Non‐mass‐dependent ¹⁷O anomalies generated by a superimposed thermal gradient on a rarefied O₂ gas in a closed system

Abstract

Cryogenic or heating methods have been widely used in experiments involving gas purification or isolation and in studying phase changes among solids, liquids, or gases for more than a century. Thermal gradients are often present in these routine processes. While stable isotopes of an element are known to fractionate under a thermal gradient, the largely diffusion-driven fractionation is assumed to be entirely mass-dependent. We report here, however, that distinct non-mass-dependent oxygen isotope fractionation can be generated when subjecting rarefied O(2) gas in a closed system to a simple thermal gradient. The Δ(17) O value, a measure of the (17) O anomaly, can be up to -0.51‰ (standard deviation (s.d.) 1σ = 0.03) in one of the temperature compartments. The magnitude of the (17) O anomalies decreased with increasing initial gas pressures. The authenticity of this phenomenon is substantiated by a series of blank tests and isotope mass-balance calculations. The observed anomalies are not the result of H(2) O contamination in samples or in isotope ratio mass spectrometry. Our finding calls attention to the importance of thermal gradient-induced isotope fractionation and to its implications in laboratory procedures, stable isotope geochemistry, and the physical chemistry of rarefied gases.

Thermal‐gradient‐induced non‐mass‐dependent isotope fractionation

“…c Data that have been reported in Sun and Bao. [10] a The gas mole ratio of the warmer and the colder compartments at the conclusion of an experiment, based on readings on capacitance monometer and on calculated compartment volume (error at AE10%). b Samples with 'r' in name are from 'small-warm' method.…”

“…The lack of interest in such a system may not be just its limited practical applications, but also the fact that thermal-diffusion-induced isotope fractionation is predicted to be totally massdependent, as can be seen from the convection-free thermal diffusion equations in Jones and Furry [1] or from recent experimental results of the Soret effect in silicate melts [3,9] (Fig. 1).However, our recent report [10] calls attention to the observation that a thermal gradient, often present in routine laboratory procedures, can generate considerable nonmass-dependent isotope fractionation for molecular oxygen (O 2 ). Potential sources of analytical artifacts (e.g.…”

“…Our experiments were carried out in a Pyrex enclosure that has three parts: one small, one bigger, and one for onsite reference gas 10. We had two sets of experiments that differed in geometry: a ‘small‐cold’ and a ‘small‐warm’ method.…”

“…SF 6 measurement was carried out after O 2 measurement to avoid memory effects of S on O 2 . Impurities, such as N 2 or H 2 O, were evaluated and considered to be negligible 10. The importance of measuring an onsite reference for each experiment cannot be over‐emphasized because the onsite reference is not only the ‘zero’ of both colder and warmer samples, but also slightly variable in its isotope composition depending on the way in which the aliquots were taken.…”

“…However, our recent report [10] calls attention to the observation that a thermal gradient, often present in routine laboratory procedures, can generate considerable nonmass-dependent isotope fractionation for molecular oxygen (O 2 ). Potential sources of analytical artifacts (e.g.…”