Search for: All records

Abstract The 3.65 Å phase [MgSi(OH)6] is a hydrous phase that is predicted to be stable in a simplified MgO-SiO2-H2O (MSH) ternary system at pressures exceeding 9 GPa. Along cold subduction zones, it is likely to transport water, bound in its crystalline lattice, into the Earth’s interior. The 3.65 Å phase consists of Mg and Si octahedral sites attached to the hydroxyl group that forms a hydrogen bond and is predicted to undergo pressure-induced symmetrization of the hydrogen bond. Therefore, in this study, we investigate the high-pressure behavior of the 3.65 Å phase using Raman spectroscopy. We have conducted five distinct compressions up to ~60 GPa using two different pressure-transmitting media—alcohol mixture and neon. At ambient conditions, we identified vibrational modes using complementary first-principles simulations based on density functional perturbation theory. Upon compression, we note that the first derivative of the vibrational modes in the lattice region stiffens, i.e., b1lattice > 0. In contrast, the hydroxyl region softens, i.e., b1OH > 0. This is indicative of the strengthening of hydrogen bonding upon compression. We noticed a significant broadening of vibrational modes related to hydroxyl groups that are indicative of proton disorder. However, within the maximum pressures explored in this study, we did not find evidence for pressure-induced symmetrization of the hydrogen bonds. We used the pressure derivative of the vibrational modes to determine the ratio of the bulk moduli and their pressure derivative. We note that the smaller bulk moduli of hydrous phases compared to the major mantle phases are compensated by significantly larger pressure derivatives of the bulk moduli for the hydrous phases. This leads to a significant reduction in the elasticity contrast between hydrous and major mantle phases. Consequently, the detection of the degree of mantle hydration is likely to be challenging at greater depths. 

Kaolinite is formed by weathering of continental crustal rocks and is also found in marine sediments in the tropical region. Kaolinite and other layered hydrous silicate minerals are likely to play a vital role in transporting water into the Earth’s interior via subducting slabs. Recent studies have experimentally documented the expansion of the interlayer region by intercalation of water molecules at high pressures i.e., pressure-induced hydration. This is counter-intuitive since the interlayer region in the layered silicates is quite compressible, so it is important to understand the underlying mechanism that causes intercalation and expansion of the interlayer region. To address this, we explore the high-pressure behavior of natural kaolinite from Keokuk, Iowa. This sample is free of anatase impurities and thus helps to examine both low-energy (0–1200 cm−1) and high-energy hydroxyl (3000–4000 cm−1) regions using Raman spectroscopy and synchrotron-based powder X-ray diffraction. Our results show that the pressure dependence of the hydroxyl modes exhibits discontinuities at ~3 GPa and ~ 6.5 GPa. This is related to the polytypic transformation of Kaolinite from K-1 to K-II and K-II to K-III phase. Several low-energy Raman modes’ pressure dependence also exhibits similar discontinuous behavior. The synchrotron-based powder X-ray diffraction results also indicate discontinuous behavior in the pressure dependence of the unit-cell volume and lattice parameters. The analysis of the bulk and the linear compressibility reveals that kaolinite is extremely anisotropic and is likely to aid its geophysical detectability in subduction zone settings. The K-I to K-II polytypic transition is marked by the snapping of hydrogen bonds, thus at conditions relevant to the Earth’s interior, water molecules intercalate in the interlayer region and stabilize the crystal structure and help form the super-hydrated kaolinite which can transport significantly more water into the Earth’s interior. 

Abstract A plausible origin of the seismically observed mid-lithospheric discontinuity (MLD) in the subcontinental lithosphere is mantle metasomatism. The metasomatized mantle is likely to stabilize hydrous phases such as amphiboles. The existing electrical conductivity data on amphiboles vary significantly. The electrical conductivity of hornblendite is much higher than that of tremolite. Thus, if hornblendite truly represents the amphibole varieties in MLD regions, then it is likely that amphibole will cause high electrical conductivity anomalies at MLD depths. However, this is inconsistent with the magnetotelluric observations across MLD depths. Hence, to better understand this discrepancy in electrical conductivity data of amphiboles and to evaluate whether MLD could be caused by metasomatism, we determined the electrical conductivity of a natural metasomatized rock sample. The metasomatized rock sample consists of ~87% diopside pyroxene, ~9% sodium-bearing tremolite amphibole, and ~3% albite feldspar. We collected the electrical conductivity data at ~3.0 GPa, i.e., the depth relevant to MLD. We also spanned a temperature range between 400 to 1000 K. We found that the electrical conductivity of this metasomatized rock sample increases with temperature. The temperature dependence of the electrical conductivity exhibits two distinct regimes. At low temperatures <700 K, the electrical conductivity is dominated by the conduction in the solid state. At temperatures >775 K, the conductivity increases, and it is likely to be dominated by the conduction of aqueous fluids due to partial dehydration. The main distinction between the current study and the prior studies on the electrical conductivity of amphiboles or amphibole-bearing rocks is the sodium (Na) content in amphiboles of the assemblage. Moreover, it is likely that the higher Na content in amphiboles leads to higher electrical conductivity. Pargasite and edenite amphiboles are the most common amphibole varieties in the metasomatized mantle, and our study on Na-bearing tremolite is the closest analog of these amphiboles. Comparison of the electrical conductivity results with the magnetotelluric observations constrains the amphibole abundance at MLD depths to <1.5%. Such a low-modal proportion of amphiboles could only reduce the seismic shear wave velocity by 0.4–0.5%, which is significantly lower than the observed velocity reduction of 2–6%. Thus, it might be challenging to explain both seismic and magnetotelluric observations at MLD simultaneously.