Table 11. Activities of 1mrnAgand llornAgFound in Livers of 1968 and 1970 Albacore by Cyanide Extraction Date pCi/sample* pCi/kgc Collected Countedd Preserved Wet wt”, kg loemAg 1 l0rnAg l08rnAg 1IOrnAg 10 Sept 68 25 Aug 70
22 Apr 71 Formalin 1.670 15 May 71 Frozen 6.050 a Original wet wt. =k one counting standard deviation. 108mAgratio = 1.1.
6 . 0 2~ 0 . 2 8.7 f 3.6 4.3 & O.lc 4 . 8 i O.lc 11OmAg/ Sample activity/kg on date of collection. d 4000 min counts. 9.9 f 0.2 26.0 i 0 . 2
0.76 i 0 . 3 1 14.7 f 0 . 2
concentrations might be followed for many years to come in fish livers and other marine organisms. During the last few years, losmAg/llornAg ratios have been measured in a large number of marine fish and squid. The interpretations of these ratios from an oceanographic point of view, in which reference is made to several possible sources of silver nuclides from nuclear tests, are discussed elsewhere (1,5,6). Experiments are now under way in this laboratory in the hope of removing other gammaemitting nuclides from wet tissues by similar methods.
65Zn. A 100-gram sample of the liver tissue was counted in the wet state for 2500 min and the silver was then extracted using the cyanide procedure. Entry number two in Table I 3hows that both silver nuclides were quantitatively removed in this case also. It may also be seen that the activity ratio remained unchanged during the procedure; this suggests that n o contamination occurred. Because of the success obtained in leaching radiosilver from wet tissue, the problem of measuring losrnAgand llomAgin the 1968 and 1970 albacore liver samples was then undertaken. The cyanide leaching process was applied t o 1.6 kilograms of livers preserved in formaldehyde since 1968 and six kilograms of frozen 1970 livers. Results are found in Table 11. To ensure that all of the radiosilver had been removed, a second batch of silver carrier was added to the liver slurries, equilibrated, and plated out. The AgCl precipitates from these “second extractions” were devoid of measurable amounts of either silver isotope. Figure 1 shows the variation of 65Zn, W o , 54Mn,llornAg, and losrnAgin livers of a specific population of albacore tuna caught in the summers off San Diego from 1964-1970 ( 4 ) . Without the cyanide extraction procedure, only upper limits of the 1968 and 1970 silver concentrations might have been reported. Thus the cyanide extraction procedure appears t o provide a convenient means by which 1osmAg and llomAg
RECEIVED for review June 21, 1971. Accepted August 25, 1971. This work was conducted under the financial support of the U.S. Atomic Energy Commission, Contract No. AT(04-3)-34, P.A. 71-15, and the U.S. Office of Naval Research Contract No. U S N N00014-69-A-0200-6011.
(4) T. R. Folsom, D. R. Young, V. F. Hodge, and R. Grismore, Paper presented at the Third National Symposium on Radioecology, Oak Ridge, Tenn., May 1&12, 1971.
(5) T. R. Folsom and D. R. Young, Nature, 206, 803 (1965). (6) R. Grismore, T. R. Folsom, V. F. Hodge, and D. R. Young, unpublished work, Scripps Institution of Oceanography, 1971.
ACKNOWLEDGMENT
The authors thank T. Otsu of the National Marine Fisheries Service, Honolulu, for supplying the Hawaiian samples ; D. R. Young, Southern California Coastal Water Research Project, Los Angeles, for the 1968 albacore samples; and P. Butram of the Westgate Cannery, San Diego, for the 1970 albacore samples.
Determination of Traces of Uranium by Radioisotope Energy Dispersive X-Ray (EDX) Analysis C. C. Bertrand and
T.A. Linn, Jr.
Metal Mining Diuision-Research Department, Kennecott Copper Corporation, Salt Lake City, Utah 841 I I
DIRECT DETERMINATION of traces of uranium in aqueous solutions by fluorimetric or spectrophotometric techniques is often precluded by interferences from diverse cations, especially iron (111) ( I , 2). Chemical separations, such as solvent extraction or ion exchange, are necessary to remove interferences and concentrate the uranium to meet sensitivity requirements (3-5). Our purpose required a more rapid (1) J. E. Currah and F. E. Beamish, ANAL.CHEM., 19, 609 (1947). (2) C. W. Sill and H. E. Peterson, ibid., p 646. (3) G. Alberti and A . Saini, Anal. Cliim. Acfa, 28, 536 (1963). (4) F. G. Sherif and A. M. Awad, ibid., 26, 235 (1962). ( 5 ) W. J. Maeck, G. L. Booman: M. C. Elliott, and J. E. Rein. ANAL.CHEM., 31, 1130 (1959).
