Potential ways to increase C- from CO2-fed sputter ion sources

Dr John Vogel1

1Univ Of California, ukiah, united states

The CO2-fed ion source was first proposed at the Third AMS Symposium in 1984. At the same gathering, the chemical method of reducing CO2 to a fullerene on fine iron powder was also presented. This latter method has been the mainstay of 14C AMS sample preparation to the present day, despite the numerous advantages of the gas sample approach. A number of laboratories built prototype gas-fed sources in the 1990’s after Middleton showed his improved gas source in 1989. Oxford, especially, had an active measurement program in the 90’s using CO2, but by 2004 it became clear that ion currents from CO2 were routinely 5-15 µAmps from actual samples, although “tank” gas occasionally produced 25 µA. Detailed analyses of the CO2-fed ion source continued, but many AMS labs, especially those having MICADAS equipment, resigned themselves to their lower-than-desired ion current and proceeded with their small sample science programs. At the last AMS meeting, I proposed that there were two ionization processes at work in the sputter ion source: surface ionization which should limit solid samples to 30 µA of C- (at Icathode = 1 mA); and volume ionization from sputtered neutral C atoms scattering off neutral, excited cesium with up to 100 µA available from recessed samples. Very large currents (300 – >400 µA) result from the volume ionization remaining active to an immersion lens located a few mm in front of the sample. The cross sections for this ionization can be calculated from equations given in my AMS-14 paper. They are in the hundreds of Å2, many of orders of magnitude greater than electron attachment or even surface ionization. I had shown in an earlier paper that other sputtered elements could compete with carbon for the Cs atoms of the desired excited state to depress C-. Oxygen particularly competes with carbon for the most plentiful Cs excited state, Cs(5d), having 62% higher cross section than carbon at this energy. The near-Gaussian cross sections reverse dominance, however. by exciting the Cs to the Cs(7s) state from the Cs(5d) with a 1359 nm laser. A competition model gives 50% more C- at that state.This is for the volume ionization only. I am uncertain how to calculate the surface ionization of gas. In his 1989 paper, Middleton said that CO gave. marginally more C- current than did CO2, but the problem of reduction was probably not worth it. This may merit revisiting, considering that the present sources are operated differently from his. He used a CaCO3 heated source of CO2, without He carrier, and the samples were generally bigger than those of today. A recent theoretical chemistry paper compared binding of CO, CO2 and CH4 to a. titanium surface and showed that while one oxygen was removed from both CO ,and CO2, the CO carbon remained on the surface while the CO2 carbon became more tightly bound to a lower level of Ti atoms.


40 years in. AMS: first at Simon Fraser Univ/McMASTER U niv, then at CAMS@LLNL. RETIRED, but still thinking about ion sources.

  • 00


  • 00


  • 00


  • 00



Nov 08 - 19 2021