Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In this work, diffusion of Fe-Mg and Mn is used to model the relative timescale of pre-eruptive processes of olivine crystals. The applied method of diffusion chronometry is based on the new, more complex concept of a magmatic system and the process of disequilibrium between crystal and melt. The first conceptual model of a magmatic system involved a large melt-dominated magma body, which was recharged by a feeder system from a deeper source. This interpretation has changed to a more open system with smaller, stacked sills, vertically connected with each other. In such a system, magma can be altered by magma replenishment, resulting in magma mixing and crystal transport, as well as by crustal assimilation. This thesis is based on the concept of magma mixing and crystal transport with the assumption that crystallization happened deep in the crust, in a primitive magma. A primitive magma is a magma directly derived from the mantle and has not changed in composition since formation. Olivine crystals that crystallize in such magma have a high magnesium concentration. After this crystallization, the crystals are carried through vertical connections towards a more shallow magma reservoir with a more evolved magma. An evolved magma has changed in composition by increasing silica and decreasing magnesium concentration. The transport of primitive magma, comprising the crystals towards the shallow magma chamber leads to magma mixing and disequilibrium between the crystals and the newly formed melt. The crystals react by diffusion of mobile elements, where elements flow in the direction of the lowest concentration. Therefore, diffusion induces a chemical concentration gradient between core and rim, which is expressed in rim zonation in olivine. This concentration gradient can be modeled with the diffusion coefficients of certain elements. The obtained model provides constraints on how long the crystals resided in different sections in the complex volcanic plumbing system up until eruption. Diffusion chronometry studies the timescale of crystals, this method is increasingly applied in the last decade, because of the accurate determination of diffusion coefficients for mobile elements in crystals. In this thesis, the focus lies on modeling the diffusion of Fe-Mg, as forsterite, and Mn in olivine. The studied olivine crystals originate from Nyamuragira volcano, situated in the Democratic Republic of Congo. Nyamuragira is a shield volcano, which is characterized by a broad domed volcano with gently sloping flanks. The volcano is located on the Western Branch of the East African Rift System and is the most western volcano of the Virunga Volcanic Province. Nyamuragira is one of the most active volcanoes on earth with an average eruption frequency of three years. The lava is classified as basanite to tephritic phonolite. The first step in this thesis was to petrographically describe the volcanic rocks to select the samples with the best examples of zoned olivine crystals. Next, the chemical bulk composition is determined with X-Ray fluorescence, in order to classify our samples. Then, the main objective of establishing the diffusion time is accomplished through best-fit diffusion modeling with the use of the MATLAB software. With our model, we obtained timescales ranging between 1 and 70 days for olivine crystals from Nyamuragira. The average diffusion time is 20 days, therefore, we can conclude that the pre-eruptive processes in the plumbing system

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Temperature exerts a major control on magmatic processes in the mantle. Spatial and temporal variations in mantle temperature are thought to drive deep planetary processes such as the mantle convection. Strong temperature anomalies near the core-mantle boundary may give rise to mantle plumes that are considerably hotter than the ambient mantle. Methods to investigate the mantle temperature at which a melt formed are needed to confirm whether such hot magmas are truly present within the mantle. Geothermometers based on olivine equilibria are a good proxy to assert the minimal melting temperature of a magma since olivine is one of the first phases to crystallise. Early olivine-based geothermometers focussed on cation equilibria between olivine and the melt, but the necessity to estimate liquid composition has proven to give rise to uncertainties. Hence, Wan et al. (2008) proposed the Al-in-olivine geothermometer which is no longer dependent on mineral-melt equilibria, but only on mineral-mineral equilibria. The Al-in-olivine geothermometer uses the partitioning of Al between forsterite-rich olivine and Cr-rich spinel as proxy for the liquidus of a melt, which is dependent on temperature and spinel Cr#. The limited calibration