a column of finer-mesh resin and eluting with a low ammonia concentration.Quantitative Measurements. The peak areas on the recorder chart were proportional to the quantities of' solutes added, within the errors of measurement, which were about k 3 Z . Peak areas must be calibrated for each solute individually. At first sight, one would expect a given mass of triethanolamine to give a bigger peak than the same mass of di-or monoethanolamine, for the refractive indices of the three amines are, respectively, 1.4852, 1.4776, and 1.4541, compared to 1.3325 for water. However, the refractive index of a mixture depends on mole fraction rather than weight fraction, and calculation shows that on a weight basis the three amines affect the refractive index almost equally. Calibration is essential, because peak areas may be influenced by displacement of ammonia from the resin. We did not try to see how small a quantity of amine could be measured, but one can form an impression of the sensitivity from the first small peak in Figure 7. This peak corresponds to 0.1 mg of triethanolamine in the injected sample. The refractometer attenuation was 4 X, and it should be possible to detect one-tenth of this amount, namely, 0.01 mg, under favorable conditions. CONCLUSIONSThese chromatographic separations of ethanolamines and aziridines illustrate generalizations we have made before (4). The strongest binding to the metal-loaded resins occurs with primary amino groups; carbon atoms attached to the amino nitrogen or to the adjacent carbon atom obstruct the binding and cause the compounds to be eluted earlier. Thus diethanolamine is eluted before monoethanolamine, and N-ethyl and 1 -methylaziridines are eluted before ethylenimine. We must consider the coordination of the metal ions with hydroxyl groups, however, as well as their coordination with nitrogen atoms. T'riethanolamine is more strongly bound

No abstract

No abstract

iron counters and reducing the source background to 50% and the cosmic background to 0.1 of its present value (advanced model). A further precision increase can be obtained by using a 1.5-mCi source and reducing the total background to 0.1 of the present one (optimum model). CONCLUSIONThe apparatus described previously shows the following advantages: (a) possibility of nondestructive testing of liquids, solid, gaseous, and chromatogram samples; (b) no necessity of sample preparation; (c) no need for highly skilled personnel; and (d) possibility of an on-line analysis. Disadvantages in comparison to mass spectrometry are: (e) lower sensitivity; (f) lower precision; (g) lack of specificity; and (h) necessity for safety precautions, e. and f. may be corrected by the improvements of the present analyzer and g. by a more advanced concept. These will be achieved at the expense of h. ACKNOWLEDGMENTA. Schoenberg contributed basic data and measurements. A. Aladjem contributed through source development. A. Kedem advised on engineering design, and Y.Nir-El in theoretical subjects.