and simple procedure for the determination of traces of uranium in a variety of aqueous samples. In addition, a versatile method was needed, which could be easily adapted for use at remote field laboratory sites. A rapid precipitation for the collection of uranium for radiochemical assay has been reported, which offers quantitative recovery of uranium (6). The separation utilizes barium sulfate as a collector for uranium (IV), and certain other 111- and IV-valent cations, in the presence of potassium (7, 8). The incorporation of radioisotope E D X analysis for (6) C. W. Sill and R. L. Williams, ANAL.CHEM., 41, 1624 (1969). (7) C. W. Sill and C. P. Willis, ibid., 36, 622 (1964). (8) Ibid., 38, 97 (1966).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 2, FEBRUARY 1972
383
the measurement of the separated uranium satisfies the other requirements mentioned above. EXPERIMENTAL
Apparatus. Micro-filtration apparatus for 24-mm diameter filters, with suction flask, glass frit, filter chimney, and clamp (Millipore Corp.) is used with 24-mm diameter glassfiber filters (Reeve Angel). Whatman No. 7 filter paper squares (25 x 25-mm) are used for filter backing. Aluminum sample cups (33-mm diameter X 5 mm deep) provide the means for mounting filters using spray enamel to cement the filter onto the back of the cup. X-ray fluorescence measurements are made using an EDX analyzer consisting of Si(Li) detector and preamplifier (Nuclear Equipment Corporation) ; linear amplifier, single-channel analyzer, timer, and display scaler (Canberra Industries, NIM modules). A lo9Cdradioisotope source (Amersham/Searle, 10 mCi), mounted in an aluminum sample holder attached to the detector housing, provides the primary X-ray excitation (9). Reagents. All reagents for the chemical separation are prepared from AR grade chemicals. A saturated solution of KzS04,0.45% solution of BaCl,, and 0.5% solution of HzSOaare made up in distilled water. Concentrated HC1 and 2 0 z TiC13 are also required for the precipitation. A stock solution of 1.000 gram of U per liter is prepared from UaOs. Uranium standard solutions are made by diluting aliquots of the stock solution with 0.5 % HzS04. Procedure. Pipet a suitable sample aliquot (containing between 5 and 100 pg of uranium) into a 150-ml beaker. If necessary, dilute to 25 ml with distilled water. Add 10-15 ml of concentrated HC1 and heat to boiling. Ten milliliters of saturated KaS04solution are added, and three or four drops of 20 TiC13 solution-until the permanent faint violet color of Ti (111) persists in the sample. While swirling the beaker, add 1 ml of 0.45% BaCla solution dropwise t o the boiling sample. Continue boiling for 1 minute; then add a second 1-ml aliquot of BaCl? solution. After 5 minutes boiling, cool the sample in cold running water. Filter the precipitate onto a 24-mm diameter glass-fiber filter (Reeve Angel 934AH) using suction and a Whatman No. 7 paper pad. Wash the precipitate and filter carefully with small portions of 0.5% H2S04solution. Dry the filter by suction on the filter unit, and prepare the back of an aluminum sample cup with spray cement for mounting the filter. Transfer the filter to the back of the A1 cup, and let dry for 10 minutes. A blank of BaS04 precipitate and uranium standards containing from 5 to 100 pg of uranium are prepared as described above. The 22-keV silver X-rays from the 109Cd radioisotope source are utilized for excitation of the uranium L-series fluorescence. The single-channel analyzer is adjusted to discriminate against all pulses except those from the 13.6-keV uranium L-alpha photopeak. One-hundred-second integrated counts are recorded. A standard curve is computed by a least-squares fit of the uranium standards data using a desk-top computer. Sample concentrations are computed directly from the sample counting data. Conventional X-ray spectrometry may be used for the determination of the separated uranium. One-hundredsecond counts a t 26.14 degrees 20 (LiF analyzer crystal) using either a tungsten- or molybdenum-target X-ray tube are adequate for the detection of 0.2 ppm of uranium.
z
(9) A. P. Langheinrich, J. W. Forster, and T. A. Linn, Jr., Anal. Instrum., 9, F-3 (1971).
384
Table I. Relative Precision for EDX Assay of Uranium. Uranium concn level, ppm Uranium measured, Re1 dev, PPm 1 2 5 10 20 a
z
1 . 0 + 0.2b 2.0 i0.1 5.0 + 0.3 10.0 + 0 . 7 20.0 j= 0 . 8
20 5 6 7 4
Ten replicates of each concentration level. Confidence limits of an average as calculated from the range.
Table 11. Comparative Uranium Assay of Solution Samples Method Sample EDX, ppm Chemical, ppm A
B
C
10 11 11 8 8.2 8.9 8.2 6.3 6.5 2.3
11 11 11 8 8.8 9.9 8.7 6.6 6.4 2.3
RESULTS AND DISCUSSION
Uranium standards of 0, 10, 20, 50, and 100 gg produce a linear calibration curve of the form y = mx 6, where typical values for m and b are 14.1 and 1065.6, respectively; and the standard error of estimate is 3.0 pg of uranium. The relative precision of the method was tested by 10 replicate analyses of process tailing solution samples. The results from this and similar tests on samples with higher levels of uranium are shown in Table I. A comparison of results with data from a solvent extraction-spectrophotometric method ( / a ) is illustrated in Table 11. The procedure has been applied successfully to samples of ground water, acid process solutions, mine waters containing large amounts of diverse metal cations [ e . g . , up t o 25 grams of mixtures of Mg (11), AI (111), Fe (11), Fe (111), and Cu (11) per liter of solution], waste waters, and brines. Standards prepared in aliquots of distilled water, culinary water, synthetic mine water, and various acid and salt solutions, all give a common standard curve regardless of the original matrix. Interferences in the X-ray determination of uranium have been reported ( / I ) ; however, the separation procedure effectively removes all but the actinide elements, which are of extremely rare occurrence in the samples studied. The separation and direct instrumental measurement of traces of uranium offers the advantages of speed and convenience, removal of interferences, sensitivity of 0.2 ppm of uranium (using a 25-ml aliquot), and application to many kinds of samples.
+
RECEIVED for review July 19, 1971. Accepted August 26, 1971. (10) C. A. Francois, ANAL.CHELI., 30, 50(1958). (11) W. J. Campbell and H. F. Carl, ibid.,27,1884 (1955).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 2, FEBRUARY 1972