range concerning pressure, temperature, oxygen fugacity (fO2) and compositional conditions gives rise to uncertainties when research is conducted on mantle plumes. The goal of this thesis was to gain further insight in the Al partitioning by investigating higher fO2 and temperature conditions for a wider range of chemical compositions. Obtaining spinel minerals from the melt has proven difficult, but doping with Cr2O3 was partially successful in generating higher spinel Cr# [=Cr/(Cr+Al)]. Secondary fluorescence during microprobe analysis gave rise to anomalous high Al2O3 concentrations in olivine. Analytical uncertainties made it difficult to properly assess the Al equilibrium. Moderate trends suggest that – other than for temperature and spinel Cr# – the Al equilibrium seems to be dependent on glass SiO2, olivine Cr# and forsterite content, spinel P2O5 and Fe3+, and fO2. Further research is needed to fully understand the influence of these factors. Recalibration of the Al-in-olivine geothermometer and application to natural samples have shown that – although a wider calibration range can be reached with the experimental results of this thesis – the original geothermometer is still more accurate. Application to natural rocks of high Cr# suggests that the original geothermometer remains realistic. More accurate results are needed to properly investigate the Al-equilibrium and recalibrate the geothermometer. Measurements with a nano-SIMS are suggested to avoid secondary fluorescence on trace elements.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Het smelten van de mantel is een van de meest belangrijke petrologische processen op Aarde. De korst van de Aarde is gevormd door het smelten van de mantel. Deze studie bestaat uit twee delen. Eerst wordt er een calibratiemethode voor de Raman spectrometer van de KU Leuven ontwikkeld om watergehaltes te kunnen meten van smelt insluitsels. Dit zijn smeltdeeltjes die gevangen worden in groeiende kristallen tijdens kristallisatie. De klassieke toestellen om deze smelt insluitsels te analyseren zijn zowel duur als tijdrovend, dus Raman spectroscopie is voorgesteld als een alternatieve methode. Verschillende instellingen zijn geëvalueerd om de beste Raman spectra te verkrijgen. Een Python script is geprogrammeerd om de Raman spectra te verwerken en om de calibratiemethode te testen. Twee verschillende calibratiemethodes van vorige auteurs zijn getest. Om de calibratiemethoden te valideren, is er nood aan smelt insluitsels met een hoog watergehalte. In het tweede deel van de studie worden natuurlijke stalen van het eiland Pico in de Azoren bestudeerd. Smelt insluitsels in olivijnkristallen van Pico hebben een hoog watergehalte. Ter hoogte van de Azoren is de oceaankorst ook zeer dik, wat ideale omstandigheden geeft om het smelten van de mantel te bestuderen. Olivijnkristallen zijn zorgvuldig met de hand uitgekozen en gepolijst tot de smelt insluitsels aan het oppervlak lagen. Hun samenstelling is geanalyseerd met Electron Probe Micro Analyser. Deze data tonen aan dat de samenstelling van de insluitsels veranderd was na kristallisate, dus is de samenstelling gecorrigeerd om hun originele samenstelling te verkrijgen. De hoeveelheid vluchtige stoffen (water, koolstofdioxide, zwavel en fluor) is gemeten met Secondary Ion Mass Spectrometry. Het watergehalte van de insluitsels is hoger dan data van vorige auteurs en van andere oceanische eilanden. De samenstelling en hoeveelheid vluchtige stoffen van de insluitsels is gebruikt om de kristallisatietemperatuur te berekenen. Ook een potentiële manteltemperatuur wordt op deze manier berekend. Deze manteltemperatuur is relatief laag en vergelijkbaar met de manteltemperatuur onder een mid-oceanische rug. Alhoewel de Azoren aanzien worden als een klassiek voorbeeld van de interactie tussen een mid-oceanische rug en een mantelpluim, wijzen deze berekende potentiële temperaturen op de afwezigheid van een mantelpluim.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

A large number of different types of rocks of Mesozoic age, dating back 250 to 66 million years, are present in the north-east part of Asia. The vast area was researched thoroughly under the Soviet regime in the search for natural wealth such as gold and tin. During this time, much data was obtained about granites, a type of rock formed by magmatic processes. A total of 4659 samples was collected and selected by Vistelius (1995), discarding data where measurement errors were made or which were incomplete. The composition of the rock samples was given in major oxides, which are the components that mainly make up a rock. The number of samples representing the over 5*10^7 km2 large area was low. The composition at locations without direct measurements was, therefore, estimated by the data points of Vistelius (1995), making