Final Report Second DE project Project Update uri icon

DCO ID 11121/9636-8134-8493-3798-CC

Date Submitted

  • 2015-09-15

Update Text

  • Because the environments within Earth’s crust where abiotic synthesis of organic compounds is thought to occur are largely inaccessible to direct sampling, laboratory simulations and theoretical models are a critical component in understanding the origin of these compounds. Accordingly, a suite of experimental and theoretical studies were conducted to investigate the reactions responsible for abiotic organic synthesis within Earth’s crust.


    A primary focus of the effort by Tom McCollom (University of Colorado) over the past two years has been investigation of abiotic reduction of carbon during laboratory simulations of serpentinization of ultramafic rocks.  One aspect of this work has involved characterization of solid carbonaceous particles found in several laboratory experiments to evaluate whether they formed by abiotic reduction of carbon during the experiments.  This work involved a new collaboration with several groups in France facilitated through the Deep Energy directorate, including researchers at the University of Lyon (Daniel, Andreani), IPGP (Martinez), and the Institut de Minéralogie, de Physique des Matériaux et de Cosmochimie in Paris (Bernard).  Solid carbonaceous products were characterized by Raman spectroscopy, synchrotron measurements including STXM and carbon-XANES, and SEM.  While these analyses were initially promising, subsequent evaluation of carbon isotopic composition of the carbonaceous particles by nano-SIMS suggests that they formed from trace organic contaminants in the reactants rather than from reduction of inorganic carbon.  During the next phase of the Deep Energy project, the McCollom effort will continue to explore for evidence of the formation of high-molecular-weight carbon products in laboratory experiments.

    A second aspect of this work has been evaluation of pathways for abiotic formation of molecular hydrogen (H2) and methane (CH4) during serpentinization.  Several recent reports have claimed abiotic production of methane from reduction of CO2 during experimental serpentinization of ultramafic rocks and minerals.  However, our results using isotopically labeled compounds to track the source of carbon demonstrated that the methane formed in serpentinization experiments at 200-300°C was derived predominantly from background sources present in the reactants rather than from reduction of dissolved CO2.  In addition, we performed a series of water-rock reaction experiments at 90°C using olivine and harzburgite as reactants to investigate formation of H2 and CH4 at low temperatures.   Although these experiments produced H2 and CH4 at levels comparable to those in several other recent studies performed at 100°C or lower, comparison with blank experiments showed that these compounds were derived predominantly from the rubber stoppers that are commonly used as part of the reaction vessel in low temperature experiments.  Overall, our results demonstrate that the production of methane during experimental serpentinization is much more limited than other recent studies have suggested.  This result is consistent with recent results indicating that the methane found in deep-sea hydrothermal vent fluids is derived from high-temperature re-equilibration of magmatic gases rather than fluid-rock interactions (McDermott et al., PNAS, 2015)

    Lastly, the University of Colorado team completed an analysis of extractable organic compounds in hydrothermal fluids from mafic- and ultramafic-hosted deep-sea hydrothermal systems along the Mid-Atlantic Ridge.  In contrast to other reports, McCollom did not find any evidence for non-volatile hydrocarbons with an abiotic origin in any of the hydrothermal fluid samples studied.  However, he did observe trace amounts of polycyclic aromatic hydrocarbons in the highest temperature fluids that may contribute to the recalcitrant dissolved organic matter in the deep ocean, and a number of carboxylic acids and other compounds of clear biological origin were observed at the moderate temperature, serpentine-hosted Lost City system. 

    In a related study, the group at the University of Lyon led by Prof. Isabelle Daniel investigated the role of aluminum on the serpentinization rate at 200° and 300°C, 2kbars in low-pressure diamond anvil cell show the following results and are the core of the PhD work of Maria Pens (graduate student from Venezuela, supported by the FUNDAYACUCHO in Venezuela). This work was allocated 2 weeks of beamtine at the ESRF, the European Synchrotron radiation facility and one week at SOLEIL, the French synchrotron facility. They confirmed the fact that Al strongly increases the serpentinization rate of olivine, as previously observed, and this time they determined the kinetic parameters of the reactions. They assessed this effect over a wide range of pH and propose that this is due to Al-Si complexes formation on the surface of olivine, possible from pH 4 to 10. It was shown that Al plays the opposite role on orthopyroxene, the other main mineral of ultramafic rocks, at least at 350°C, possibly because the charge surface of pyroxene differs from that of olivine during the first stages of dissolution. This was particularly surprising since available data on olivine and pyroxene serpentinization under similar P-T conditions in pure water report faster rate for orthopyroxene serpentinization than for olivine. The other surprising result was the fact that their data demonstrated a general increase of olivine serpentinization rate with pH in presence of Al, while all available data on olivine dissolution display the opposite, i.e. a decreasing rate with pH. This effect of pH is also observed in the series of experiments that were run using HCO3- -enriched solution that displayed fast reaction rates, the fastest observed being when both Al and alkaline conditions are gathered. These results are currently being written in two papers that will be submitted within the next months, while Maria will be defending her thesis within the next year.

