Ferric iron stabilization at deep magma ocean conditions

Data files

Jan 30, 2026 version files 63.89 KB

MO_redox.zip

51.85 KB
README.md

12.04 KB

Abstract

Fe₂O₃ produced in a deep magma ocean in equilibrium with core-destined alloy sets the early redox budget and atmospheric composition of terrestrial planets. Previous experiments (≤28 GPa) and first-principles calculations indicate that a deep terrestrial magma ocean produces appreciable Fe³⁺, but predict Fe³⁺/ΣFe that conflict by an order of magnitude. We present Fe³⁺/ΣFe of glasses quenched from melts equilibrated with Fe-alloy at 38-71 GPa, 3600-4400 K, analyzed by synchrotron Mössbauer spectroscopy. These indicate Fe³⁺/ΣFe of 0.056-0.112 in a terrestrial magma ocean with mean alloy-silicate equilibration pressures of 28-53 GPa, producing sufficient Fe₂O₃ to account for the modern bulk silicate Earth redox budget and surficial conditions near or more oxidizing than the iron-wüstite buffer, which would stabilize a primitive CO- and H₂O- rich atmosphere.

https://doi.org/10.5061/dryad.612jm64b0

Overview

This dataset contains raw experimental data and analysis scripts supporting the study on the stabilization of ferric iron (Fe³⁺) under deep magma ocean conditions. The research investigates iron redox behavior using laser-heated diamond anvil cell (LH-DAC) experiments combined with synchrotron Mössbauer spectroscopy (SMS) and thermodynamic modeling.

Data & Code Description

The repository contains three primary compressed files:

1. MO_redox.zip

This archive contains Python scripts and associated data files for data analysis, thermodynamic modeling, and figure generation. Key components include:

db/Folder: Contains essential input data files:
- D20_molar.csv: Thermodynamic data parameters from Deng et al. 2020.
  - Terminology:
    - W_Fe: the interaction Margules parameter of components FeO and FeO1.5
    - W_Mg: the difference between the interaction Margules parameter of components FeO_1.5-MgO (W_FeO1.5-MgO) and the interaction Margules parameter of components FeO-MgO (W_FeO-MgO)
    - W_Si: the difference between the interaction Margules parameter of components FeO_1.5-SiO₂ (W_FeO1.5-SiO2) and the interaction Margules parameter of components FeO-SiO₂ (W_FeO-SiO2)
    - W_Al:the difference between the interaction Margules parameter of components FeO_1.5-AlO_1.5 (W_{FeO1.5-AlO1.5}) and the interaction Margules parameter of components FeO-AlO_1.5 (WFeO-AlO1.5)
    - W_Ca:the difference between the interaction Margules parameter of components FeO_1.5-CaO (W_FeO1.5-CaO) and the interaction Margules parameter of components FeO-CaO (W_FeO-CaO)
    - W_Na:the difference between the interaction Margules parameter of components FeO_1.5-NaO_0.5 (W_{FeO1.5-NaO0.5}) and the interaction Margules parameter of components FeO-NaO_0.5 (W_FeO-NaO0.5)
    - W_K:the difference between the interaction Margules parameter of components FeO_1.5-KO_0.5 (W_FeO1.5-KO0.5) and the interaction Margules parameter of components FeO-KO_0.5 (W_FeO-KO0.5)
    - W_Ph:the difference between the interaction Margules parameter of components FeO_1.5-PO_2.5 (W_FeO1.5-PO2.5) and the interaction Margules parameter of components FeO-PO2.5 (W_FeO-PO2.5)
    - W_Ti: the difference between the interaction Margules parameter of components FeO_1.5-TiO₂ (W_FeO1.5-TiO2) and the interaction Margules parameter of components FeO-TiO₂ (W_FeO-TiO2)
- DeltaIW-high_adiabat.csv; DeltaIW-low_adiabat.csv: Calculated oxygen fugacity (fO₂) values along magma ocean high temperature and low temperature adiabats, respectively.
  - Terminology:
    - T(K): the temperature along the adiabat in Kelvin;
    - P(GPa): the pressure along the adiabat in Gpa;
    - IW* : the calculated oxygen fugacity of reaction Fe_liq+1/2O₂=FeO_liq at this pressure and temperature
    - IW : the calculated oxygen fugacity of iron-wustite at this pressure and temperature;
    - IW*-IW: the difference between* IW and IW
    - delta IW: the relative oxygen fugacity of magma ocean to IW
- MoleWeights.csv: Major element chemical compositions of planetary bodies in wt. %.
  - Terminology:
    - Oxide: the oxides name (e.g., SiO₂, FeO);
    - cation (g/mol): the mole weight of oxides in one cation basis (e.g., for Al₂O₃, it might refer to the molar weight per Al cation).
    - oxides (g/mol): the mole weight of the oxide compound itself (e.g., molecular weight ofAl₂O₃).
    - atom (g/mol): the mole weight of cation atom (e.g., atomic weight of Al).
- planets.csv: Major element chemical compositions of planetary bodies in wt. %.
  - Terminology:
    - Group: the planets name.
    - SiO2: Silicon dioxide content (wt%).
    - Al2O3: Aluminum oxide content (wt%).
    - FeO: Iron(II) oxide content (wt%).
    - MgO: Magnesium oxide content (wt%).
    - CaO: Calcium oxide content (wt%).
    - K2O: Potassium oxide content (wt%).
    - Na2O: Sodium oxide content (wt%).
    - TiO2: Titanium dioxide content (wt%).
    - P2O5: Phosphorus pentoxide content (wt%).
    - Total: Total oxide sum (wt%, data quality check).
    - References: Source reference (citation).
- pre_exp.csv: Compilation of relevant previous experimental data.
  - Terminology:
    - No.: Experiment or sample number.
    - REF: Source reference (citation).
    - T_K: Temperature in Kelvin.
    - P_GPa: Pressure in Gigapascals.
    - log fO2: Logarithm of oxygen fugacity.
    - FeRatio__uncorrected_Moss: Fe³⁺/ΣFe ratio resolved from mossbauer spectra directly.
    - FeRatio_c_corrected: Fe³⁺/ΣFe ratio resolved from mossbauer spectra and corrected with recoiless fractions of Fe³⁺ and Fe²⁺.
    - FeRatio_f_Ru_corrected: e³⁺/ΣFe ratio resolved from mossbauer spectra, corrected with recoiless fractions of Fe³⁺ and Fe²⁺, and corrected with Ru effect.
    - Feratio_f: the Fe³⁺/ΣFe ratio used in the model.
    - SiO2: Silicon dioxide content (wt%).
    - TiO2: Titanium dioxide content (wt%).
    - Al2O3: Aluminum oxide content (wt%).
    - Cr2O3: Chromium(III) oxide content (wt%).
    - FeO: Iron(II) oxide content (wt%).
    - MnO: Manganese(II) oxide content (wt%).
    - MgO: Magnesium oxide content (wt%).
    - NiO: Nickel(II) oxide content (wt%).
    - CaO: Calcium oxide content (wt%).
    - Na2O: Sodium oxide content (wt%).
    - K2O: Potassium oxide content (wt%).
    - P2O5: Phosphorus pentoxide content (wt%).
    - RuO2: Ruthenium(IV) oxide content (wt%).
    - PtO: Platinum(II) oxide content (wt%).
    - Total: Total oxide sum (wt%, data quality check).
- Z24.csv: Experimental data from the LH-DAC experiments published in this work.
  - Terminology:
    - SampleName: Experimental number
    - Temp_average: LH-DAC experimental averaged temperature in Kelvin
    - Pressure: LH-DAC experimental pressure in GPa
    - Ratio: Fe³⁺/ΣFe ratio resolved from ESMS directly.
    - Ratio_SS: 1 sigma standard deviation of Fe³⁺/ΣFe ratio resolved from ESMS
    - Syncmoss_C: Fe³⁺/ΣFe ratio resolved from ESMS and corrected with recoilless fraction of Fe³⁺ and Fe²⁺
    - logfO2: the calculated oxygen fugacity of experiments
    - logFe3_Fe2_H22chem: the calculated log(X_Fe3+/X_Fe2+) with the EOS model from Hirschmann (2022) without considering the pressure effect. the X_Fe3+ and X_Fe2+ are the molar percent of Fe³⁺ and Fe²⁺ in the silicate melt, respectively.
    - delta_IW: the relative oxygen fugacity of the experiments to Iron-wustite buffer, which is calculated with Hirschmann (2021).
Analysis Scripts:
- FeRatioCal.py: Calibrates Fe³⁺/ΣFe ratios using different thermodynamic models, including the new model developed in this study.
- PtFe_fO2_func.py: Calculates oxygen fugacity (log fO₂) values for the LH-DAC experiments.
- thermofit_reox_lsfit1array.py: Performs thermodynamic model fitting using the new LH-DAC data (db/Z24.csv) combined with previous experimental data (db/pre_exp.csv). Outputs fitted model parameters to the output/folder (para_ox_lsfit1.json, para_re_lsfit1.json).
- thermofit_result.py: Calibrates Fe³⁺/ΣFe ratios using different thermodynamic models, including the new model developed in this study.
Figure Generation Scripts:
- fig2.py: Generates Figure 2 of the publication.
- fig3_fig4.py: Generates Figures 3 and 4 of the publication.
- Output figures can be saved in the Figures/folder within this archive.
Figures/Folder: The empty folder here is just the directory for generated figures.
output/Folder: Contains the primary results from the thermodynamic fitting script:
- para_ox_lsfit1.json, para_re_lsfit1.json: Fitted parameters for the fit 1 model.
- para_ox_12p5.json, para_re_12p5.json, para_ox_25.json, para_re_25.json, para_ox_FeO.json, para_re_FeO.json: fitted parameters from '12p5', '25', and 'FeO' models of Deng et al. (2020)
- All JSON files share the same variable names:
- ircal.csv: The calculated Fe³⁺/ΣFe ratio through different models of the new LH-DAC experiments published in this paper and previous piston cylinder and multi-anvil experiments published.
  - Terminology:
    - H22: the EOS model from Hirschmann (2022)
    - D20: the "12p5" EOS model from Deng et al. (2020)
    - fit 1: the "fit 1" EOS model from this study
    - fit 2: the "fit 2" EOS model from this study
    - fit 3: the "fit 3" EOS model from this study
    - fit 4: the "fit 4" EOS model from this study
- result_arrays_exp.csv: The calculated Fe³⁺/ΣFe ratio with "fit 1" EOS model from this study at different oxygen fugacity and 4000K.
  - Terminology:
    - Ir_cal_minus05_fit 1_4000: the calculated Fe³⁺/ΣFe ratio at IW-0.5 condition
    - Ir_cal_0_fit 1_4000: the calculated Fe³⁺/ΣFe ratio at IW condition
    - Ir_cal_5_fit 1_4000: the calculated Fe³⁺/ΣFe ratio at IW+0.5 condition
    - Ir_cal_10_fit 1_4000: the calculated Fe³⁺/ΣFe ratio at IW+1.0 condition

