Birefringence, the Lost and Forgotten Optical Property
Petrologists and mineralogists could more effectively exploit birefringence and its derivative properties,
retardation and interference color, to characterize minerals in thin section. Mineralogy texts and courses largely
confine their treatments to the principal birefringences: γ - α, β - α, and
γ - β in biaxial crystals and |ε - ω| in uniaxial crystals. Each section
through a biaxial or uniaxial crystal has a birefringence and the birefringences range from zero to a maximum
value for the substance under examination. The distribution of birefringence values on the indicatrix is not
random; rather it follows a regular pattern. The pattern reveals itself in stereographic projection and it can be
quantitatively depicted if the principal refractive indices are known. This pattern, when combined with the optical
orientation of the crystal, places limits on the crystallographic orientation of the crystal plate in thin section.
Birefringence can be used to estimate the composition of binary solid solutions displaying moderate to high
interference colors if the optical orientation can be established. Computer color management techniques
provide estimation of retardation values derived from interference colors to within a few nanometres in the
range 250 to 1650 nanometres. These color management techniques can also be used to create charts and
diagrams with retardation values as one variable and other mineral properties and compositions as the other
variable. On such diagrams, interference colors can be painted, the color bands normal to the retardation axis
like Michel-Levy charts.
Core Formation and Extreme Fractionation of Highly Siderophile Elements (HSE)
Relative abundances of the HSE in Earth's silicate mantle are nearly the same as in chondritic meteorites, but absolute concentrations are ∼100-fold lower. To explain these observations as a consequence of equilibrium separation of Earth's Fe-Ni core requires metal-silicate partitioning to be similar for all the HSE, with values of 200-800. Although low temperature experiments document partitioning in excess of 100,000 for all the HSE, higher temperature experiments reveal lower partitioning for some of these elements. However, convergence of partitioning at high temperature for all the HSE has hitherto been absent. To close this knowledge gap, we have performed piston-cylinder experiments to measure the solubility and estimate the metal-silicate partitioning of Os, Ir and Au at 2 GPa, 2173-2588 K, and IW-1.5 to +3. Experiments are also in progress to characterise the behaviour of Re. Samples consist of synthetic basalt plus a metal mixture encapsulated in graphite or Fe-Ir alloy. Time-resolved LA-ICPMS spectra obtained from experiments which were initially metal undersaturated show a uniform distribution of Os, Ir and Au, consistent with homogeneous solution of these elements in the silicate melt. We have confirmed equilibrium in our experiments by approaching metal solubilities from both over- and undersaturated conditions. At the same relative oxygen fugacity of IW + 2, the slopes of log Dmetal/silicate versus temperature for all three elements yield indistinguishable values, but partitioning of Os and Ir into the metal phase is 10,000x greater than that of Au. Metal-silicate partitioning of Os and Ir show a general increase with decreasing oxygen fugacity consistent with solution of both metals as an oxide species in the silicate melt. Partition coefficients for Au show a slight drop with decreasing oxygen fugacity, consistent with Au solution in molten silicate as an Au-silicide species. For conditions approximating core-formation, partitioning of osmium and iridium likely exceeds 10 million, and the relative partitioning of osmium and iridium to gold is >10,000. Such extreme HSE fractionation cannot be reconciled in the context of equilibrium core formation models, requiring the addition of HSE-bearing material to the mantle following metal extraction.
