|Home||Archived October 29, 2018||(i)|
Major Sediment Characteristics
In the laboratory, the bagged samples were first weighed to determine the water loss. A wet aliquot was placed in pre-cleaned pre-weighed porcelain evaporating dish, dried at 40 deg. C, cooled and re-weighed to determine water loss. This sample was dissolved in 6N HCl. The filtrate was washed three times with distilled water and dried. The residue was reweighed and reported as percent insoluble residue.
A separate aliquot (ca. 20 g wet weight) was weighed then sieved (62 micron mesh) using distilled water to obtain the amount of fine sediment in each interval. These samples were freeze-dried. Portions of dried samples were ground to a fine powder (75-100 mesh) in a grinding mill to obtain a homogeneous sample for analysis. Sample splits were made and 5g placed into pre-cleaned, pre-weighed crucibles and heated in a muffle furnace at 450 deg. for 6 hours so as to obtain a stable weight. The samples were then cooled and reweighed to determine loss on ignition. The remain fraction was submitted for 210Pb analysis.
Total extracted lead-210
The analysis of sediments at the USGS-Denver is based on counting of Po-210, in secular equilibrium with its parent Pb-210, and exploits the ability of polonium to self-plate onto silver and certain other metals (FLYNN, 1968). Five grams of the >62 micrometer dried, ground sample was transferred to a 100 ml Teflon beaker and mixed with 5-10 ml of reagent grade 16N nitric acid. An amount of NIST-calibrated Po-209 spike was added and the sample swirled to mix the spike. The beaker was covered with a watch glass and allowed to stand overnight. The solution with solids was allowed to evaporate under heat lamps at 90 deg. C. The sample was washed from the sides of the beaker using 8N Hydrochloric acid and swirled again to insure proper mixing. The solution is evaporated again and allowed to cool. One ml of 30% hydrogen peroxide was then added to the sample and the resulting mixture evaporated to dryness. These latter steps, adding peroxide and drying the mixture were repeated an additional two times. The 8N HCl was added to the sample and allowed to dry out. This step was repeated so as to ensure that nitric acid, which hinders efficient plating is virtually eliminated from the mixture. Finally 5ml of 8N HCl were added to the dried sample and the mixture transferred to a 100ml beaker using additional amounts of de-ionized water to insure essentially complete transfer. To minimize interference of several ions with plating, 5ml of hydroxylamine hydrochloride and 2ml of 25% sodium citrate were added to each sample. Additionally, 1ml of a hold back carrier, bismuth nitrate, was added to prevent deposition of Bi-212. A plastic coated magnetic bar was added to each beaker for stirring during autoplating. The pH of the solution was adjusted to between 1.85 and 1.95 using ammonium hydroxide. The beaker was placed on a hot plate and heated between 85 and 90 deg. C for 5 minutes to reduce iron, chromium and oxidants present. Then a Teflon holder which exposes one side of a silver foil disc was placed in the solution for a minimum of 90 minutes. Rinsed and air-dried silver discs were then counted for polonium isotopes by alpha spectroscopy. Combined analytical and counting errors in determining lead-210 values are about 3%.
For purposes of intercalibration, fifteen samples of sieved sediment (from Bob Allen Core 6C) were analyzed at the GLERL laboratory for lead-210 by counting its decay product, polonium-210. Two gram samples were placed in beakers, 1 ml of calibrated polonium-209 standard solution added to each, followed by 30 ml of concentrated HCl transferred slowly to avoid excessive foaming. After subsidence of foaming (in 24 hrs) the mixtures were placed in centrifuge tubes and allowed to stand for two weeks at room temperature. The mixture was then centrifuged and the supernatant decanted. The residue was rinsed, re-centrifuged and the second aliquot of supernatant added to the first. The above procedure was then repeated on the rinsed residue with the additional step of first slowly adding 10 ml of 30% hydrogen peroxide to oxidize organic matter and presumably release any bound lead-210 into the extractant. To both sets of supernatant solutions, one gram of NH2OHúHCl was added as in the Denver method to reduce the interference of dissolved iron with polonium plating (FLYNN, 1968) and the solutions brought to a volume of 100 ml with distilled water and adjusted to a pH of 1.5 via small additions of NH4OH. Polonium isotopes were self-plated onto polished, Mylar-backed copper disks placed in the solutions (at 85 C) and gently agitated for eight hours. Exposed disks were alcohol-rinsed and counted by alpha spectrometers consisting of surface barrier detectors coupled to a multichannel analyzer. The polonium-209 solution was calibrated by comparison with standard solutions of polonium-208 (NIST SRM 4327) and with polonium-210 in secular equilibrium with lead-210 in an NIST-traceable standard solution obtained from the U. S. Environmental Protection Agency (Environmental Monitoring Systems, Las Vegas, Nevada). The activity of the polonium-208 solution has an overall uncertainty of 1.4 percent. The activity of the polonium-209 solution is consistently determined by the two independent methods and presently known with an overall uncertainty of about 2.5%. Uncertainty due to random errors decreases in time because the activity of the solution is re-checked every few years by intercomparison with new NIST-traceable aliquots of polonium-208. Overall uncertainty in counting were about 2%.
