A gas phase standard (SF6) with known d34S value was used for
tuning and calibration of isotope ratio values. In our earliest analyses,
the Ar ‘sample gas’ for the plasma was preheated by flowing
through the GC oven before entering the custom-built heated
transfer line. With this arrangement, both sensitivity and d34S values
drifted systematically (by nearly 10‰ in the latter case) as the
GC oven was ramped from 80–300 C. We now understand that
this effect arises from an imperfect response of the Neptune mass
flow controllers to changing flow resistance with GC oven
temperature. To account for such effects, sample d34S values were