E. Morenzoni

Precision measurements of 14C in AMS — some results and prospects

Abstract Some of the factors affecting the precision in AMS measurements will be discussed and the specific developments undertaken to reduce the errors at the ETH facility are described. Based on a large number of 14C measurements we show the present limitations of our system and consider what improvements might be possible. Further, a comparison is made between high precision measurements of 14C and current achievements relating to Be, Al and Cl.

Mass-spectrometric 10Be dating of deep-sea sediments applying the Zürich tandem accelerator

Abstract We have measured 10 Be in deep-sea sediment cores from the Central North Pacific (GPC-3, 30° N 157° W) and from areas with high biological activity, off West Africa and the South Atlantic near Antarctica. In core GPC-3 we observe a decrease with depth which can be decomposed into three exponentials. Sediment accumulation rates of 2 mm/ka for the upper 2.2 m, 1.1 mm/ka for the section between 2.2 m and 4.7 m and 0.5 mm/ka from 4.7 m to 10 m have been derived. The changes in the accumulation rate occurred at 1.1 Ma and 3.3 Ma BP. In core GPC-3 the 10 Be chronology agrees (20%) with paleomagnetic dating of the uppermost 4.2 m, corresponding to a time range of 2.43 Ma (Matuyama-Gauss Boundary). The average 10 Be flux into the sediments at locality GPC-3 during the past 1.1 Ma is comparable to the assumed depositional flux from the atmosphere. On sediment cores from off West Africa and the South Atlantic we observe, on the contrary, fluxes of 10 Be which exceed more than 10 times the depositional flux. This indicates that the concentration of 10 Be in the open ocean at localities in the proximity of highly productive areas may be strongly modulated by variations of bioactivity throughout time.

The camp century 10Be record: Implications for long-term variations of the geomagnetic dipole moment

Abstract 10 Be concentrations measured in ice samples from Camp Century, Greenland, show short term variations which in general correspond to the 100–200 year “wiggles” in the 14 C tree ring record. There is, however, no evidence for a long term variation over the last 5000 years. This constancy is in contrast to the approximately sinusoidal variation of the atmospheric 14 C concentration which has generally been attributed to a changing geomagnetic dipole moment. This discrepancy implies that the 14 C trend might stem from other causes such as changes of oceanic circulation processes or from higher production rates during the Wisconsin rather than from variation in the geomagnetic field.

Diffusion and supply rates of 10Be and 230Th radioisotopes in two manganese encrustations from the South China Sea

Ages of two Mn encrustations estimated by their 230Th and 10Be distributions are compared with K-Ar ages and micropaleontological datings of their nuclei to discuss possible diffusion and supply effects of the radioisotope distribution and their influence on the reliability of age determinations. Based on comparable results obtained by the different methods the effective diffusion coefficient of 10Be can be calculated as D∗ ≤ 1.0·10−8cm2/y. This coefficient is 3–8 times smaller than the best estimates available at present. In both nodules we observe lower 10Be concentrations in the uppermost 2–3 mm (1.3 m.y.), which suggests that 10Be uptake has been reduced since the middle Pleistocene. The 2.7-fold increase of the growth rate starting 3.2 m.y. ago coincides with the initiation of the northern hemisphere glaciation.

10Be dating of the inner structure of Mn-encrustations applying the Zürich tandem accelerator