Water Quantification and Speciation in Silicate Melt Using Raman Spectroscopy
Water plays a fundamental role in the dynamics and evolution of magmas in the deep interior and during volcano eruptions. In particular, extraction of magma can change drastically as a function of dissolved water content. In addition to temperature, pressure and main chemical components, volatiles exert a strong influence on the physical properties of magmas. However their speciation, which is pressure and temperature dependent, must be understood to fully assess their effects on the properties of magmas. Therefore we need to quantify the proportion and speciation of volatile in silicate melts. To quantify water content in silicate glass or lava, bulk analysis like fire loss or water extraction are very well used and accurate, but they are destructive and a sufficient quantity of material is needed. Raman spectroscopy allows measuring water contents of hydrous silicate glasses. It offers several advantages in the study of natural silicate glasses like pumice: (i) the high spatial resolution of 1-2 μm, (ii) the non-destructive nature of this analysis and (iii) sample preparation is not necessary. Two parts can be distinguished on Raman spectra of a hydrated silicate glass: the low wave-number region, which correspond to vibrations of the silicate network (0-1500 cm-1), and the high wave-number region, which correspond to the OH- stretching vibration and H2O molecular vibration (3100 - 3750 cm-1). Behrens et al. (2006) have shown, using Long correction and normalizing spectra with the highest intensity, that water content can be determine. However, there calibrations are SiO2 dependent. We present now a new chemical independent internal calibration. Using cubic-spline baseline to fit the bases of silicate peaks in the low wave-number region of Raman spectra allow us to take into account the changes in this region induced by chemical variations in our set of glasses (basaltic to rhyolitic). That allows us to study basaltic glasses or rhyolitic glasses with the same calibration line. Using this calibration, we can discuss some results obtained on the water speciation at high temperature in silicate melts.
Diffusion in Natural Melts: Kinetics and Mechanisms of Redox Reactions
Diffusion in silicate melts plays a fundamental role in all magmatic processes in nature as well as in the glass industry. The diffusivity contrast that occurs between the so-called network-former (e.g., Si, Al) and network- modifier (e.g., alkali and alkaline-earth) cations is of particular importance. Whereas the diffusivities of all these cations tend to converge at the high-temperature limit, a strong decoupling is observed at glass transition temperatures. The diffusivity of oxygen and network-former cations becomes much lower than that of network- modifier cations. This decoupling exerts an influence on the kinetics and mechanisms of redox reactions. In redox reactions, diffusion of oxygen is the rate-limiting factor only at superliquidus temperatures whereas at lower temperatures, the kinetics of these reactions is controlled instead by diffusion of alkaline-earth or alkaline cations coupled to a flux of electron holes. For iron redox reactions, we have investigated these effects quantitatively from the glass transition up to 2100 K by in situ high temperature X-ray Absorption Near Edge Structure (XANES) experiments at the iron K-edge. To compare in a simple way the observations made, we have introduced the concept of redox diffusivity from the time required to achieve redox equilibrium at a given temperature. Comparisons of these redox diffusivities with the diffusivities of oxygen, network-forming and network-modifying cations then allow one to distinguish the temperature range where a given redox mechanism predominates. The results obtained in this way for a variety of natural iron lavas will be presented. In particular we will discuss the kinetics of iron redox reactions obtained on depolymerized melt such as basalt and komatiite or on polymerized melt such as phonolite and rhyolite.
