Extended Data Fig. 7: Water content calibration line and reproducibility of hydrogen isotope analysis established on apatite and silicate glass standards.

a, A water calibration curve was established using the following standards: two apatite (Kovdor and Durango), three basaltic glasses (ALV-1833-11, ALV-519-4−1, and SWIFT MORB glass), and two anhydrous background (San Carlos olivine and Suprasil 3001). The datasets are listed in Supplementary Data Table 4. The analytical uncertainties are 0.42 % and 0.54 % (1σ) for apatite and glasses, respectively. b, Reproducibility of the hydrogen isotope analysis of the standards and the CE6 apatite and melt inclusions over the analytical session. The reproducibility of δD analysis throughout the whole analytical sessions was ±54 ‰ (2σ), estimated on the Kovdor apatite standard. The average δD values measured for the SWIFT MORB glass and Durango apatite are −98 ± 70 ‰ and −97 ± 100 ‰, respectively, consistent with their recommended values (−73‰ and −120‰)16,55 within analytical errors. The average δD values measured for the basaltic glass standards AVL−1833-11 and ALV-519-4−1 are −101 ± 60 ‰ and −94 ± 58 ‰, consistent with previous measured values of −101 ± 86 ‰ and −132 ± 158 ‰, in ref. 9. Data are listed in Extended Data Tables 2, 3 and Supplementary Data Tables 4, 5.