    They are currently investigating the effect of Al ± HCO3- at lower pressure conditions using large volume reactors that also allow fluid sampling for liquid and gas analyses. This work is being conducted by Steve Peuble, a DCO post-doctoral fellow and has benefitted from the acquisition of a dedicated and optimized micro-GC supported by the labex LIO. Steve will continue the work for another year under the auspices of the ANR grant deepOASES to B. Menez. A first series of experiments has been realized using pure olivine powder reacting at 200-250°C and 200-250 bars with Al- and HCO3-- enriched saline water of initial neutral pH. Blank experiments (with same solution ± inert solids) were realized at the same time to test for possible carbon contamination. After a month, 4-5% of serpentinization has been observed, which is much lower than expected. These preliminary results already suggest that the Al effect is lower at lower P. However, after 4 days, first detections of H2 and CH4 are observed, and ethane is detected after 20 days. Their concentrations are much higher than the background ones measured in blanks. The H2 content tends to decrease rapidly while CH4 one increases. Solid product characterization confirms the formation of serpentine and magnetite plus a carbon-rich material locally that is under identification. These first results are extremely encouraging since they validate the possible reduction bicarbonate ions to both solid and gas carbon compounds during serpentinization.

    Concerning related projects, M. Andreani recently organized the ECORD-DCO “Rainbow workshop” in Lyon, 10th-12th June 2015 ( that aimed at preparing a possible IODP drilling proposal on the Rainbow massif that hosts the ultramafic hosted hydrothermal field Rainbow. A summary of the workshop will be send soon to the DCO and ECORD, and it will be also highlighted in the ECORD Newsletter.

    Catalytic Reactions

    A second major experimental activity in the second Deep Energy project was to explore the constraints on redox reactions and carbon transfer. The group at IPGP investigated the catalytic properties of a single phase. A first set of experiments was conducted on magnetite (Fe3O4) following the established protocol typically by the chemistry group (Poitiers). In these experiments, 1 g of finely divided magnetite and 20 g of ultrapure water were charged into a stainless steel reactor. The experiment was run at 80 bar (by means of 40 bar CO2 and 40 bar H2) and heated at 180 ° C for 72 hours under continuous magnetic stirring. A sampling valve allowed to monitor the composition of the gas mixture, through gas chromatography. The presence of methanol (CH3OH) was detected after 72 h. A blank experiment (without magnetite) showed that methanol was not produced by simple reactivity of CO2-H2 mixture at high pressure and temperature. Several unidentified products were formed and their identification is currently underway. TEM observation on the solid products did not show any C phases produced during the experiment. 

    In parallel, two sets of experiments were performed at IPGP by mixing silicates synthesized in finely divided form (fayalite (Fe2SiO4) and olivine (FeMgSiO4)), with catalytic phases present in natural systems. In this case, the reactants were loaded into gold capsules together with a sodium bicarbonate solution. The capsules are then reacted at 200 °C and 200 bar for 3 weeks. After the experiment the capsule is first pierced in a helium flushed syringe to recover the gas products, which are then injected in a GC for analysis. Finally the capsule is opened and the recovered solid studied by conventional microscopy techniques. They did not detect any gas phase in the capsule, probably because there was not enough gas for GC analysis. Next experiment will be run at 300°C which should enhance the production of hydrogen and methane. The solid phases showed interesting reactions: run products are mainly phyllosilicate (talc still to be confirmed by DRX), magnesium carbonates (derived from carbonation of olivine), and magnetite. Furthermore, in experiments where sulfides are present (Cu2S and FeS2), they observed large carbon-rich phases (several microns in size), sometimes associated with magnesite which then show smaller grain sizes. These phases were not detected in the reaction products where the magnetite was present.   While it remains to be examined in more details, it seems that sulfides are more effective in the CO2 reduction reaction than magnetite and chromite.

