Experimental Constraints on Fluid-Rock Reactions during Incipient Serpentinization of Harzburgite

Friday, 19 December 2014: 11:20 AM
Frieder Klein1, Niya G Grozeva2, Jeffrey Seewald1, Thomas M McCollom3, Susan E Humphris4, Bruce M Moskowitz5, Thelma Souza Berquo6 and Wolf-Achim Kahl7, (1)WHOI, Woods Hole, MA, United States, (2)Massachusetts Institute of Technology, Cambridge, MA, United States, (3)Univ Colorado, Boulder, CO, United States, (4)Woods Hole Oceanographic Inst, Woods Hole, MA, United States, (5)University of Minnesota, Institute for Rock Magnetism, Minneapolis, MN, United States, (6)Concordia College, Moorhead, MN, United States, (7)University of Bremen, Bremen, Germany
The exposure of mantle peridotite to water at crustal levels leads to a cascade of interconnected dissolution-precipitation and reduction-oxidation reactions - a process referred to as serpentinization. These reactions have major implications for microbial life through the provision of hydrogen (H2). To simulate incipient serpentinization and the release of H2 under well-constrained conditions, we reacted uncrushed harzburgite with chemically modified seawater at 300°C and 35 MPa for ca. 1.5 years (13441 hours), monitored changes in fluid chemistry over time, and examined the secondary mineralogy at the termination of the experiment. Approximately 4 mol % of the protolith underwent alteration forming serpentine, accessory magnetite, chlorite, and traces of calcite and heazlewoodite. Alteration textures bear remarkable similarities to those found in partially serpentinized abyssal peridotites. Neither brucite nor talc precipitated during the experiment. Given that the starting material contained ~3.8 times more olivine than orthopyroxene on a molar basis, mass balance requires that dissolution of orthopyroxene was significantly faster than dissolution of olivine. However, the H2 release rate was not uniform, slowing from ~2 nmol H2(aq) gperidotite-1 s-1 at the beginning of the experiment to ~0.2 nmol H2(aq) gperidotite-1 s-1 at its termination. Serpentinization consumed water but did not release significant amounts of dissolved species (other than H2) suggesting that incipient hydration reactions involved a volume increase of ~40%. The reduced access of water to olivine surfaces due to filling of fractures and coating of primary minerals with alteration products led to decreased rates of serpentinization and H2 release. While this concept might seem at odds with completely serpentinized seafloor peridotites, reaction-driven fracturing offers an intriguing solution to the seemingly self-limiting nature of serpentinization. Indeed, the reacted sample revealed a number of textural features diagnostic of incipient reaction-driven fracturing. Reaction-driven and tectonic fracturing must have far reaching impacts on the release rate of H2 in peridotite-hosted hydrothermal systems and therefore represent key mechanisms in regulating the supply of reduced gases to microbial ecosystems.