Challenging developments in three decades of accelerator mass spectrometry at ETH: from large particle accelerators to table size instruments.

Accelerator mass spectrometry (AMS) was invented for the detection of radiocarbon at natural isotopic concentrations (10(-12) to 10(-15)) more than 30 years ago. Meanwhile this method has also been applied for the analysis of many other long-lived radioisotopes, which are found at very low concentrations. The first investigations were made at large tandem accelerators originally built for nuclear physics research and operating at voltages of 6-12 MV. Today dedicated instruments are mostly used for AMS, which are optimized for associated applications. In the past 15 years, a new generation of much smaller instruments has been developed. For many years it was believed that accelerators with voltages of 2 MV or higher are needed to eliminate the molecular interferences. At these energies the ions are predominantly stripped to charge state 3+, thereby removing the binding electrons of the molecules. In contrast, the new compact facilities use 1+ or 2+ ions. In this case the molecular destruction process is based on molecule-atom collisions in the gas cell. The cross sections for this destruction are sufficiently large that the intensity of molecular components such as (12)CH(2) and (13)CH can be reduced by 10 orders of magnitude. These new facilities can be built much smaller due to the lower energies. Universal instruments providing analysis for many isotopes over the whole range of periodic table have a space requirement of about 4 x 6 m(2); dedicated radiocarbon facilities based on a 200 kV accelerator have a footprint of about 2.5 x 3 m(2). This smallest category of instruments use special technologies: The high voltage terminal with the gas stripper canal is vacuum insulated and the gas is pumped to ground potential through a ceramic pipe. A conventional 200 kV power supply provides the terminal voltage from outside. A review of this new generation of compact AMS facilities is given. Design considerations and performance of these new instruments will be presented. With these developments, new AMS instruments are not much larger than conventional mass spectrometers, allowing a significant reduction in cost.