A Sensitive Analytical Method for the Determination of Very Low Level Plutonium in Humans (Geneva Conference Paper 1958 PP 759)

Page 1

A SENSITIVE ANALYTICAL METHOD FOR THE DETERMINATION OF VERY LOW LEVEL PLUTONIUM IN HUMANS General Electric Company, Richland, Washington. Hanford Laboratories Operation, Hanford Atomic Products Operation, THEearly detection of possible plutonium deposition in individuals working with this element is of prime importance in providing radiation protection for these people within established limits. Analysis of urine samples from exposed individuals reflects the body burdcn of plutonium on the assumption that the excretion rate is a function of the quantity of dcposited plutonium. Because of the low maximum permissiblc body burden of plutonium, 0-04 pc (0-6 and the attendant very low excretion rate, it is essential that sensitive methods be available to detect the element in urine. For an acute plutonium exposure to soluble plutonium compounds resulting in deposition of the maximum permissible body burden, only 6 disintegrations per min ( 4 . 2 ~10-5 pg) per day will be excreted in the urine 100 days after the exposure(2).For incipient chronic intake which eventually builds up to a maximum permissiblc body burden, even smaller quantities of plutonium are excreted per day because of the nearly complete immobilization of plutonium which occurs with time. Since i u s important to dctect a small fraction of the permissible body burden in,qjer to permit timely corrective action, the urine analpiis for plutonium sfiould be capable of measuring much less than 1 dis/min (disintegrations per minute) in a sample. The objcctive of the work reported was to develop an adequately sensitive and reproducible method for plutonium in urine. The routine procedure used at Hanford prior to this development required evaporating the urine sample to dryness, muffling, and dissolving remaining salts. Any plutonium present was then co-precipitated with lanthanum as the fluoride. After solution of the lanthanum precipitate, thenoyltrifluoroacetone ('ITA) was used to chelate the plutonium now in the +IV oxidation state. The plutonium was subsequently released from the organic phase to the aqueous phase with 8N HNO3, and the resulting acid evaporated to near dryness. Evaporation was continued to dryness after transfer of the acid to a 14-in. counting dish. Low-background alpha counters were used to record the alpha particles entering the sensitive volume of the counter. The air chambers used were sensitive to noise and other minor disturbances and much maintenance and Geneva Conference Paper ( I 928) Pi759.

Page 2

Analytical Method for the Determination of Very Low-Level Plutonium in Humans 327 operator time was expended in keeping the sets in control with low backgrounds. A measurement time of up to 2 hr was needed to give a detection limit of about Nuclear track emulsions were considered to replace the electronic counters as the alpha-particle detector, since several of the disadvantages of the then currently used electronic detectors could be eliminated or minimized. The analysis as visualized would separate the plutonium from the sample and deposit it unifornily onto a small disk. This disk would be held against the nuclear track emulsion for a period long enough to produce a significant number of alpha tracks upon development. These would be counted by viewing the exposed arca under a microscope. The quantity of plutonium present could then be calciilated froiii the emulsion efficiency, area examined, area of deposit, and exposure time. It was first necessary to develop a method for depositing the plutonium coiitaincd in tlic urinc sample uniformly onto a small area. An electrodeposition procedure was dcvclopcd i n which plutonium was first oxidized in a basic medium, thcn elcctrodeposited on a 7 mm diameter area in the center of a 4 in. dianietcr stainless steel disk(3).In 1-2N KOH and in the presence of sodium hypochlorite it was found that nearly quantitative recovery of pc of plutonium was achicvcd. Thc snniplc preparation, lanthanum fluoride and thenoyltrifluoroacctonc method for prior scparation of plutonium from urine was retained. Hydrochloric acid was substituted for 8 N HNO3 in the extraction, when it was found that nitratc ion caused low electrodeposition yields under the conditions cmployed. An clcctrodcposition ccll was dcveloped and multi-position equipment designcd to pcrmit simultaneous analysis of as many as twcnty samplcs. Figure 1 shows thc lucite clcctrodcpositioii cells and the twenty-sample electroIn the app li diy of thc procedure, urine samples "spiked" with a known quantity of plutonium and blank samples are processed with cach group of actual saniplcs in order to provide data on the control of thc process. Detailed procedures Tor thc chcmical processing of thcse samples cannot bc prcscnted in this report, due to space limitations, but will be made available(4). NUCLEAR TRACK FILM DETECTION The nuclear track emulsion chosen to record the alpha particles cmittcd from the 7 mm dianicter area was Kodak NTA emulsion 25 p thick on 1 x 3 i n . microslicks. A study was undertaken to determine optimum condiriotis for eradication of background latent tracks, exposure, proccssing, and rcading the exposed slides. As received, the emulsion contains latent track images from alpha-cmlttiii: impurities. The background tracks from these sources must bc extrcmcly few to provide maximum sensitivity. Background tracks in the cn~~ilsionWCI'C effectively eradicated before development by exposing the ernu l \ i i ) i i at room

