Philip Skemer

Modeling olivine CPO evolution with complex deformation histories—Implications for the interpretation of seismic anisotropy in the mantle

Relating seismic anisotropy to mantle flow requires detailed understanding of the development and evolution of olivine crystallographic preferred orientation (CPO). Recent experimental and field studies have shown that olivine CPO evolution depends strongly on the integrated deformation history, which may lead to differences in how the corresponding seismic anisotropy should be interpreted. In this study, two widely used numerical models for CPO evolution—D-Rex and VPSC—are evaluated to further examine the effect of deformation history on olivine texture and seismic anisotropy. Building on previous experimental work, models are initiated with several different CPOs to simulate unique deformation histories. Significantly, models initiated with a preexisting CPO evolve differently than the CPOs generated without preexisting texture. Moreover, the CPO in each model evolves differently as a function of strain. Numerical simulations are compared to laboratory experiments by Boneh and Skemer (2014). In general, the D-Rex and VPSC models are able to reproduce the experimentally observed CPOs, although the models significantly over-estimate the strength of the CPO and in some instances produce different CPO from what is observed experimentally. Based on comparison with experiments, recommended parameters for D-Rex are: M* = 10, λ* = 5, and χ = 0.3, and for VPSC: α = 10–100. Numerical modeling confirms that CPO evolution in olivine is highly sensitive to the details of the initial CPO, even at strains greater than 2. These observations imply that there is a long transient interval of CPO realignment which must be considered carefully in the modeling or interpretation of seismic anisotropy in complex tectonic settings.

In Situ Measurement of Magnesium Carbonate Formation from CO2 Using Static High-Pressure and -Temperature 13C NMR

We explore a new in situ NMR spectroscopy method that possesses the ability to monitor the chemical evolution of supercritical CO(2) in relevant conditions for geological CO(2) sequestration. As a model, we use the fast reaction of the mineral brucite, Mg(OH)(2), with supercritical CO(2) (88 bar) in aqueous conditions at 80 °C. The in situ conversion of CO(2) into metastable and stable carbonates is observed throughout the reaction. After more than 58 h of reaction, the sample was depressurized and analyzed using in situ Raman spectroscopy, where the laser was focused on the undisturbed products through the glass reaction tube. Postreaction, ex situ analysis was performed on the extracted and dried products using Raman spectroscopy, powder X-ray diffraction, and magic-angle spinning (1)H-decoupled (13)C NMR. These separate methods of analysis confirmed a spatial dependence of products, possibly caused by a gradient of reactant availability, pH, and/or a reaction mechanism that involves first forming hydroxy-hydrated (basic, hydrated) carbonates that convert to the end-product, anhydrous magnesite. This carbonation reaction illustrates the importance of static (unmixed) reaction systems at sequestration-like conditions.

Quantitative Identification of Metastable Magnesium Carbonate Minerals by Solid-State 13C NMR Spectroscopy

In the conversion of CO2 to mineral carbonates for the permanent geosequestration of CO2, there are multiple magnesium carbonate phases that are potential reaction products. Solid-state (13)C NMR is demonstrated as an effective tool for distinguishing magnesium carbonate phases and quantitatively characterizing magnesium carbonate mixtures. Several of these mineral phases include magnesite, hydromagnesite, dypingite, and nesquehonite, which differ in composition by the number of waters of hydration or the number of crystallographic hydroxyl groups. These carbonates often form in mixtures with nearly overlapping (13)C NMR resonances which makes their identification and analysis difficult. In this study, these phases have been investigated with solid-state (13)C NMR spectroscopy, including both static and magic-angle spinning (MAS) experiments. Static spectra yield chemical shift anisotropy (CSA) lineshapes that are indicative of the site-symmetry variations of the carbon environments. MAS spectra yield isotropic chemical shifts for each crystallographically inequivalent carbon and spin-lattice relaxation times, T1, yield characteristic information that assist in species discrimination. These detailed parameters, and the combination of static and MAS analyses, can aid investigations of mixed carbonates by (13)C NMR.

Ultramylonite generation via phase mixing in high-strain experiments

Dynamic recrystallization and phase mixing are considered to be important processes in ductile shear zone formation, as they collectively enable a permanent transition to the strain-weakening, grain-size sensitive deformation regime. While dynamic recrystallization is well understood, the underlying physical processes and timescales required for phase mixing remain enigmatic. Here, we present results from high-strain phase mixing experiments on calcite-anhydrite composites. A poorly mixed starting material was synthesized from fine-grained calcite and anhydrite powders. Samples were deformed in the Large Volume Torsion apparatus at 500°C and shear strain rates of 5 × 10−5 to 5 × 10−4 s−1, to finite shear strains of up to γ = 57. Microstructural evolution is quantified through analysis of backscattered electron images and electron backscatter diffraction data. During deformation, polycrystalline domains of the individual phases are geometrically stretched and thinned, causing an increase in the spatial density of interphase boundaries. At moderate shear strains (γ ≥ 6), domains are so severely thinned that they become “monolayers” of only one or two grains width and form a thin compositional layering. Monolayer formation is accompanied by a critical increase in the degree of grain boundary pinning and, consequently, grain-size reduction below the theoretical limit established by the grain-size piezometer or deformation mechanism field boundary. Ultimately, monolayers neck and disaggregate at high strains (17 <γ <57) to complete the phase mixing process. This “geometric” phase mixing mechanism is consistent with observations of mylonites, where layer (i.e., foliation) formation is associated with strain localization, and layers are ultimately destroyed at the mylonite-ultramylonite transition.

