Contents lists available at ScienceDirect Nuclear Inst. and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb Chlorine measurements at the 5MV French AMS national facility ASTER: Associated external uncertainties and comparability with the 6MV DREAMS facility R. Braucher a, ⁎ , K. Keddadouche a,b , G. Aumaître a,b , D.L. Bourlès a,b , M. Arnold a,b , S. Pivot a , M. Baroni a , A. Scharf c , G. Rugel c , E. Bard a a Aix Marseille Univ, CNRS, IRD, Coll. France, CEREGE, Plateau de l’Arbois, BP 80, 13545 Aix en Provence, France b ASTER-Team, Aix Marseille Univ, CNRS, IRD, Coll. France, CEREGE, Plateau de l’Arbois, BP 80, 13545 Aix en Provence, France c Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, Bautzner Landstr. 400, 01328 Dresden, Germany ARTICLE INFO Keywords: Accelerator mass spectrometry Chlorine Standard Cosmogenic nuclides ABSTRACT After 6 years of 36 Cl routine operation, more than 6000 unknown samples have been measured at the 5MV French accelerator mass spectrometry (AMS) national facility ASTER (CEREGE, Aix en Provence). This paper presents the long term behavior of ASTER through the analysis of the measurements of the most used chlorine standards and reference materials, KNSTD1600, SM-Cl-12 and SM-CL-13 over a 46 months’ time period. Comparison of measured chlorine concentrations (both 35 Cl and 36 Cl) from ice samples on two AMS facilities operating at 5MV (ASTER) and 6MV (DREAMS, Helmholtz-Zentrum Dresden-Rossendorf) and normalizing to two different reference materials agree within uncertainties making both reference materials (SM-Cl-12 and KNSTD1600) suitable for 36 Cl measurement at ASTER. 1. Introduction In April 2017, ASTER, the 5MV French Accelerator Mass Spectrometry (AMS) national facility hosted by CEREGE in Aix-en- Provence will pass a decade since the acceptance tests. Since then, ASTER is routinely measuring 10 Be and 26 Al [1,2]. The ions extracted from a SO110 hybrid ion source [3]are first energy-analysed by a 54 degree electrostatic deflector before being mass-analyzed by a 90 de- gree magnet equipped with a fast bouncing system that sequentially injects the isotopes of interest with a repetition rate of ∼ 100 Hz. A fast beam blanking unit defines with nanosecond resolution the exact duration during which the different isotopes are injected through the accelerator. The HVE model 4150 Tandetron TM accelerator [4] is equipped with an all-solid-state power supply. The high energy-spec- trometer features a 90 degree analyzing magnet with Faraday cups installed for measurement of the stable isotopes. Faraday cups are fol- lowed by SiN absorber foil and a 35 degree electrostatic deflector. Two sets of Q-pole doublets just before and after the electrostatic deflector serve for proper focusing of the beam that suffers from substantially emittance growth by scattering in the foil. A 30 degree magnet reduces background from ions that have been scattered on the deflector elec- trodes. The magnet is mounted vertically to uncouple the dispersive action of the electrostatic deflector and the magnet. The measurement of all isotopes is gated in synchronization with their corresponding injection periods. The rare isotope is detected in a high-resolution, 4- anode gas ionization chamber. Because regarding volatile elements ( 36 Cl, 129 I) the initial Cs-sputter ion source SO110 induced sample-to-sample cross-contamination, the source and aperture designs as well as the shape of the cathodes have been upgraded to reduce the resulting memory effects to significantly less than ∼0.1%, allowing routine measurements of these volatile elements [5,6] since 2010. More specifically, an average number of 1000 36 Cl unknown targets are measured per year, all being normalized to AgCl materials (KNSTD1600; 36 Cl/ 35 Cl = 2.112 × 10 -12 ) prepared by Kuni Nishiizumi [7]. To monitor the long term variability of ASTER and to determine the external uncertainties associated to the 36 Cl/ 35 Cl, 36 Cl/ 37 Cl and 35 Cl/ 37 Cl ratios, it has be decided to measure together with the KNSTD1600 standard, the SM-Cl-X reference materials. Pre- pared by S. Merchel under the auspice of the European project “CRONUS-EU”, these SM-Cl-X reference materials have been cross-ca- librated through an interlaboratory comparison involving eight AMS facilities worldwide. The resulting round-robin data are fully presented in Merchel et al. (2011) [8]. The aim of this paper is therefore not to refine the SM-Cl-X standards calibration with more data points but to https://doi.org/10.1016/j.nimb.2018.01.025 Received 16 March 2017; Received in revised form 29 November 2017; Accepted 25 January 2018 ⁎ Corresponding author. E-mail address: braucher@cerege.fr (R. Braucher). Nuclear Inst, and Methods in Physics Research B 420 (2018) 40–45 0168-583X/ © 2018 Elsevier B.V. All rights reserved. T