2. EOS.zip (Uploaded to Zenodo)

This archive contains Equation of State (EOS) calibration routines, primarily based on Deng et al. (2020, Nature Communications). Includes:

G_cal.py: Python script for Gibbs free energy calibration.
eos_JD.py: A script written by Deng Jie for performing EOS calibration.
mie_grueneisen_debye.py: This file is part of BurnMan, released under the GNU GPL v2 or later. It implements a Mie-Grüneisen-Debye equation of state.
modified_vinet.py: This file is part of BurnMan, released under the GNU GPL v2 or later. It contains a modified Vinet EOS and has been adapted by Deng Jie.

3. MS_raw_data.zip (Uploaded to Zenodo)

This archive contains raw Mössbauer spectroscopy data:

E-SMS/ Folder: Contains raw energy domain Synchrotron Mössbauer Spectroscopy (SMS) data files
- synthesized samples recovered from LH-DAC experiments
- secondary magic glasses standards
- α-Fe foil
- spectra labels (ES1330_Hongluo_spectra labels.xlsx)
CMS/ Folder: Contains Conventional Mössbauer spectra.
- secondary magic glasses standards
- Starting Material (B3)， which is used in the LH-DAC experiments.
- α-Fe foil

Code/Software Requirements

Python: Version 3.10 is required to run the provided scripts.
Python Packages: The scripts rely on common scientific libraries.
Annotations: The Python scripts include basic annotations to explain their functionality.

Reproducibility

The provided scripts (MO_redox.zipand EOS.zip) are designed to reproduce the key data analysis, thermodynamic modeling results, and figures presented in the associated publication:

Unzip MO_redox.zip and EOS.zip, and move the EOS folder into the MO_redox folder.
Ensure required Python packages are installed.
Run the analysis scripts (e.g., PtFe_fO2_func.py, thermofit_reox_lsfit1array.py, FeRatioCal.py) to generate model outputs and calibration results.
Run the figure generation scripts (fig2.py, fig3_fig4.py) to recreate Figures 2, 3, and 4. The figures could be saved in the Figures/directory within MO_redox.

Usage Notes

The raw Mössbauer data (MS_raw_data.zip) is provided for transparency and potential re-analysis.
The db/folder within MO_redox.zipcontains all necessary input data files referenced by the Python scripts. Ensure the scripts can locate these files relative to their own paths.
The output/folder contains the key JSON files resulting from the thermodynamic model fitting procedure (thermofit_reox_lsfit1array.py).

Further questions you can contact zhanghl@cugb.edu.cn.