³⁶Cl-AMS at the DREAMS facility

The DREAMS (DREsden Accelerator Mass Spectrometry) facility [1] went operational in August 2011. It is based on a state-of-the-art 6 MV Tandetron. The target wheels of the two Cs sputter ion sources can be loaded with up to 200 samples each. A fast bouncing system at the 90° injection magnet allows almost simultaneous measurements of the radioisotope and its stable isotope(s). For the measurement of ¹⁰Be and ³⁶Cl a 1 µm Si₃N₄-post-acceleration stripping-foil in combination with a 35° ESA suppresses their stable isobars ¹⁰B and ³⁶S. An ionization chamber with four anodes is used for the final identification of the radioisotopes.

At the moment routine measurements of ¹⁰Be, ²⁶Al and ⁴¹Ca are performed at DREAMS. In the case of volatile elements like Cl and I, the key issue for precise AMS-measurements is the understanding and minimization of ion source memory effects [2,3]. For this purpose the main focus for ³⁶Cl-measurements at DREAMS was set to the mechanical modification and improvement of the original SO110 ion source from High Voltage Engineering (HVE) [4]. The new DREAMS design has a more open geometry to improve the vacuum level and a modified target loading and positioning system, which allows exchanging the individual cathode aperture with each target.

In order to evaluate this improvement in comparison to other up-to-date ion sources, an interlaboratory comparison with three AMS labs had been initiated. The long-term memory effect in the four Cs sputter ion sources of VERA [5] (Vienna Environmental Research Accelerator, NEC ion source: MC-SNICS), ASTER [3] (Accélérateur pour les Sciences de la Terre, Environnement, Risques, modified HVE SO110) and DREAMS (original HVE SO110 and modified HVE SO110 ion source) had been investigated by running samples of natural ³⁵Cl/³⁷Cl-ratio and samples containing highly-enriched ³⁵Cl (³⁵Cl/³⁷Cl > 500). We choose to perform these measurements with the two stable Cl isotopes instead of using ³⁶Cl standards and blanks to avoid limitations of accuracy by counting statistics on low level ³⁶Cl samples.

The primary goals of these experiments are the determination of the time constants of the recovery from the contaminated sample ratio to the initial ratio of the sample and the level of the long-term memory effect in the sources.

[1] S. Akhmadaliev et al., Nucl. Instr. Meth. B 294 (2013) 5.
[2] R. Finkel et al., Nucl. Instr. Meth. B 294 (2013) 121.
[3] M. Arnold et al., Nucl. Instr. Meth. B 294 (2013) 24.
[4] M.G. Klein et al., Nucl. Instr. Meth. B 249 (2006) 764.
[5] M. Martschini et al., Nucl. Instr. Meth. B 269 (2011) 3188.