Page 3

for 3 4 hr restored the original sensitivity of the film. A reduction of background to less than 10% of the original track density was thus easily accomplished. Backgrounds are commonIy experienced of 0-002-0.004 dislrninlcm2, which is The latent image of a bona-fide track in an emulsion will be destroyed over a period of time by tlie same processes which are deliberately accelerated in background eradication. High humidity and warmth are particularly deleterious to retention of the latent tllack image. It was found that imperceptible fading of latent images occurred in 8 weeks after exposure if the exposed emulsion was held at 5" C during the period following exposure. Optimum developing time was determined to be six minutes in D-19 developer at 68" F&2". Developing proceeds with no agitation. Fixing requires 45 inin with agitation, and was1iing;one hour. Slides are dried in a dust-free A necessary part of the development was a device to support the $in. diameter disks against the emulsion. A "radioautographic camera" described in Fig. 2 was constructed. All cameras used are identical and each permits the simultaneous exposure of eight disks. The eradicated NTA slidc is inscrtcd emulsion-sidc up i n t o tlie milled recess of the brass base block. The thumb clamp draws thc slidc against three edge-index bosses corresponding to thc cdge contact points of the mcclunical stage of the microscope used. This three-point indexing permits thc co-ordinates of the disk centerlines to be establishcd reproducibly with rcspcct to the stage vernier scale, even when some variation occurs at the cdges of thc cmulsion slides. The lucite block is placed over the slide, indexing to dowels in the base block. The 5 in. disks are placed in the holcs, platcd sid; against thc cniulsion,

Page 5

Analytical Method for the Determination of Very Low-Level PIutonium in Hunlans 329 and the cover place located. Printed forms are used to assist in the identification and placing of the source disks. The NTA slides are handled with as little exposure to safelight (Wratten Series OA) as possible to keep the single grain background to a minimum. After the disks have been exposed at 5" C to the emulsion for the exposure time, generally 168 hr. the slides are removed and processed. A microscopic count of tracks is then made in the exposed areas of the emulsion. Dark field illumination and a 43 x objective and 10 x eyepiece are used giving a total magnification of 430. Immersion oil is used between the condenser lens and A scanning system of counting is employed. A rectangular reticule in the eyepiece delineates a 0.10 mmx0-2 mm area of the emulsion. From the coordinates of the area to be examined the lateral limits of the scan (4 mm long) are determined. Two stops are then positioned on the right-left traverse bar of the mechanical stage. These stops limit the motion of the stage to the length of the scan. A rachet-stop attachment to the knob for forward-back motion permits the stage to be moved in 4 mm increments perpendicular to the lateral motion. These devices permit several traverses to be made on each exposed area with the total area scanned in the standard procedure, about 4 mm2. The number of tracks per unit area is thus determined and a conversion to disintegrations per min made using exposure time, exposure area, and film efficiency. The film eficiency is routinely checked by exposing a standard source of known disintegration rate for a known period of time. The frequency distribution of the number of tracks found in 3.8 mm2 of emulsion for 545 blank urine samples analyzed by this procedure is shown in Fig. 3. The frequency distribution calculated from the Poisson function is plotted adjacent to the observed frequency. Data accumulated subsequently have shown ever increasing agreement with the Poisson distribution. It is assumed for *poses of defining a detection limit that samples containing a significant q u w of plutonium will also be distributed to conform to the Poisson function. The detection limit is then defined as that quantity which, if actually present, would give a positive result above the background distribution in 99% of the samples analyzed. These data indicate that 0-033 dis/min actually present in the sample would result in a measured dis/min greater than the 99% upper limit in the background distribution 99% of the time. It should be noted that values greater than the 99% upper limit of the background distribution but lower than the 0-033 dis/min indicate positive values and should be treated as such. Figure 4 illustrates the result of the analysis of 553 samples each "spiked" with 0.57 dis/min of plutonium as determined by repeated calibrations on a PC/2A proportional counter. These samples were analyzed concurrently with the regular bioassay urine samples. The standard deviation of the 553 results shown is 12%. If only random variations from reading the tracks were present, a standard deviation of 7% should be demonstrated. The remaining variation results from all other random errors in the process. The long-time avxa2c of

Page 6

BLANK URINE SAMPLES A MEAN OF 13 TRACKS FIG.4. Frequency distribution of 553 urine samples spiked with the overall process yield is 89%, assuming a 50% emulsion geometry. Emulsion geometry will, in general, be slightly less than 50%, principally because of the inability of the microscopist to record tracks perpendicular to the plane of the The particular advantage of this method is that there is virtually no limit to its ultimate sensitivity since by a combination of reducing source size and increasing the exposure time almost any desired sensitivity should be obtained.

Page 7

Analytical Method for the Determination of Very Low-Level Plutonium in Humans 33 1 Experiments have shown that electrodeposition on 1 and 2 mm diameter areas is feasible. Another advantage is the complete freedom from the somewhat unpredictable beliaviour of low background electronic counters, and the attendant high initial cost and subsequent maintenance costs. For incidents in which plutonium present in urine must be known within a very short time of the incident the method described with the standard exposure time would require too long an interval. However, in these cases the plutonium will appear in the urine in much higher concentrations because of the short delay between exposure and sample collection. For these cases the usual alphacounting after a chemical separation is generally adequate. In a routine bioassay program designed to discover incipient plutonium exposures and identify such individuals who may require work with less chronic exposure liability, this method has proved reliable, sensitive,‘and adequately precise. In this application the lag time required bctween sampling and recording of results is of I . Maximum permissible amounts of radioisotopes in Ihe human body and maximunl permissible concentrations in air and water. Nutiotzal Burcart of Staizdards Handbook 52 . of Pcrsorz)icI, presented at the mceting of Medical and Laboratory Directors of the Atomic Encrgy Conmission (28 Septemtcr 1950). Ei?iirlsiorzs10 lie Aiialysis of Uriiiefor Vcry Low L c v c l Piiirorziuiii, H W-22680 ( I 5 November in urine, submitted to Nrrclconics.