    A complementary set of experiments were conducted by Jeff Seewald at WHOI to determine how potential catalytic phases might influence the isotopic exchange of reduced species. In the first, a laboratory experiment was designed to elucidate mechanisms of hydrogen exchange between aqueous methane and water at elevated temperatures and pressures with the goal of understanding factors that regulate the clumping of carbon and hydrogen isotopes in methane. Clumped isotopes represent a potentially powerful tool to examine the formation temperature for methane in subsurface environments, provided thermodynamic isotopic equilibrium is attained. The experiment he conducted was designed to test the hypothesis that sulfur and aromatic hydrocarbon radicals will catalyze the exchange of hydrogen between methane and water and facilitate attainment of isotopic equilibrium. A U.S. Gulf Coast crude oil with added methane was heated at 325°C and 350 bar in the presence of D2O and a pyrite-pyrrhotite-magnetite redox buffer that also served as a source of sulfur. Mass spectrometric analysis of the methane during the course of the experiment revealed that deuterium was not significantly incorporated into methane, suggesting that hydrogen/deuterium exchange with D2O was not facilitated by the presence of aromatic radicals from the oil and/or sulfur radicals form the redox buffer. He will further test our hypothesis in a second experiment of this type conducted under more reducing conditions with greater oil and sulfur abundances.

    A second experiment currently underway at WHOI is being conducted to assess the extent to which pH and temperature influence the catalytic activity of Fe- and Cr-bearing minerals at near critical hydrothermal conditions. Abiotic formation of low molecular weight hydrocarbons has been shown to occur in the presence Fe- and Cr-bearing minerals, but the extent of their catalytic activity is presently unclear due to the absence of mineral-free control experiments. The experiment he is conducting examines the polymerization of methane under hydrothermal conditions in the absence of added minerals and constrain the potential for abiotic synthesis in a pure aqueous system.

    In addition to the experiments described above, Jeff has been working with John Eiler at Caltech and Shuhei Ono at MIT to provide hydrothermal fluid samples for measurement of the rare isotopologues of methane. Seewald carefully assessed our large collection of methane-rich hydrothermal fluids from peridotite influenced hydrothermal systems and distributed selected samples that are optimally suited for clumped isotopic analysis. These data will provide some of the first clumped methane isotopic data for methane at a sediment-deficient mid-ocean ridge system and constrain temperatures of formation. Results will also allow comparison of two mass spectrometric and spectroscopic approaches for the measurement of clumped methane isotopes.

    Behavior of Reduced Carbon Species in Porous Matrices

    The goal of this portion of the project was to employ molecular dynamics approaches to understand the structure and dynamics of fluids containing hydrocarbons, water, and oxygenated compounds confined within narrow pores that could be found in sub-surface formations. The theoretical results obtained by Prof. Alberto Striolo and his students are compared systematically to experimental data obtained by collaborator Prof. David Cole of the Ohio State University, and his team. The distinctive feature of this project is the inclusion of both organic and aqueous fluids. Additionally, access to the DCO Cluster greatly facilitated the molecular dynamics simulations for some of the systems investigated.

    Building on the detailed summary of the behavior of hydrocarbons in nanoporous systems summarized in Chapter 19 in the 2013 Carbon in Earth MSA volume, the Cole-Striolo team have systematically explored the effects of nanoconfinement on the structure and dynamics of assorted alkanes as pure fluids as mixtures with either water or CO2. Thus far, the team has conducted a number of different kinds of experiments focusing on the behavior of propane, ethane, methane, carbon dioxide or their mixtures with or without water present interacting with different types of mesoporous matrices (e.g., SiO2, Al2O3, TiO2, ZrO2). High temperature-high pressure gravimetric measurements on these fluids have revealed profound fluid densification in nanopores as the density (pressure) approaches that of the bulk critical density followed by a dramatic density decrease (fluid depletion). Densification of propane in model silica pores has been observed in MD simulations, in qualitative agreement with experiments. Quasielastic neutron scattering (QENS) experiments have been conducted at Oak Ridge National Laboratory on the system propane-mesoporous silica with or without CO2 present. Results from these experiments are interpreted in terms of translational “diffusive” motion of propane, residence times between diffusion jumps and the jump distances. Interestingly, the presence of CO2 seems to enhance the mobility of propane in the mesopores.

    To establish a close connection with the experimental data obtained by Cole and coworkers, Striolo’s team simulated the adsorption of propane in slit-shaped silica pores. The results, in qualitative agreement with the experiments, show a maximum in the density of the confined fluid that occurs at conditions approaching the gas-to-liquid transition of the bulk fluid. Mixed water-methane and water-ethanol systems have also been investigated using MD simulations, suggesting that preferential adsorption might be responsible for the results observed experimentally in mixed systems. Using MD the team interrogated mixtures of water and methane confined within narrow pores carved out of silica, considered a proxy for cristobalite and other rocks. The pores were slit-shaped and of width ~ 1nm. The results suggested that the solubility of methane in water within the confined space can be up to 1 order of magnitude larger than that in the bulk. The results show that the structure of confined water, template by the solid surface, is responsible for this enhanced methane solubility. The structure of water molecules surrounding one methane molecule is reminiscent of that of hydrates. These results have been recently confirmed by high pressure-high temperature Magic Angle Spinning (MAS) NMR experiments conducted at the Environmental Molecular Science Laboratory.