use of interpolation techniques. Two different classifications systems were used on both the interpolated data and the original samples of the dataset of Vistelius (1995), to unravel differences in chemical composition and time of formation of granites located in the research area. The so-called QAPF classification subdivides the data according to the presence of the minerals quartz (Q) or feldspathoids (F), plagioclase (P), and K-feldspar (K) in the samples. Since the composition of the data points is given by oxides, a transformation on the data had to be done to obtain the composition in terms of mineral occurrence. The second classification system is purely mathematically and is called the cluster analysis. The QAPF classification was not useful to differentiate between regions as the majority of the samples belonged to the same class of monzogranite. Nonetheless, the classification based on the cluster analysis using the estimations from the interpolation showed good results. Different large regions in the research area were distinguished and could be linked to the literature. However, the number of data points for such a vast area proved to be too low to describe the research area in detail. New conclusions about the research area could, therefore, not be drawn. However, the approach outlined in this thesis should prove to be very useful given a more densely populated dataset.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

As surface and interior compositions of rocky exoplanets still cannot be directly determined, rocky exoplanets are assumed to have similar refractory element ratios as their stars considering they have formed from the same cloud of gas and dust. This assumption allows to examine different compositions and their effect on the interior of a planet with that composition. The aims of this MSc research are (1) to determine the melting behaviour of non-peridotitic mantles of low-Mg planets, and (2) to determine the composition of melts produced in a non-peridotitic mantle and the formation of a secondary volcanic crust, and (3) to get first order constraints on the atmospheric composition. The selected compositions are based on the EH composition of Berthet et al. (2009), as enstatite chondrites are assumed to be one of the potential building blocks of reduced planetary bodies. From the three selected compositions, with a varying molar Ca/Al ratio (0.76, 0.85, and 0.95), synthetic samples are prepared using the following pure oxide powders: SiO2, Al2O3, Fe2O3, MgO, CaSiO3, and Na2SiO3. High temperature experiments are run with the samples in a vertical tube gas-mixing furnace at five different temperatures (1300°C, 1325°C, 1350°C, 1375°C, and 1400°C), at an fO2 of FMQ-2 (Fayalite-Magnetite-Quartz) for 50 to 100 hours. Afterwards, the samples are chemically analysed by the Scanning Election Microscope, developing a thermodynamic model of mantle melting for various lithological sources. The SEM is also used to attain BSE images to determine the distribution of melt and crystals in the samples. Electron Dispersive X-ray Spectroscopy is used to chemically analyse the experimental products. The main crystallising phases present in the experimental products are orthopyroxene and quartz. They both occur as anhedral to euhedral crystals within the melt, with quartz occasionally being trapped in Opx crystals at lower temperatures. The composition of the melt has decreasing amounts of Si and Al, and increasing amounts of Mg with increasing temperatures, i.e. degree of partial melting. The melt composition is also slightly influenced by the increasing Ca/Al ratio in the starting composition, where an increase of Mg and Ca is visible and a decrease of the amounts of Na and Si. However, these are not highly significant differences as they only vary up to 2 oxide% with increasing Ca/Al ratio. Further, the estimated fraction of melt in the samples range from 25% to 55%. This wide range is mostly due to the occurrence of melt blobs in certain samples, greatly affecting the estimations. An increasing Ca/Al ratio also increased the melt fraction present in the experiments, although only at higher temperatures. At lower temperatures a more dubious trend is visible. In general only a minimal influence of the Ca/Al ratio is visible within the samples of this research, causing more CaO and slighly less Al2O3 to be present in the melt at higher Ca/Al ratios. Lastly, predictions for volatile saturation in the produced melts show the following general evolution in dominant volatile species: H2-CH4 -> H2 -> CO -> CO-CO2. An increasing H as H2O concentration shift the atmospheres with reduced oxygen fugacities (IW-3 to IW-6) from a CO-rich to H2 and CH4 rich atmospheres. At more oxidized conditions (IW to IW+3) the atmospheres are always CO-CO2 rich.