o90 FRED M. UBER 



sent or negligible in most cases, this is not true for high concentrations 

 of deuterium {2^, p. 220). In fact, deuterium as a tracer should not 

 be employed in concentrations higher than about 10% unless its selec- 

 tive behavior is also investigated. Heavy water may be definitely 

 toxic to organisms when its deuterium content exceeds 25%. 



4. Minimum Size of Test Samples 



The minimum quantity of gas required for a mass spectrometer 

 test sample depends somewhat on the particular instrument. Two 

 frequently stated estimates of a satisfactory volume are either 100 

 ml. of a gas at 1 mm. of mercury pressure or 0.1 ml. of gas at a pres- 

 sure of 1 atmosphere. Using either of these criteria for carbon di- 

 oxide, one arrives at a weight of approximately 0.2 mg. Expressed 

 in terms of carbon content only, something less than 0.1 mg. should 

 be sufficient for an analysis. On the same basis hydrogen would re- 

 quire approximately 0.01 mg. for a minimum test sample. In work- 

 ing with such small quantities of matter, it is usually necessary to 

 allow greater tolerances on reaction yields than would otherwise be 

 considered normal. For example, organic samples to be tested for 

 C^' or N'^ might well contain approximately 1 mg. of either carbon 

 or nitrogen previous to combustion. 



5. Limiting Dilution Factors 



The factor by which an isotope having a specified original concen- 

 tration can be diluted and still be detected reliably depends both on 

 the accuracy with which concentrations can be measured and on the 

 natural abundance of the isotope. The permissible dilution is 

 greater for isotopes whose natural abundance is smaller. Where 

 ratio determinations can be made with an accuracy of 1% and where 

 initial concentrations of the isotope are in the neighborhood of 50 

 atom per cent excess, N^* can be detected following a 10,000-fold 

 dilution. An equivalent value for S»^ is 1000-fold, for C'^ 5000-fold, 

 and for deuterium 5000-fold {2, p. 563). These dilution factors are 

 more than adequate for numerous types of investigation. They are 

 not as large, however, as can be obtained by the use of artificial radio- 

 active isotopes, which are greatly favored in this respect because their 

 natural concentration is zero. Since, with extreme care, higher ac- 

 curacy can be obtained than assumed above (/, p. 29), some improve- 

 ment over the ratio values cited is possible. Most studies to date 

 have not even approached the above dilution limits. 