POWTEX - A new High-Intensity Powder and Texture Diffractometer at FRM II, Garching Germany
In recent years, neutron diffraction has become a routine tool in Geoscience for experimental high-field (HP/HT/HH) powder diffraction and for the quantitative analysis of the crystallographic preferred orientation (CPO). Quantitative texture analysis is e.g. involved in the research fields of fabric development in mono- and polyphase rocks, deformation histories and kinematics during mountain building processes and the characterization of flow kinematics in lava flows. Secondly the quantitative characterization of anisotropic physical properties of both rock and analogue materials is conducted by bulk texture measurements of sometimes larger sample volumes. This is easily achievable by neutron diffraction due to the high penetration capabilities of the neutrons. The resulting geoscientific need for increased measuring time at neutron diffraction facilities with the corresponding technical characteristics and equipment will in future be satisfied by this high-intensity diffractometer at the neutron research reactor FRM II in Garching, Germany. It will be built by a consortium of groups from the RWTH Aachen, Forschungszentrum Jülich and the University of Göttingen, who will also operate the instrument. The diffractometer will be optimized to high intensities (flux) with an equivalent sufficient resolution for polyphase rocks. Furthermore a broad range of d-values (0.5 to 15 Å) will be measurable. The uniqueness of this instrument is the geoscientific focus on different sample environments for in situ-static and deformation experiments (stress, strain and annealing/recrystallisation) and (U)HP/(U)HT experiments. A LP/LT or atmospheric-P deformation rig for in situ-deformation experiments on ice, halite or rock analogue materials is planned, to allow in situ-measurements of the texture development during deformation and annealing. Additionally a uniaxial HT/MP deformation apparatus for salt deformation experiments and an adapted Griggs- type deformation rig are also designated. Furthermore an uniaxial stress frame for in situ stress investigations is planned to conduct simultaneous measurements of stress, elastic or plastic deformation and texture. Other sample environments for geoscientific application will be HP/HT furnaces and pressure cells for powder diffraction investigations. Furthermore the diffractometer will be built in combination with a high-pressure multi anvil up to 25 GPa and 2500 K built by the University of Bayreuth at the same beam line. The detector concept allows single shot texture measurements and therefore the measurement of larger geological sample series as necessary for the investigations of complete geological structures. This concept is complementary to the geoscience neutron texture diffractometer in Dubna, Russia and the stress diffractometer STRESS-SPEC located also at the Garching research reactor. For powder diffraction the diffractometer will be complementary to the existing high-resolution powder diffractometer SPODI at the FRM-II. It will offer the possibility of short, high-intensity parametric powder diffraction measurements in dependency of temperature, electrical, magnetic and stress fields due to the higher flux at the sample. The optimization to high-intensities and therefore short measuring times will also allow time-resolved measurements of kinetic reactions even of small sample volumes.
Calculated Olivine and Augite Distribution Coefficients From a Differentiated Proterozoic Sill, Victoria Island, NWT
A series of sills related to the late Proterozoic Franklin Igneous event are exposed in the Minto syncline on Victoria Island NWT, Canada. A continuous drill core through a single 43m thick sill with upper and lower chilled margins, has been examined in detail. The drill core was split vertically and 0.3 to 0.5m half-sections were homogenized to produce 92 bulk rock samples. A plot of Mg# vs. height has a classic S-shaped profile with a maximum Mg# of 80 reached 6.5m above the base and a minimum Mg# of 35 at the 37m level. The upper and lower chilled margins have Mg# 70. The lower part of the sill is an olivine accumulation zone in which the percentage of olivine increases systematically from 10% at the lower chilled margin to 48% at the 8m level. The olivine zone is overlain by a 2m thick augite zone in which the percentage of accumulated augite decreases upward from 44% to 6%. The upper 33m of the sill is medium-grain diabase that becomes progressively enriched in Fe, Ti, and excluded elements upward from the 10m level to the 37m level where Fe and Ti reach maximum values and then decrease upward to the upper chilled margin. The Fe and Ti rich zone appears to have formed as a result of the expulsion of differentiated interstitial melt from compaction of the lower part of the sill. Enrichment factors (upper zone/chilled margin): Fe = 1.4, Ti = 1.7, Na = 2.5, K = 3.0, P = 1.7, Ba = 1.8, Hf = 2.6, Nb = 1.7, Rb = 1.5, Ta = 1.5, W = 3.0, Y = 2.8, Zr. = 2.5, REE-total = 1.6, suggest that this zone may represent residual melt after 50 to 70% crystal fractionation. The lower olivine and augite accumulation zones appear to have formed as a result of the settling of phenocrysts in the initial magma. Fourteen samples from the olivine zone have linear trends for Mg and a number of trace elements when plotted against % olivine suggesting the accumulation of Fo 84 olivine with apparent xl/lq distribution coefficients: Ni = 14; Cr = 10; Co = 3.9; V = 0.01; Sc = 0.17; Zn = 2.1; and REE-total = .01. The high coefficient for Cr suggests that the olivine contains 0.6% chromite inclusions. Four samples from the augite zone indicate accumulation of Mg# 81 augite with apparent xl/lq distribution coefficients: Ni = 4.5; Cr = 20; Co = 1.3; V = 0.66; Sc = 1.5; Zn = 0.28; and REE-total = 0.26.
