Zhang and anyone else interested,

        Our lab has measured nitrogen (and oxygen and argon) isotopes in
air (and in trapped gases in ice cores and extracted from seawater) for the
past dozen years on a Finnigan MAT 251 and two 252 mass specs. If you are
measuring the isotopic values of these gases in an air mixture, then you
might be experiencing the following: On our mass specs, the measured
isotopic values are a function of the elemental ratios of the mixture. In
other words, the delta 15N of N2 is a function of the relative amounts of
O2/N2 in the mixture, and the delta 18O of O2 is a function of the N2/O2.
**Each of the mass specs has different relationships for the isotopes and
each relationship needs to be determined experimentaly** by mixing
experiments in which you measure your working std relative to a calibration
std (in our case dry air) to which you add varying amounts of pure O2 or N2
depending on which isotope you are calibrating.

        If you have removed the O2 from your samples, we have not seen any
relationship between the delta15N of N2 and the amount of Ar in samples
over a wide range of delta N2/Ar values (several hundred permil) on our
machines.

        One other possibility, the measurement of nitrogen isotopes is very
sensitive to water in the samples. If your numbers are consistently
enriched, they might be wet.

                                        Good Luck,
                                                Joe

Joe Orchardo
The Graduate School of Oceanography
University of Rhode Island
S. Ferry Rd.
Narragansett, RI, 02882
ph 401-874-6220
fax 401-874-6811