Nanoporosity in serpentinites and its consequences for fluid transport: a combined multi-scale electron microscopic imaging and molecular dynamics study

Hydration of upper mantle rocks leads to serpentinization with drastic consequences for the geophysical and geochemical properties of the Earth’s lithosphere. Serpentinization takes place via a dissolution-precipitation process in which the fluid phase plays a key role both in the transport of dissolved constituents and in the supply of reactants. A limiting factor for serpentinization is the likelihood of pore-space clogging due to the large solid-volume increase and potential negative repercussions on reaction-induced fracturing [1]. However, small-angle neutron scattering [2] has shown that porosity remains abundant at the nanometer-scale ensuring that aqueous fluids can penetrate to the reaction front allowing serpentinization to progress. To further determine the nature of serpentinite nanoporosity and explore the consequences of fluids confined to nanometric dimensions, we couple multi-scale correlative electron microscopy to molecular dynamics simulations. In the analytical part, we combined electron backscatter diffraction (EBSD) with focused-ion beam scanning electron microscopy (FIB-SEM) nanotomography and transmission electron microscopy (TEM) to investigate two partially serpentinized peridotites from the mid-Atlantic ridge (ODP Leg 209) and the Røragen ultramafic complex, Norway. We determined the crystallographic orientation of the host olivine grains to constrain any potential orientational relationship between the host and the reaction-induced porosity within serpentine. No apparent correlation was found. Based on the EBSD maps, we excavated 23 FIB-SEM cross-sections across serpentine veins and the serpentine-olivine interface. At a pixel resolution of 3 nm, only three out of 23 cross-sections showed apparent pore space. Subsequent, FIB-SEM nanotomography of these three regions showed that vein porosity (average φ_FIB-SEM <50 nm) is concentrated at the olivine-serpentine interface and devoid in the vein middle. To further investigate pore space beyond the FIB-SEM resolution, we prepared eight electron-transparent foils for high-resolution TEM analysis. All foils show an apparent porosity between 1 to 4% with an average pore size of 5 nm. TEM-based energy-dispersive X-ray analysis reveals a 100-nm wide brucite-layer separating serpentine from olivine. Within the brucite-layer, total porosity ranges from 10 to 20% with pore size >10 nm. A higher porosity within brucite-bearing domains is also apparent at the SEM-scale, where we observe larger brucite-rich veins with a high density of nanopores. Hence, our microstructural investigations suggest that continued fluid transport to the reaction interface in the potential absence of reaction-induced fracturing could be sustained through a combination of a nanoporous serpentine network, porous brucite-rich veins, and a highly nanoporous brucite-layer at the olivine reaction interface. Overall, our observations that serpentinite porosity is constrained to the nanoscale have first-order implications for fluid transport and behaviour, because critical physicochemical properties, such as the dielectric constant ε, differ significantly in nanoscale-confined fluids when compared to their bulk counterparts. Initial molecular dynamics simulations of aqueous fluids confined in brucite nanochannels indicate that with the reduction of the width of the nanochannel, the perpendicular component of ε drops drastically which likely has a profound impact on minerals solubility hence overall reaction progress.

[1] Plümper et al. Geology (2012) 40(12): 1103-1106.

[2] Tutolo et al. Geology (2016) 44(2): 103-106.