Oxygen isotopic fractionation of O 2 during adsorption and desorption processes using molecular sieve at low temperatures

RATIONALE Cryogenic trapping using molecular sieves is commonly used to collect O 2 extracted from silicates for 17 O/ 16 O and 18 O/ 16 O analyses. However, gases which interfere with 17 O/ 16 O analysis, notably NF 3 , are also trapped and their removal is essential for accurate direct measurement of the 17 O/ 16 O ratio. It is also necessary to identify and quantify any isotopic fractionation associated with the use of cryogenic trapping using molecular sieves. METHODS The oxygen isotopic compositions of O 2 before and after desorption from, and adsorption onto, 13X and 5A molecular sieves (MS13X and MS5A) at 0°C, −78°C, −114°C, and −130°C were measured in order to determine the oxygen isotopic fractionation at these temperatures. We also investigated whether isotopic fractionation occurred when O 2 gas was transferred sequentially into a second cold finger, also containing molecular sieve. RESULTS It was confirmed that significant oxygen isotopic fractionation occurs between the gaseous O 2 and that adsorbed onto molecular sieve, if desorption and adsorption are incomplete. As the fraction of released or untrapped O 2 becomes smaller with decreasing trapping temperature (from 0 to –130°C), the isotopic fractionation becomes larger. Approximately half of the total adsorbed O 2 is released from the molecular sieve during desorption at –114°C, which is the temperature recommended for separation from NF 3 (retained on the molecular sieve), and this will interfere with 17 O/ 16 O measurements. CONCLUSIONS The use of a single cold finger should be avoided, because partial desorption is accompanied by oxygen isotopic fractionation, thereby resulting in inaccurate isotopic data. The use of a dual cold finger arrangement is recommended because, as we have confirmed, the transfer of O 2 from the first trap to the second is almost 100%. However, even under these conditions, a small isotopic fractionation (0.18 ± 0.05‰ in δ 17 O values and 0.26 ± 0.06‰ in δ 18 O values) occurred, with O 2 in the second trap being isotopically enriched in the heavier isotopes. Copyright © 2014 John Wiley & Sons, Ltd.