diameter. Tamura (1974, 1975, 1976a) also studied soil samples from safety shot areas in NTS. These studies of particle size association included separation of different silt and clay sizes and showed that the association was primarily with che coarse silt (50 - 20 um) and fine sand (125 - 50 um) sizes. have been extracted. Since the citrate does not reduce the iron and manganese in the samples under the extraction conditions, the citrate-extractable plutonium may not be bound with those elements, lLron and manganese form surface oxide coatings on particles; thus the dithionite treatment coupled with citrate would serve to remove the surface coatings and possibly remove surface-sorbed plutonium as well. The particle size segregation of Pu-contaminated soils by Tamura (1974) was made without benefit of chemical or physical treatment to induce particle dispersion. This approach was taken in order to evaluate the degree of assoclation of plutonium as it exists in the soil. Later data were reported on size association of the plutonium with and without ultrasonic treatment for dispersion (Tamura, 1976b). In addition to comparing the effect of ultrasonic treatment on two NTS soils, data was included on a soil from the floodplain in Oak Ridge (ORNL) and a segment of bottom sediment from the canal at Mound Laboratory (ML). The same data are reproduced in Tables 2 and 3. Bondietti et al. (1976) evaluated the amount of organically associated plutonium in a floodplain soil in Oak Ridge. They used a resin technique and estimated that about 15% of the plutonium in the soil may have been associated with soil organic material. Sodium-resin, prebuffered at the soil pH, was used to maintain constant pH and decalcify the soil organic matter. This is in contrast to sodium hydroxide extraction of organic matter which raises the pH. Destruction of organic matter with sodium hypochlorite released 82% of the plutonium. They concluded that since the alkaline hypochlorite treatment would not dissolve the iron oxides, most of the plutonium was surface sorbed rather than intimately assoctated with iron. Since the fioodplain soil and the Lake Michigan sediments represent different matrices, the mode of association could be different as postulated by these investigators. It would be informative if the different treatments employed by the investigators could be applied in a similar manner to both samples so that deductions regarding the association could be evaluated. The ultrasonic treatment increased the plutonium contribution of the finer size fractions of the total soil (Table 3). In general, the increase in the plutonium contribution in the finer resuspendible sizes was also accompanied by an increase in mass contribution in these finer sizes (Table 2). This suggests that the plutonium contribution in the finer sizes was primarily due to plutonium-soil particle disaggregation and not necessarily to plutonium detachment from the soil particle. Considering the less than 5 um sizes, the ultrasonic treatment increased the percentage contribution by two to five times, although the total percentage contribution in the NTS samples was still less than 10%. In the sample from Mound Laboratory, the ultrasonic treatment increased the Pu in the less than 5 ym size from 34% to 81%. More significantly, The high Kd's of over 100,000 assoctated with aged (years) sources described above, as well as with the groundwater particulates at Enewetak Atoll and the sediment from the canal near Mound Laboratory, compared to the lower Kd of about 1,000 for shorter term (days) simulated tests in the laboratory suggest a biogeochemical process which increases retention. The Process of increasing retention has implications in reduction of uptake by plants as well as in the resuspension process, both in terms of the size of the particles being resuspended and the depositional patrern in the lungs. the Pu in the less than 2 pm size fraction in Mound Laboratory sample increased from 17 to 70%. The importance of the association of the plutonium with the fine size of soil particles will be discussed further in the later section of this paper. Muller and Sprugel (1977) determined the distribution of plutonium in three soils located within 50 miles of Mound Laboratory in Ohio. Two of the sofls In discussing plutontum particle size, two tmmediate considerat ions become apparent. First, when Pu releases occur from stacks, the particles to be considered are pure metal or metal oxides. When these particles deposit on the ground surfare, interactions with the soil particles occur. Depending on the character of the initial source, the intensity of the association with soil particles may be different. This paper is concerned with the plutonium in soils and sediments, hence it does not consider the initially released particles. That plutonium attaches itself to host soil particles has been verified directly by microscopic and alpha-track measurement techniques (Hayden, 1976; Nathans and Holland, 1971: Nathans et a7., 1976). Mork (1970) studied the size assoc tation of plutonium in Yucca Flat on the Nevada Test Site (NTS) and showed that most of the plutonium was aasoctated with soil particles greater than 44 um 100 -: were “background samples reflecting fallout levels; the third was a soil PLUTONIUM SIZE CHARACTERISTICS sample taken one mile east of the Laboratory, containing sufficient ??3py originating from the operations to differentiate it from fallout levels. The samples were treated with a carbonate-oxalate dispersant and the fractions were analyzed for plutonium. Between 60-75% of the plutonium was assoc {lated with particles less than 4 pm size, and about 907 of the activity was assoctated with particles less than 45 um (silt and clay}. For comparison, a bottom sediment sample taken in the canal at Mound Laboratory showed very similar distribution with 81% of the Pu in the less than 5 um and 100% in the combined silt and clay fraction (Tamura, 1976b). Plutonium associations with different particle sizes in soiis of Rocky Flats have been reported by Little et al. (1973}. The mode of segregation involved sieves with only mechanical shaking of oven-dried samples. The particle size distribution showed that the less than 104 um size particles in the soil contributed 3.74 in one sample and 11.2% another. The 3.7% and 11.2% weight fractions reported by Little et al. (1973) contributed 11 and 377, respectively, to the total soil activity. The low amount of fine materials likely caused by the oven drying which would tend to "cement" particles, the short shaking 101