Physico-chemical processes involved in the construction of a composite pluton: an example from Stewart Island, New Zealand
The ~140Ma calc-alkaline Bungaree Intrusives of Stewart Island, New Zealand, were emplaced incrementally over a period of at least 0.4Ma. The felsic magma chamber was periodically interrupted by multiple injections of mafic magma in the form of a sequence of mingled sheets and enclaves. Tilting has exposed a cross-section through the magma chamber, which allows for interpretations regarding magmatic processes to be made with respect to height or depth within the chamber. Processes operating during the construction and evolution of the Bungaree pluton have been identified through detailed analysis of structures and textures in the field, coupled with comprehensive geochemistry and geochronology. Field interpretation of magma mingling and mixing structures and textures has enabled the identification of several physical processes operating within the pluton. These include; multiple mafic magma pulses that extensively mingled and sometimes mixed locally and at depth; differences in the style of mingling due to variations in magma viscosity, temperature, volume and rate of replenishment; magmatic mineral fabrics and aligned mafic enclaves characteristic of magmatic flow; shortening of the chamber during crystallization due to magmatic loading; and crystal accumulation at the base of the chamber. Geochemical and geochronological data were obtained to constrain the age, and determine the conditions that operated during the construction and evolution of the magma chamber. Transects on hornblende and zircon crystals reveal that conditions within the chamber fluctuated through time. Hornblende crystals exhibit variations in Al, Ti and Cl, which are interpreted to represent variations in temperature and water content within the crystallizing magma. Zircon crystals also exhibit variations in chemistry from core to rim, with several zircons having reverse cooling trends, with hotter rims than cores. These crystals are also depleted in Y and have higher Zr/Hf ratios relative to zircons that exhibit normal cooling trends. These changes in crystal chemistry were probably caused by the input of hot, wet mafic magmas into the felsic chamber.
Construction and Evolution of an Ice-Confined Basaltic Eruptive Fissure Complex: Sveifluhals, SW Iceland
Ice-confined fissure-fed basaltic eruptions were common in Iceland during the Pleistocene. Most of these generated complexes consisting of closely spaced, sub-parallel, multi-vent ridges. Individual ridges are constructed mostly of numerous linked steep-sided mounds (point-source vents) and short ridges (fissure vents) of subaqueous lavas, most of which are draped by phreatomagmatic tephra. There are more than 1000 such ridges in Iceland and they represent an important, largely untapped, database on North Atlantic terrestrial ice conditions. Some of these complexes were of comparable length to the famous Laki fissure eruptions of 1783-1785. Sveifluhals is a 21.5-km long formerly ice-confined fissure complex of unknown age, but similar nearby centers have been dated at about 47ka. Ice-thickness estimates, based on volatile analysis of pillow rind glass vary from 70-400m. This is the first detailed study of such a complex, which focuses on how it was constructed in space and time, how it differs from published studies of simple single short fissure ice-confined centers, and how it interacted with the overlying ice. Initial results have identified vents with an average spacing of 0.7km on at least nine sub-parallel fissures, which are spaced about 0.1-0.5km apart. The vents are dominated by inward and outward-dipping rotated (oversteepened) blocks of bedded tephra, which overlies mostly slumped massive tephra. The tephra drapes subaqueous lava mounds that display marginal steep- ridges of subaqueous lavas, that may have been emplaced in short sub-ice tunnels. We have also identified several now dry lakes that formed in some inter-ridge areas, but whose age relationship to ridge construction is as yet uncertain. We present here our preliminary data on vent locations and estimates of product volumes, volumes of ice melted, and location of possible meltwater pathways. We discuss the most important differences of the processes and products of such ridge complexes compared to published examples of short "monogenetic" ridges, and the implication of these differences for modeling ice melting, meltwater drainage and hazards.