PLUTONIUM-BEARING PARTICLES FROM FUEL REPROCESSING 137 



counting the tracks, a small portion of the film containing the particle was isolated, the 

 emulsion removed, the polycarbonate dissolved, the track replicas oxidized, and the 

 elemental composition of the ^^^Pu-bearing particle determined by electron-microprobe 

 analysis. These data were compared with data from natural aerosol particles. 



Most of the collected particles were composed of aggregates of crustal materials. Of 

 the particles, 3.6% was organic and 1.7% was metallic, viz., iron, chromium, and nickel. 

 High enrichment factors for titanium, manganese, chromium, nickel, zinc, and copper 

 were evidence of the anthropogenic nature of some of the particles. The amount of 

 plutonium in most particles was very small (less than 1 fCi of ^^^Pu). Thus plutonium 

 concentrations had to be determined by the fission-track counting method. Only one 

 particle contained sufficient plutonium for detection by electron-microprobe analysis. 

 This was a l-jum-diameter particle containing 73% PUO2 by weight (estimated to be 170 

 fCi of ^^^Pu) in combination with FejOa and mica. The plutonium-bearing particles 

 were generally larger than natural aerosols. The geometric mean diameter of those 

 collected from the mechanical line exhaust was larger than that of particles collected 

 from the wet-cabinet exhaust (12.3 /jm vs. 4.6 /im). Particles from the mechanical hne 

 also contained more plutonium per particle than those from the wet cabinets. The 

 amount of plutonium per particle decreased with the distance of each sampling point 

 from the mechanical Une. 



The size and ^^^Pu content distribution among particles collected from the sand 

 filter effluent and at the 50-ft level of the 291-F stack were almost the same. The 

 geometric mean and standard deviation of the diameter of ^^^Pu-bearing parricles at the 

 50-ft level was 5.43 ± 2.69 jum. The relatively large size of these particles is believed to be 

 due to coagulation of submicrometer particles by thermal and turbulent mechanisms to 

 fomi larger agglomerates. The elemental composition of these particles, which contain 

 very small amounts of plutonium in combination with crustal elements not used in the 

 recovery process, supports this assumption. Scanning electron micrographs, such as 

 Fig. 5(c), also show these particles to be agglomerates of smaller dissimilar particles. 



Fleischer and Raabe (1977) have observed alpha-decay-induced fragmentation of 

 ^^^Pu02 particles probably caused by the heavy recoiling nuclei. When suspended in 

 water, these particles produce fragments, or subparticles, which contain from 50 to 

 10,000 ^^^Pu atoms, the abundance of which follows a power-law relation v^th the 

 largest particles being the least abundant. The possibiHty exists that PUO2 particles, large 

 enough to be trapped on HEPA filters, fragment owing to alpha decay. The small 

 fragments then pass through the filters where they coagulate with dust composed of 

 crustal elements. The larger dust particles may not have passed through the HEPA filters 

 but entered the exhaust system through leaks in the ducts, as illustrated in Fig. 10. Such 

 leaks might remain undetected as long as the exhaust system remained under negative 

 pressure with respect to the atmosphere. 



The geometric mean and standard deviation of the number of fission-fragment tracks 

 per ^^^Pu-bearing particle collected from the 50-ft level during July, August, and 

 September 1977 was 17.01 ± 1.65 tracks. One femtocurie of ^^^Pu in a mixture of 

 low-irradiation plutonium will produce 52.6 fission fragments when irradiated with a 

 fluence of 8.64 x 10^ '^ thermal neutrons/cm^. Only about half, or 26.3, of the fragments 

 will produce tracks in the polycarbonate film. Thus the calculated geometric mean 

 radioactivity on the ^^^Pu-bearing parricles leaving the stack is 0.65 fCi/parricle. During 

 these 3 months a total of 82 juCi of ^^^Pu was discharged to the environment. This 



