Two manuscripts recently have been accepted in RapidCommunications in Mass Spectrometry that detail aninternational collaboration to identify, understand at afundamental level, and correct measurement biasesmanifested in gas isotope ratio mass spectrometers.These biases arise from memory effects in ion sourcesand affect the reliability and comparability of carbon-13 and oxygen-18 measurements used in studiesrequiring high interlaboratory reproducibility, includingresearch relevant to climate change. In this regard,measurements of light stable isotopes in carbondioxide, methane, and other atmospheric trace gasesprovide a unique means to better understand theirsources, fates, and contributions in biogeochemicalcycles. This collaboration already has resulted in thetangible improvement of the measurement tools usedby dozens of international researchers, and improvedthe value assignments of isotope reference materialsdistributed by NIST and the International AtomicEnergy Agency (IAEA).The first manuscript details an intercomparisonexercise that was successful in providing extensivemeasurements of sufficient quality needed to constructmechanistic models of memory effects in the massspectrometers. The models were consistent with thepresence of two instrumental memory sources oneshort-lived (10 s to 20 s) and the other long-lived(6 min to 10 min) that could not be compensatedaccurately by normal background corrections.Observed biases were substantial and dependent uponseveral operational parameters. The results, originallypresented in December 1995 to the IAEA Consultants Meeting for Light Stable Isotope Reference Materialsin Vienna, Austria, initiated research and developmentactivities to address these effects at several otherorganizations in collaboration with NIST.The second manuscript describes the results of acollaboration with an instrument manufacturer, to testnew focusing plates and slits in the NIST ion sourceand in an identical instrument at the Max-Planck-Institute for Chemistry in Mainz, Germany. The newion source materials were designed and engineeredto minimize the specific memory effects previouslyidentified, and historical data from the NIST instrumentprovided the benchmarks needed to compareprecisely and document the change in observedperformance. Biases in carbon-13 and oxygen-18measurements were decreased by up to a factor offive, which enabled up to a 50 % reduction in theuncertainty of reference material value assignments.02The newly engineered ion source is incorporated in theisotope ratio instruments, which has benefited theresearch of dozens of research groups worldwide. Thismanuscript reveals the technology to a wider audienceof instrument manufacturers and offers several practicalrecommendations to the measurement communitiesregarding the minimization of the biasing effects.
Citation: Journal of Research (NIST JRES) -
Volume: 108 No. 2
NIST Pub Series: Journal of Research (NIST JRES)
Pub Type: NIST Pubs