An offline sample preparation system and water exchange reaction method for the measurement of δ2H of non-exchangeable hydrogen in organic matter

The measurement of δ²H of non-exchangeable-H (δ²H_n) in organic matter (OM) by isotope-ratio mass spectrometry is often hampered by the difficulties in controlling H isotope exchange of the exchangeable H fraction (f_ex) and removal of residual moisture. The determination of δ²H_n in organic matter requires control of the isotopic composition of f_ex. This can be achieved through dual water H isotope exchange experiments. However, these experiments are laborious, sensitive to the method used (e.g. prior sample treatment, temperature, time) and are costly. This has resulted in a wide range of reported f_ex for known isotopic references. Moreover, it is not always clear that samples are completely dry following the H exchange experiments, leading to even larger variations. The δ²H_n data is typically used in ecological studies for source attribution due to the large observed separation between contributing end members. However, it is not clear to what degree the analytical errors in δ²H determined by incomplete H isotope exchange of f_ex and the residual moisture impact the source attribution. Here we developed a simple offline sample preparation system, the Isobox, and a protocol for the measurement of δ²H_n in natural OM as well as pure organic compounds. We performed dual water H isotopic exchange experiments with both liquid and vapor water at near 0 and 105°C respectively. We analyzed three keratin reference materials (KHS, CBS and USGS42), two amino acids (Isoleucine and Threonine), along with caffeine (USGS62) and polyethylene (IAEA-CH-7) as drying references. We have also used natural samples of demineralized soil and green algae to create known mixtures to test these methods and their analytical uncertainty impact on the source attribution. We show that the liquid water exchange experiments led to f_ex close to the theoretical expectations for both keratin and pure compounds. Depending on the research question careful determination with controlled dual water procedure for the determination of δ²H_n may be required. However, simple sample treatment with exposure to a single isotopically known water can be used to derive δ²H for source attribution. The offline sample preparation system and equilibration method we developed is simple, accurate and cost effective and can be implemented in virtually any laboratory for the analysis of a wider range of OM types.