RADIOCHEMICAL FRACTIONATION CHARACTERISTICS 10 These experiments have proved that important radiochemicalfractionation information can be obtained from single particles. Both the limitations and the advantages of this technique have been delineated. INTRODUCTION The 1961 Dominic test series presented the first opportunity tc study radionuclide fractionation in single particles from high~yield ait bursts. In these air bursts the condensation matrices were dependent only on the weapon composition. Masses of contaminating materials such as tower, water, and salt, which would present indeterminable factors, were not present. The radionuclide distributions in selected particles obtained from several tests in this series were analyzed radiochemically and by gamma spectrometry. Of particular interest was the distribution of volatile and refractory nuclides for each particle size. The nuclides radiochemically analyzed were Mo, ™sr, *°Sr, *°Zr., 123m Te, 1p e BIO. Mp, Mice Mice, Ing 23tyy and ?°Np. It soon be- came apparent that the radiochemical method was too time consuming, and as an alternative a nondestructive gamma-spectrometric technique. which also yielded the required data on radionuclide behavior, was employed. A computer technique was developed for resolving the spectra. and 334 of the particles were examined by this method. Analysis of the radionuclide concentrations in individual particles so far has been confined to the determination of correlations between the data and the particle diameters. PROCEDURE The radioactive particles were located by autoradiography andisolated from the filter-paper matrix. After particle size and other physical parameters were measured, the particles were subjectedto sequen- tial radiochemical analysis (see Fig. 1). Particles ranged in size from 4 to 20 pand contained between 10° and 10° fissions based on *?Mo analysis. Aggregates of particles representing groups 4to5 yu and 4.5 to 9 were Similarly analyzed to determine radionuclides that were in too low abundance to be measured in single particles. A complete study was made on integral papers. The nuclides were assayedby standard counting techniques, and the results were converted to R* values for fission products or atoms for induced radionuclides. *An R value is defined as the ratio of the ratio (r) of calculated fissions in isotope (i) to fissions as calculated from °*Mo to a similar ratio (r) determined from thermal-neutron fission of 95U; i.e., (ry,99)X/(1j,99) th.