Abstract We have measured 10 Be and 230 Th depth profiles in Mn-nodules and crusts from the Central North Pacific (2 samples), from the South China Sea (2 samples) and from the South Pacific (1 sample). Comparison of the nucleis ages derived via 230 Th and 10 Be datings and other methods (K/Ar and paleontology) shows very good agreement and proofs that distortion of 10 Be and 230 Th ages due to radionuclide diffusion in Mn-encrustations is negligible ( In all samples we observe a smooth exponential decrease for larger sections, suggesting constant depositional 10 Be fluxes within the time resolution of 0.4 to 0.8 Ma. (resulting from the thickness of the sampled intervals). Growth rates range from 2 mm/Ma to up to 20 mm/Ma. Common to all samples is a remarkable change of growth rate accompanied by a visible change of the mineralogical inner texture, dated at 6.2 Ma. We observe slower growth rates during the last 6.2 Ma than in the previous periods. Provided that the Mn-concentration in deep water has remained constant throughout time (as supported by nearly constant trace metal/manganese ratios of the samples) the data suggest that the bottom water circulation pattern has slowed down since 6.2 Ma BP. Visible changes in the inner texture corresponding to paleooceanographic time marks such as 1.3 Ma (a presumed variation of the ice volume during Middle Pleistocene), 3.3 Ma (Glaciation of the Northern Hemisphere), and 12–14 Ma ago (maximum Glaciation of the Antarctica) strongly suggest that the growth pattern of Mn-encrustations has been influenced by the response of the oceans to the history of climatic changes.

10Be annual fallout in rains in India

The 10Be concentrations of annual rainfall collections during 1979–1981, at eight stations in India, ranged from 0.43 × 107 to 8.48 × 107 atoms/l and the corresponding 10Be fallouts are in the range of 0.31 × 10 6 to 2.73 × 106 atoms cm−2 a−1. The estimated 10Be global fallout based on the presently available data is 1.55 × 107 atoms−2 a−1 or 5 × 10−2 atoms cm−2 s−1. Most of the measured rates of fallout and deep sea deposition of 10Be are a factor of 2–3 lower than the present estimate.

14C dating of polar ice

Abstract We have dated ice core samples by 14C AMS to show that this method extends other ice dating methods and enables direct comparison with 14C related climatic events found elsewhere. For the measurement we use the CO2 that was occluded together with the air at the time of ice formation. To extract the CO2 we crush the samples with a milling cutter. For a 14C AMS measurement the CO2 content of about 10 kg of cleaned ice (0.25 cm3) has to be converted to amorphous carbon. First 14C results on the Dye 3 (Greenland) ice core are in good agreement with other age determinations.

14C measurements on foraminifera of deep sea core V28-238 and their preliminary interpretation

Abstract In this paper first results obtained by AMS dating foraminifera are presented. The amount of material used for these studies was 7 to 10 mg calcium carbonate. A preliminary interpretation is given.

Spallogenic nuclides in meteorites by conventional and accelerator mass spectrometry

Light noble gases and 10Be were determined in 15 ordinary chondrites by conventional mass spectrometry and accelerator mass spectrometry, respectively. Aliquots of these samples had previously been studied for 26Al and 53Mn. Be was extracted from 10–20 mg of meteorite and 10Be was determined by means of the 6 MV tandem Van de Graaff accelerator at ETH-Zurich with a reproducibility better than 3%. A detection limit of 0.7 dpm/kg meteorite was achieved. Based on the noble gas data a 10Be production rate of 17.8 ±0.4 dpm/kg was established for chondrites with H-group chemistry. Short exposure ages of meteorites were determined via the 10 Be concentrations. A slight correlation of 10Be production rates with shielding could be observed. During the last 4 million years the long-term averaged cosmic ray intensity did not vary by more than 30%.

Target preparation for milligram sized 14C samples and data evaluation for AMS measurements

Abstract Our preparation technique produces in a glow-discharge an amorphous carbon deposit on a copper substrate. The process starts with 1.6 cm3 CO2 STP (900 μg carbon) which is reduced over hot zinc to CO and subsequently cracked in the discharge. The yield of the process is typically 80%. With these targets in the Zurich ion source ion currents up to 20 μA are obtained. The background of samples prepared with this technique is presently around 30 ka (2.5% MODERN). The precision after half an hour measuring time for a modern sample is 0.7% and 2.7% for a three half-lives old sample, including the errors of the background and the NBS oxalic acid measurement. The method we use to correct for the background of the preparation and the accelerator as well as for the fractionation in the accelerator is presented.