The reduction of sample memory effects in the Chalk River AMS ion source

Abstract

The mechanism underlying Cl and I sample-to-sample interference in the new Chalk River AMS ion source has been studied and compared with the interference observed in an earlier ion source of different internal geometry. The distribution of sputtered material and its degree of migration was measured with the radioactive tracer, 82Br. The temperature dependence of the surface constitutent was measured with the elastic recoil detection (ERD) technique and the effect of sample geometry and ion source cleaning was studied with elevated (5 × 10−10) 36ClCl and 129II samples. These measurements indicate that a hot (>350°C) aperture plate ahead of the sample can prevent the sputtering of contaminated regions near the sample. The plate itself remains relatively free of Cl or I itself since these elements or their Cs-gettered compounds are desorbed at this temperature. A small, fixed quantity of Cl or I on this surface is observed, which if sputtered by Cs+ ions, may contribute to ion source memory. Relative sample-to-sample interference for both Cl and I is about 10−3 after 20 min or 10−4 after 60 min.