We have studied isotopic fractionation effects which occur during the
analysis of oxygen isotope ratios by secondary ionization mass spectro
metry. Variable instrumental isotopic fractionation has been well docu
mented in the past as making reliable ion microprobe analyses of oxyge
n isotopes elusive. We report here techniques for minimizing these eff
ects by careful optimization of an Isolab 54 ion microprobe and a meth
od of integrating the secondary-ion beam to eliminate a major source o
f fractionation caused by unequal focusing of isotopes on the source s
lit of the mass spectrometer. A further improvement has been the insta
llation of new ion detectors for measuring O-18- and O-17- secondary i
on beams in which the ions strike a conversion dynode (Conversion dyno
de system-CDS) to release electrons into a channeltron. The CDS detect
ors have a uniform, high and stable gain which, coupled with the beam
integration techniques referred to above, yield reproducible oxygen is
otope ratios on both conducting and insulating mineral standards (less
-than-or-equal-to 1.2 parts per thousand on O-18/O-16 during a day and
< 2 parts per thousand over periods of days and weeks).