Abiotic methane formation during experimental serpentinization of olivine

Abstract

Fluids circulating through actively serpentinizing systems are often highly enriched in methane (CH₄). In many cases, the CH₄ in these fluids is thought to derive from abiotic reduction of inorganic carbon, but the conditions under which this process can occur in natural systems remain unclear. In recent years, several studies have reported abiotic formation of CH₄ during experimental serpentinization of olivine at temperatures at or below 200 °C. However, these results seem to contradict studies conducted at higher temperatures (300 °C to 400 °C), where substantial kinetic barriers to CH₄ synthesis have been observed. Here, the potential for abiotic formation of CH₄ from dissolved inorganic carbon during olivine serpentinization is reevaluated in a series of laboratory experiments conducted at 200 °C to 320 °C. A ¹³C-labeled inorganic carbon source was used to unambiguously determine the origin of CH₄ generated in the experiments. Consistent with previous high-temperature studies, the results indicate that abiotic formation of CH₄ from reduction of dissolved inorganic carbon during the experiments is extremely limited, with nearly all of the observed CH₄ derived from background sources. The results indicate that the potential for abiotic synthesis of CH₄ in low-temperature serpentinizing environments may be much more limited than some recent studies have suggested. However, more extensive production of CH₄ was observed in one experiment performed under conditions that allowed an H₂-rich vapor phase to form, suggesting that shallow serpentinization environments where a separate gas phase is present may be more favorable for abiotic synthesis of CH₄.

Keywords: abiotic methane; hydrothermal systems; serpentinization.

Fig. S1.

Back-scattered electron images and energy-dispersive spectrometer (EDS) analyses showing apparent occurrences of native metals in experiment OLIVOPX230. (A) Overview of reaction products in polished grain mount. (B) Expanded view from A, with EDS of the area centered on the bright particle outlined by the red circle shown in C. The EDS is dominated by Ni and Fe with little O, indicating that the small bright particles are composed of an Ni-Fe alloy. The Mg and Si in the EDS data are likely from serpentine surrounding the alloy. (E) EDS of area outlined by the red circle in D. The prominent Pt peak and lack of significant peaks for S or O indicate that the bright particle in the red circle in D is native Pt. Brc, brucite; Mgt, magnetite; Ol, olivine; SRP, serpentine.

Fig. 1.

Measured concentrations of dissolved volatiles during serpentinization experiments. (A and B) Dissolved H₂. (C and D) Dissolved CH₄. For experiments OLIV230 and OLIVOPX230, additional fluid was injected into the reaction cells during heating to replace fluid removed for chemical analysis, resulting in a transient decrease in dissolved concentrations. Except where shown, estimated analytical errors are smaller than the size of the symbols. Methane was not measured at all sample points to conserve fluid. The measured concentrations are provided in Table S1, with additional data on fluids compositions given in ref. .

Fig. S2.

Isopleths of constant mole percentage of H₂ in the H₂-H₂O system as a function of pressure and temperature. The red X denotes the P–T conditions of experiment OLIV300_sat2. Extrapolation of the isopleths indicates that the experiment is within the two-phase field for systems containing about 0.3 mol % H₂, indicating that the fluid should become saturated and exsolve a separate H₂-rich vapor phase at aqueous concentrations around 170 mmol kg⁻¹. It should be noted, however, that the isopleths in the P–T region of the experiment as portrayed by Seward and Franck (31) are defined by only a few measurements, and therefore, there is substantial uncertainty in the position of the isopleths. Furthermore, the diagram is drawn for pure water and H₂, and the presence of dissolved salts and additional volatile species (CO₂ and CH₄) will cause the isopleths to deviate somewhat from the pure system. Modified from ref. .

Fig. 2.

Representative mass fragmentograms for CH₄ generated during serpentinization experiments. Also shown for reference is analysis of a methane gas standard. For the reference standard, prominent peaks occur at m/z 15 and 16, corresponding to m/z values for the ions ¹²CH₃⁺ and ¹²CH₄⁺, respectively. The small peak at m/z 17 in the standard reflects a small amount of ¹³CH₄⁺ (∼1 mol % at natural abundance). Higher amounts of m/z 17 in some experiment samples indicate contributions from ¹³CH₄ formed through reduction of dissolved inorganic carbon [¹³CO_2(aq) or H¹³CO₃⁻] during the experiments.

Fig. S3.

Calculated equilibrium CH_4(aq)/CO_2(aq) ratios (solid lines) as a function of the concentration of H_2(aq) at the experimental temperatures and 35 MPa, with measured fluid compositions at the end of the serpentinization experiments shown for comparison (symbols). With increasing temperature, the position of the line shifts to the right, reflecting the increased thermodynamic stability of CO₂ relative to CH₄ at higher temperatures. In all cases except OLIV320, equilibrium CH_4(aq)/CO_2(aq) ratios at the experimentally measured H_2(aq) values are >10³, indicating there is a strong thermodynamic drive for the conversion of dissolved inorganic carbon to CH₄ to approach thermodynamic equilibrium. Experiment OLIV320 is close to equilibrium, indicating little or no thermodynamic drive for formation of CH₄. Thermodynamic data for the equilibrium calculations were obtained using SUPCRT92 (38) with parameters for dissolved compounds from Shock et al. (39) and Shock and Helgeson (40). For the purposes of the calculations, activity coefficients of all dissolved species were assumed to be equal to one. The concentration of dissolved CO_2(aq) for the experiments was calculated from the measured ΣCO₂ values at the estimated in situ pH using speciation calculations performed with the program EQ3 (29).