• mass spectrometry;
  • nitrogen isotopes;
  • triple collection

[1] We describe a new system for the simultaneous static triple-collection of nitrogen isotopes at the <10μcm3 STP [N2] (<1 × 10−5 cm3STP; <0.5 nmol) level using a modified VG-5440 noble gas mass spectrometer. The system consists of an internal N2-STD with aδ15N value of −0.11 ± 0.22 ‰ (1σ) calibrated against an air-standard (Air-STD). The N2-STD was measured repeatedly with an average uncertainty on an individualδ15N measurement being 0.03 ‰ (1σ) versus an average single day reproducibility of 0.38 ‰ (1σ). Additional refinements include (1) monitoring of interfering CO contributions at mass 30, allowing a comprehensive CO correction to be applied to all samples, (2) quantification of procedural N2 blanks (n = 22) in both size (4.2 ± 0.5 μcm3 STP) and isotopic composition (δ15N = 12.64 ± 2.04 ‰), allowing consistent blank corrections to all samples, and (3) independent measurement of N2/Ar ratios using a quadrupole mass spectrometer (QMS). The new system was tested by measuring nitrogen isotopes (δ15N), concentrations and N2/Ar ratios on 11 submarine basalt glasses. Results show that the uncertainty on the δ15N data is improved as a consequence of multiple standards being run per day. Reduced analytical times, afforded by triple collection, also minimize sample depletion and memory effects, thus improving measurement statistics. Additionally, we show that CO corrections can be accomplished using mass 30 to monitor CO interferences, leading to substantial improvements in reproducibility and the overall accuracy of results when the contribution of CO is significant.