A new reservoir with extremely light Si isotopes formed by rodingitization process in the altered oceanic lithosphere

Supplementary Materials 1 Supplementary computational methods 1.1 Monte Carlo estimation of Si loss during rodingitization 1.2 Inversion of mineral modal abundances using mass-balance methods 1.3 Monte Carlo isotope mass balance for inferred fluid δ30Si and apparent rock–fluid fractionation 1.4 Computational implementation and reproducibility 2 Supplementary Figures: Figs. S1-S5 Fig. S1. Diagram of δ30Si vs. δ29Si of the studied samples and reference materials Fig. S2. Sensitivity of estimated Si loss to alternative Monte Carlo model parameterizations Fig. S3. Mineral modal proportions reconstructed by constrained mass-balance inversion Fig. S4. Equilibrium Si isotope fractionation between albite and H3SiO40, H4SiO4− or fluid phases Fig. S5. Monte Carlo constraints on apparent rock–fluid Si isotope fractionation factor and the isotopic composition of inferred Si-bearing fluids 3 Supplementary Tables: Tables S1-S10 Table S1. Compiled petrographic, geochemical, and isotopic data for global rodingites and metarodingites. Table S2. Complete parameter settings for the strict, preferred, and permissive Monte Carlo Si-loss models. Table S3. Complete parameter settings for the constrained mineral mass-balance inversion. Table S4. Complete parameter settings for the Monte Carlo Si isotope mass-balance and population-level KDE calculations. Table S5. Silicon isotope compositions of Luqu gabbros, rodingites, and reference materials. Table S6. Compiled published silicon isotope data used for comparison with the Luqu samples. Table S7. Sample-specific Si-loss estimates and Monte Carlo diagnostics for the strict, preferred, and permissive models. Table S8. Mineral mass fractions reconstructed using the preferred five-oxide and three-oxide constrained inversion. Table S9. Equilibrium silicon isotope fractionation equations and calculated albite–fluid fractionation factors. Table S10. Sample-specific and population-level estimates of exported-fluid δ30Si and apparent rock–fluid fractionation factors. Code: 1. Fluid Si isotope calculation 2. Mineral Percentage calculation 3. Si_loss calculation