November 17 @ 12:00 pm - 1:00 pm CST
Departmental Research in Earth, Environmental and Planetary Sciences
Graduate Interdisciplinary Earth Science Symposia (GIESS)
By Proteek Chowdhury, Damanveer Grewal, and Jia Shi
Chowdhury Abstract: Effect of sulfate on the liquidus and sulfur concentration at anhydrite saturation (SCAS) of hydrous basalt at subduction zones
Sulfur (S) as sulfide minerals, melts, and as S2- species in silicate melts is prevalent in many different tectono-magmatic settings in Earth. Yet, S as anhydrite or as SO42- species in fluids and melts is thought to be relevant for subduction zones, where the presence of sulfate over sulfide is argued to play a key role in processes such as mobility of chalcophile element, oxidation of mantle and mantle-derived magmas, and release of excess S-rich gases. However, it remains unclear what role the slab-released SO42-, dissolved in fluids or melts plays in magma genesis in sub-arc mantle. Furthermore, although oxidized arc magma is thought to transport SO42- from mantle to volcanic arc crust and atmosphere, the SO42- carrying capacity of arc basalts at mantle conditions are unknown as the existing S concentration at anhydrite saturation (SCAS) experiments are restricted to 1 GPa and mostly on felsic compositions.
We performed piston-cylinder experiments in Au-Pd capsules at 1-3 GPa and 1000-1325 °C to investigate (a) the effect of variable dissolved SO42- (0-2 wt.% S) on the liquidus of a primary hydrous arc basalt with ~4 wt.% H2O and (b) the SCAS of hydrous mafic magmas. Dissolved SO42- in the silicate melt was confirmed by S Ka X-ray peak position using electron microprobe. S-free hydrous liquidus of cpx at 2 GPa is ~25 °C hotter than the liquidus with ~0.1 wt.% S as SO42- and the liquidus depression with further S enrichment to anhydrite saturation (~2 wt.% S) can be fitted by an empirical power function. Experiments on more mafic compositions show that SCAS increases with increasing temperature and CaO and decreases with SiO2. Calculations using a new SCAS model, fitted with our new data and previous experiments, and assuming 150-550 ppm S in the arc mantle show that <10% melting would exhaust anhydrite, if present. The S content as SO42- of hydrous arc basalts produced by 10-20% melting will be 500-4000 ppm, which is comparable to the melt inclusion S contents from various arcs. The SO42- undersaturated basalts may assimilate crustal sulfate and lead to high observed SO2 flux.
Grewal Abstract: Simultaneous alloy-silicate fractionation of carbon, nitrogen, and sulfur at high pressures and temperatures: Implications for establishing the volatile budget of the Earth
Constraining the origin, distribution and evolution of volatiles such as carbon (C), nitrogen (N) and sulfur (S) in terrestrial planets is essential to understand planetary differentiation, habitability and comparative planetology. C/N ratio of Bulk Silicate Earth (BSE) is superchondritic (40 8), while C/S ratio is nearly chondritic (0.49 0.14). Accretion, core formation, and magma ocean (MO) crystallization are the key processes that could have set the relative budgets of C, N and S in different planetary reservoirs. However, experiments using either C-N or C-S-bearing systems have shown that C is more siderophile than N and S, consequently core formation would have left behind subchondritic C/N and C/S ratios in BSE. Accretion of extremely C-rich bodies during core formation or/and as a late veneer along with an early atmospheric blow-off are amongst the scenarios that have been suggested to explain C/N ratio while the addition of a differentiated body with a C-rich mantle has been suggested to explain C/S ratio in BSE. However, no internally consistent explanations exist on the origin of all the volatile elements.
We performed piston cylinder and multi-anvil experiments, using Fe-Ni-N-C±S alloy with variable amounts of S and mafic-ultramafic silicate mixtures in graphite saturated conditions at 1-7 GPa, 1600-1800 C, and fO2 ranging from IW of -1.1 to -0.3. EPMA and SIMS were used to determine major elements and volatile abundances in the coexisting alloy and silicate melt phases, while the speciation of the volatiles was determined using Raman spectroscopy. Our experimental data reveals that C becomes less siderophile in the presence of N and S during core-mantle differentiation involving an S-rich alloy. Using a set of inverse Monte-Carlo simulations, we propose that a disequilibrium merger of a Mars-sized planetary embryo with a C-saturated, S-rich core to a volatile-depleted proto-Earth during the main stage of accretion could have simultaneously satisfied C-N-S abundances and ratios in BSE along with setting up the stage of for the presence of NH3 and HCN in the Earth’s early atmosphere via MO degassing.
Shi Abstract: On the parallel computations of Earth’s normal modes via novel Lanczos algorithms
We present a novel parallel computational approach to compute Earth’s normal modes. We discretize Earth via an unstructured tetrahedral mesh and apply the continuous Galerkin finite element method to the elasto-gravitational system. To resolve the eigenvalue pollution issue, following the analysis separating the seismic point spectrum, we utilize explicitly a representation of the displacement for describing the oscillations of the non-seismic modes in the fluid outer core. Effectively, we separate out the essential spectrum which is naturally related to the Brunt-Väisälä frequency. We introduce two Lanczos approaches with polynomial and rational filtering for solving this generalized eigenvalue problem in prescribed intervals. The polynomial filtering technique only accesses the matrix pair through matrix-vector products and is an ideal candidate for solving three-dimensional large-scale eigenvalue problems. The matrix-free scheme allows us to deal with fluid separation and self-gravitation in an efficient way, while the standard shift-and-invert method typically needs an explicit shifted matrix and its factorization. The rational filtering method converges much faster than the standard shift-and-invert procedure when computing all the eigenvalues inside an interval. Both two Lanczos approaches solve for the internal eigenvalues extremely accurately, comparing with the standard eigensolver. In our computational experiments, we compare our results with the radial earth model benchmark, and visualize the normal modes using vector plots to illustrate the properties of the displacements in different modes.