EMIL H. W HUE 187 



been exaniiiied in the j^ast almost exrlnsively in a system composed 

 ot" water, sodium or j)otassium hydroxide, hydrogen peroxide, the 

 hydrazide, and an 'activating" agent (2) . The hitter reagent usually 

 consists of either an oxidizing agent such as sodium hypochlorite or 

 potassimn terricyanide, or a transition metal. A chelating agent is 

 used with the metal to prevent the precipitation of the metal oxide 

 or hydroxide in the basic solutions which are required for chemi- 

 luminescence. Typical metal "activators" are iron in the form of 

 hemin (2) or the jjolyamine or Versene complexes (30) , and copper 

 in the form of its annnonia complex (30) . The role of the metal 

 is a double one; in its reaction with hydrogen peroxide, two necessary 

 reactants, oxygen gas and hydroxy or hydroperoxy radicals are formed 

 (16) . In a few instances hydrogen-peroxide-free systems have been 

 used, but iHider these conditions oxygen must be supplied and an 

 oxidizing agent such as potassium ferricyanide must be used to get 

 any appreciable amount of light. 



The quantum yield has been measured for the luminol reaction 

 inider a number of conditions with various combinations of reagents; 

 a convenient summary of the results has been given by Bremer (6a). 

 To date, the highest quantum yield measured for a water system has 

 been found to be 0.003. 



Mechanism of tJie Reaction. Very few facts are available on which 

 to base a mechanism. No product other than nitrogen (26) has been 

 isolated from a chemiluminescent reaction of luminol in water. 

 The chemiluminescent reaction of the unsubstituted phthalic hy- 

 drazide has yielded phthalic acid (8, 32) but this is largely a product 

 of the "dark" reaction. With a quantum yield of 0.003 for the 

 chemiluminescence in water of the most efficient hydrazide, the possi- 

 bility of isolating the products of the "light" reaction is, at best, slim. 



In view of the nitrogen formed in the reaction and in view of the 

 nature of the reaction in organic solvents (vide infra) , it is rather 

 certain that luminol is not a "catalyst" in the chemiluminescence; 

 it is consumed in the reactions leading to the excited state of some 

 emitter. The belief that luminol is not consumed in the chemi- 

 luminescence stems from the work of two investigators in particular. 

 Drew showed that in a sodium hypochlorite— hydrogen peroxide sys- 

 tem, ca. 70 moles of hydrogen peroxide were consumed per mole of 

 luminol (8) . Actually, large amounts of oxygen are evolved, and 

 the excess of hydrogen peroxide is merely lost in this side reaction 

 which has nothing to do with luminol. Oxygen evolution from the 

 hydrogen-peroxide-based systems occurs whenever a strong oxidizing 



