Hello Peter,

Your assessment of the Oxford GC-AMS paper is the understatement of the 
decade.  Perhaps you could post your information publicly, so that labs 
around the world will not make Canada's mistake of "assuming" that the 
paper was complete or accurate.

I personally would like any information you can provide as to how to make 
the Oxford setup work on a regular basis.

And for anybody who is interested: in AMS, mass 12 and 13 are produced in 
microamperes and their Faraday cups are not usually optimised to make 
precise measurements.  Rather, these two isotope currents are used to fudge 
the C14 numbers for recovery, fractionation, etc when the ionisation starts 
from solid C targets.

Also, AMS is sometimes set up to inject three isotopes sequentially: the 
measurement time for masses 12 and 13 is milliseconds to hundreds of 
milliseconds because you need to spend most of your time measuring mass 14.

For CO2 gas input to AMS, the sputtering fractionation originates from a 
different phenomenon.  This depends on the design of the ion source, and a 
reliable discussion has not yet appeared in the refereed literature.  

A research project would need 2 years (starting from a properly-operating 
source and interface) to measure and program the corrections for gas input 
to AMS. Ideally, the samples would be paired with IRMS analysis, and then a 
real determination of equivalence would be available.

Irene Ellis