mo- 

 lecular oxygen. Apparently this oxidation can be carried out by any 



