N 2 : a potential pitfall for bulk 2 H isotope analysis of explosives and other nitrogen-rich compounds by continuous-flow isotope-ratio mass spectrometry Wolfram Meier-Augenstein 1,2 * , Helen F. Kemp 2 and Claire M. Lock 3 1 Centre for Anatomy and Human Identification, University of Dundee, Dundee DD1 5EH, UK 2 Scottish Crop Research Institute, Stable Isotope Laboratory, Invergowrie, Dundee DD2 5DA, UK 3 Environmental Forensics and Human Health Lab., Environmental Engineering Research Centre, Queen’s University Belfast, Belfast BT9 5AG, UK Received 20 March 2009; Revised 6 May 2009; Accepted 6 May 2009 Observations made during the 13 C isotope analysis of gaseous CO 2 in the simultaneous presence of argon in the ion source of the isotope ratio mass spectrometer prompted us to investigate what influence the simultaneous presence of nitrogen would have on both accuracy and precision of bulk 2 H isotope analysis of nitrogen-rich organic compounds. Initially an international reference material, IAEA-CH7, was mixed with silver nitrate in various ratios to assess the impact that N 2 evolved from the pyrolysis of nitrogen-rich organic compounds would have on measured d 2 H-values of IAEA-CH7. In a subsequent experiment, benzoic acid was mixed with silver nitrate to mimic the N:H ratio of organic-rich nitrogen compounds such as cellulose nitrate and RDX. The results of both experiments showed a significant deterioration of both accuracy and precision for the expected d 2 H values for IAEA-CH7 and benzoic acid when model mixtures were converted into hydrogen and nitrogen, and subsequently separated by gas chromatography using standard experimental conditions, namely a 60 cm packed column with molecular sieve 5 A ˚ as stationary phase held at a temperature of 85-C. It was found that bulk 2 H stable isotope analysis of nitrogen-rich organic compounds employing published standard conditions can result in a loss of accuracy and precision yielding d 2 H values that are 5 to 25% too negative, thus suggesting, for example, that tree-ring 2 H isotope data based on cellulose nitrate may have to be revised. Copyright # 2009 John Wiley & Sons, Ltd. Since the first reports on on-line techniques for bulk and compound-specific 2 H and 18 O isotope analysis of organic compounds by means of pyrolytic or reductive sample conversion into H 2 and CO gas in the late 1990s, 1–5 research using 2 H/ 18 O continuous-flow isotope-ratio mass spectrom- etry (CF-IRMS) either in conjunction with other isotope analyses or on its own as a means to providing a new or additional source of information on authenticity, biochemical/ environmental fate or provenance of a given compound or material has generated new insights and opened new doors in many areas of applied analytical chemistry including forensic sciences. 6–18 One area of systematic analytical research underpinning all areas of applied stable isotope ratio mass spectrometry techniques that has recently been the focus of attention is the impact of the concurrent presence of a gas other than the analyte gas in the ion source of an IRMS instrument. Inter- ferences caused by the presence of a contaminant gas can impair both accuracy and precision of isotope ratio measure- ment, and, hence derived d-values. Studies published thus far have concentrated on the issue of isobaric interference of nitrous oxides with CO 2 measurement. Isobaric interference in IRMS analysis occurs when gas molecules giving rise to ions of the same mass/charge ratio as the analyte gas are present in the ion source at the same time. One example for isobaric interference would be that between CO and N 2 affecting both major and minor abundant ions of interest, m/z 28 and m/z 29. Other, more pernicious isobaric interferences are the effects of NO þ 2 (m/z 46) on the acquisition of the 12 C 16 O 18 O þ (m/z 46) ion current or the influence of NO þ (m/z 30) on the acquisition of the minor abundant CO ion 12 C 18 O þ (m/z 30). 19,20 However, we are looking at a pitfall here that is an inter- ference, which in the absence of a more apt description we have termed ‘Ionization Quench’ [IQ], meaning that molecules of the interfering gas X compete for or prevent ionization of the target gas T. We originally made this observation more than 15 years ago when we prepared and analyzed CO 2 gas samples comprised of 5 vol% CO 2 in argon for use as standards in the 13 C isotope abundance analysis of breath gas CO 2 , first on a dual-inlet IRMS instrument and again later on a continuous-flow IRMS system. Gas mixtures comprising CO 2 in argon invariably yielded measured d 13 C-values that were significantly lower (apparent d 13 C ¼59.8  1.2%) than known accepted d 13 C-values of 32.6  0.1% for the CO 2 gas RAPID COMMUNICATIONS IN MASS SPECTROMETRY Rapid Commun. Mass Spectrom. 2009; 23: 2011–2016 Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.4112 *Correspondence to: W. Meier-Augenstein, Centre for Anatomy and Human Identification, University of Dundee, Dundee DD1 5EH, UK. E-mail: w.meieraugenstein@dundee.ac.uk; wolfram.meier-augenstein@scri.ac.uk Copyright # 2009 John Wiley & Sons, Ltd.