Torsion experiments on coarse-grained dunite: implications for microstructural evolution when diffusion creep is suppressed

Abstract Large strain deformation experiments in torsion were conducted on a coarse-grained natural dunite with a pre-existing lattice preferred orientation (LPO). Experiments were conducted at conditions where deformation by diffusion creep is initially negligible. Microstructural evolution was studied as a function of strain. We observe that the pre-existing LPO persists to a shear strain of at least 0.5. At larger strains, this LPO is transformed. Relict deformed grains exhibit LPO with [100] crystallographic axes at high angles to the shear plane. Unlike previous experimental studies, these axes do not readily rotate into the shear plane with increasing strain. Partial dynamic recrystallization occurs in samples deformed to moderate strains (γ>0.5). Recrystallized material forms bands that mostly transect grain interiors. The negligible rate of diffusion creep along relict grain boundaries, as well as the limited nature of dynamic recrystallization, may account for the relatively large strains required to observe evolution of microstructures. Our data support hypotheses based on natural samples that microstructures may preserve evidence of complex deformation histories. Relationships between LPO, seismic anisotropy and deformation kinematics may not always be straightforward.

Grain‐size sensitive rheology of orthopyroxene

The grain-size sensitive rheology of orthopyroxene is investigated using data from rheological and microstructural studies. A deformation mechanism map is constructed assuming that orthopyroxene deforms by two independent mechanisms: dislocation creep and diffusion creep. The field boundary between these mechanisms is defined using two approaches. First, experimental data from Lawlis (1998), which show a deviation from non-linear power law behavior at low stresses, are used to prescribe the location of the field boundary. Second, a new orthopyroxene grain-size piezometer is used as a microstructural constraint to the field boundary. At constant temperature, both approaches yield sub-parallel field boundaries, separated in grain size by a factor of only 2–5. Extrapolating to lithospheric conditions, the deformation mechanism transition occurs at a grain size of ~150–500 µm, consistent with observations from nature. As the transition from dislocation to diffusion creep may promote shear localization, grain-size reduction of orthopyroxene may play a prominent role in plate-boundary deformation.

Low Temperature plastic rheology of olivine determined by nanoindentation

Low-temperature plasticity is a deformation mechanism that occurs mainly at high stress and low temperatures and may be important in the shallow lithosphere, at the tips of cracks, and in laboratory experiments. Previous studies investigating the low-temperature plasticity of the mineral olivine have exhibited wide variability in their extrapolations to the athermal flow strength or Peierls stress. To better constrain the rheology of olivine, nanoindentation tests were performed on samples in the temperature range of 0–175°C. The indentation properties were converted to uniaxial properties using a finite element-based method. The data were fit to a standard flow law for low-temperature plasticity, and Peierls stresses between 5.32 and 6.45 GPa were obtained. These results provide increased confidence in the extrapolation of high-pressure and high-temperature laboratory experiments to low-temperature conditions and illustrate the applicability of nanoindentation methods to the study of mineral rheology.

Viscous anisotropy of textured olivine aggregates: 2. Micromechanical model

The significant viscous anisotropy that results from crystallographic alignment (texture) of olivine grains in deformed upper-mantle rocks strongly influences a large variety of geodynamic processes. Our ability to explore the effects of anisotropic viscosity in simulations of these processes requires a mechanical model that can predict the magnitude of anisotropy and its evolution. Unfortunately, existing models of olivine textural evolution and viscous anisotropy are calibrated for relatively small deformations and simple strain paths, making them less general than desired for many large-scale geodynamic scenarios. Here we develop a new set of micromechanical models to describe the mechanical behavior and textural evolution of olivine through a large range of strains and complex strain histories. For the mechanical behavior, we explore two extreme scenarios, one in which each grain experiences the same stress tensor (Sachs model) and one in which each grain undergoes a strain rate as close as possible to the macroscopic strain rate (pseudo-Taylor model). For the textural evolution, we develop a new model in which the director method is used to control the rate of grain rotation and the available slip systems in olivine are used to control the axis of rotation. Only recently has enough laboratory data on the deformation of olivine become available to calibrate these models. We use these new data to conduct inversions for the best parameters to characterize both the mechanical and textural evolution models. These inversions demonstrate that the calibrated pseudo-Taylor model best reproduces the mechanical observations. Additionally, the pseudo-Taylor textural evolution model can reasonably reproduce the observed texture strength, shape, and orientation after large and complex deformations. A quantitative comparison between our calibrated models and previously published models reveals that our new models excel in predicting the magnitude of viscous anisotropy and the details of the textural evolution. In addition, we demonstrate that the mechanical and textural evolution models can be coupled and used to reproduce mechanical evolution during large-strain torsion tests. This set of models therefore provides a new geodynamic tool for incorporating viscous anisotropy into large-scale numerical simulations.