Total lead-210 and radium-226
A larger aliquot of wet sediment was washed through a 62 micron stainless steel sieve using distilled water. The fine fraction was allowed to settle, overlying water decanted and the final slurry was freeze-dried and disaggregated. At USGS-Woods Hole, 10-50 grams of the freeze-dried fraction were transferred to a plastic counting jar with a tight sealing plastic screw cap. Sealed counting jars were stored for at least 20 days to allow for the in-growth of radium-222 and lead-214 to approximate equilibrium values.
Samples were counted on a gamma detector system consisting of a germanium detector for low energy gamma rays (2000 mm2 area) and a 4096 channel multichannel analyzer. Samples were typically counted for 48 hours (depending on sample size) or until counting errors were <5%. The system has been calibrated using EPA standard pitchblende ore in the same geometry as the samples. A self-absorption correction (CUTSHALL et al., 1983) has been applied to each sample. Results of the lead-214 analysis (for radium-226) of the Canadian ore standard DH-1a were within 3.5% of the published value. Total lead-210 results on the same standard were 1952 ñ14 dpm/g (n=2) compared with the published value of 1848 ñ54 dpm/g, a 5.6 % difference.
At the GLERL laboratory radium-226 was determined on 4.00 ñ0.02 g of dry, desegregated sediment less than 62 micrometer. Samples were packed into counting vials to a constant height 4.0 ñ0.05 cm to assure a fixed counting geometry and packing density. Vials were sealed with epoxy cement and stored for at least three weeks to allow for in-growth of radon-222 and its gamma-emitting progeny. Samples were placed in an intrinsic (HpGe) well detector and counted for up to two days to achieve acceptable counting statistics. Specific activity of radium-226 was determined from counts associated with the gamma photopeaks of Pb-214 (295 and 351 keV) and Bi-214 (609 KeV). Activities calculated from the three peaks were combined to yield a weighted mean reported value and standard deviation. The system was calibrated and frequently recalibrated by counting standards prepared by doping portions of Bay sediment with precisely known amounts of a radium standard solution (NIST SRM-4959). This solution has a 0.4% uncertainty in activity at the 99% confidence level due to random errors and an additional 0.8% uncertainty due to assessable systematic errors. Reported standard deviations in radium activity, including random errors associated with detector calibration, were typically 5-7%. The limit of detection was about 0.1 dpm/g for this matrix.
The GLERL well and planar detectors were also calibrated for cesium-137 using sediment doped with a standard gamma-emitting, mixed radionuclide standard (Amersham Searle, QCY.46) traceable to the British Calibration Service. The system efficiency has also been calibrated using several NIST standard reference materials, (SRM-4354, Standard Lake Sediment; SRM-4353 Rocky Flats soil Number 1 ; SRM-4353, River Sediment). In order to achieve acceptable counting statistics, all samples having significant radiocesium activity on the 1 to 2 day count for radium were recounted for up to five days each. The resulting overall detection limit was about 0.02 dpm/g corresponding to a minimum uncertainty in reported activities of about 10%. Cesium-137 was also determined at USGS-Woods Hole on selected samples by counting the 661.7 KeV gamma ray. Results for IAEA Standard 4018 were within 3% of published values. The detection limit was also about 0.02 dpm/g.
Trace metal analysis (lead and others)
To determine concentrations of selected trace metals (lead, uranium, barium and mercury) 0.2 grams of freeze-dried whole sediment from Bob Allen 6A, 6C and Russell Bank 19C were placed in microwave digestion vessels to which was added 20 ml of 10% Ultrapure nitric acid. The mixtures were processed by microwave digestion at 160 deg. C for 1 hour. Ten ml of the resulting solutions were diluted 2.5x and analyzed for metals using an ICP-M system (Perkin Elmer Elate 5000A) focused to detect Pb-208 and U-238. The system, situated in a class 100 laboratory, was calibrated linearly using standard concentrations of 1, 10 and 100 ppb of lead and uranium and checked using synthetic water reference (NIST-1643C) as the external standard. Long term replication of that standard indicated a probable uncertainty in the determination of both metals as less than or equal to 5%. Concentrations of uranium reported as ppb were re-calculated in terms of specific activities (dpm/g) for comparison with lead-210 and radium-226 results.
Isolation of the Non-Carbonate Fraction
Starting from the top of core Bob Allen 6C, roughly equal weights of dried, sieved sediment from 10 contiguous 2 cm sections of core were combined to form a composite sample weighing 65 g spanning a 20 cm segment. Eight such composite samples were prepared for each 10 successive 2-cm intervals downcore which together spanned the entire 160 cm length of the core. Four grams of each ground, homogenized composite was packed into standard vials and sealed for later counting while a 60g sample was placed in a 8 L container to which a stochiometric excess of dilute (10%) acetic acid was added to dissolve carbonates. Following the primary phase of gas evolution, the entire mixture was quantitatively transferred to large beakers, the solids allowed to settle and the supernatant decanted. Fresh 10% acetic acid was added along with a magnetic stirring bar. The mixtures were stirred continuously at ambient temperature (ca 20 deg. C) for an additional three days to ensure complete dissolution of carbonates. Subsequently, the mixtures were allowed to settle, the supernatant discarded and residues rinsed with distilled water three times to remove soluble products. Following drying, 4 grams of each residue were packed into standard vials and sealed for subsequent gamma counting.
|Home||Archived October 29, 2018|