p. C. WILHELMSEN, R. LUMRY AND H. EYRING 81 



very small quantities of oxygen greatly affect the efficiency of the 

 chemiluminescence. 



A difference in the rates of the reaction with oxygen in the presence 

 and absence of an additional oxidant might be inferred from the 

 difference in quantities of oxygen required to produce a visible light 

 in each case. However, a better idea of the relative rates might be 

 obtained from a study of the length of times required for the reactions 

 to proceed to a certain fraction of completion. The reaction involving 

 only dissolved oxygen requires weeks or months to proceed even a 

 small fraction of the way to completion. The reaction where an addi- 

 tional oxidant such as ferricyanide is present is essentially complete 

 in a matter of minutes. Figure 2 shows a typical plot of intensity 

 versus time for a solution that was 0.05M luminol, 0.005A/ ferricya- 

 nide and l.ON NaOH. 



The kinetic study of the reaction between the DPD's and oxygen 

 would at best be very time consuming if the reaction were carried to 

 completion. However, as was stated earlier, the reaction rate can be 

 related to the intensity of the chemiluminescence if the temperature 

 dependence of (f> is known. An experimental determination of this 

 temperature dependence will now be discussed. 



Experimental Temperature Dependence of </> 



The reaction between ferricyanide and luminol in the presence of 

 oxygen was chosen for a study of the temperature dependence of </> 

 because of its relative freedom from complications and tlie ideal speed 

 of the reaction for kinetic studies. 



Measurements were made by using the equipment diagrammed in 

 Fig. 3. The solution of luminol and oxygen was mixed with the ferri- 

 cyanide solution by means of a thermostated mixer, 5, similar to the 

 one employed by Chance (1940). The solutions were thermostated 

 and stored under pressure in flasks 2 and 3, and when the double 

 stopcock 4 was opened, the solutions rapidly mixed and flowed along 

 a 1-mm capillary. The solutions were allowed to flow until the light 

 emitted reached a steady value. The flow was then stopped and the 

 intensit>' of the light from a small quantity of solution in a short 

 section of the capillary was studied. The Hght was measured by a 

 photomultiplier, 7, which was in turn connected with an oscilloscope, 



