Search

The Chicxulub impact event at ca. 66 Ma left in its wake the only complex crater on Earth with a preserved peak ring, characterized by a well-developed magnetic anomaly low. To date, little is known about its magnetic properties. The joint Integrated Ocean Drilling Program (IODP) and International Continental Scientific Drilling Program (ICDP) Expedition 364 drill core M0077A revealed that the peak ring consists of uplifted and strongly deformed granitoid basement rocks overlain by a 130-m-thick impact melt and suevite layer. Pre- and postimpact hydrothermal systems affected this basement with maximum temperatures up to 450 °C. We used microscopy, mineral chemistry, temperature-dependent magnetic susceptibility, and hysteresis properties to characterize the magnetic mineralogy of pre-, syn-, and postimpact rocks. Compared to its amount of pure, stoichiometric shocked magnetite, the granitoid basement shows low magnetic susceptibility, which is in line with earlier experimental studies indicating that shock reduces magnetic susceptibility. Cation-substituted magnetite with varying compositions in the melt rocks carries a higher induced and remanent magnetization compared to the basement. In the granitoid basement, magnetite was partially oxidized to hematite by a pre-impact hydrothermal event, but at lithological contacts with high-temperature impact melt rock, this hematite was locally retransformed back to magnetite. Elsewhere in the granitoid basement, the temperature reached in the hydrothermal system was too low for hematite retransformation. It was also too low to anneal all the lattice defects in the shocked magnetite, which likely occurs above 540 °C. The presence of shocked magnetite in the granitoid basement well explains the magnetic anomaly low due to its unusually low induced magnetization.

Single crystals of five (Al,Ge)-mullites incorporating Pb, and four of which also incorporating foreign cations (Fe,Cr,Nd,Sm) were grown by flux techniques in a PbO-MoO3 flux. They were characterized by scanning electron microscopy, electron microprobe analyses, single-crystal X-ray diffraction. In addition, the refractive indices of mullite containing Nd were determined by spindle-stage optical investigations. Careful inspection of the single-crystal X-ray diffraction data revealed that weak superstructure reflections observed in all doped crystals violating the reflection conditions can be attributed to λ/2 contributions in the primary X-ray beam. Consequently, all crystal structures were refined in space group Pbam, thus avoiding a symmetry lowering to a noncentrosymmetric subgroup as done in earlier work on a (Al,Ge)-mullite doped with Pb and Nd (Saalfeld & Klaska, Z. Kristallogr. 1985, 172, 129–133). The following phases with chemical compositions used in the refinements were obtained: undoped mullite (Al4.50Ge1.50O9.75; a = 7.6559(4) Å, b = 7.7763(4) Å, c = 2.9233(2) Å, V = 174.04(2) Å3); (Pb,Fe)-doped mullite (Pb0.02Fe0.68Al3.95Ge1.37O9.70; a = 7.7125(7) Å, b = 7.8527(7) Å, c = 2.9528(2) Å, V = 178.83(3) Å3); (Pb,Cr)-doped mullite (Pb0.01Cr0.63Al3.90Ge1.47O9.75; a = 7.6917(6) Å, b = 7.8168(6) Å, c = 2.9522(2) Å, V = 177.50(2) Å3); (Pb,Nd)-doped mullite (Pb0.06Nd0.02Al4.82Ge1.18O9.69; a = 7.6585(7) Å, b = 7.7666(7) Å, c = 2.9164(3) Å, V = 173.47(3) Å3); (Pb,Sm)-doped mullite (Pb0.06Sm0.02Al4.55Ge1.45O9.79; a = 7.6563(3) Å, b = 7.7873(3) Å, c = 2.9236(1) Å, V = 174.31(1) Å3); Pb is only incorporated into the crystal structure when a co-dopant element is present. Then it resides together with Nd or Sm in the oxygen-vacancy sites created by the formation of triclusters of AlO4 and GeO4 tetrahedra. In the case of (Pb,Fe)-doped mullite, Fe shares the same position as Al and Ge. In contrast to the (Al,Si)-mullites, Ge is located in both tetrahedral sites T and T*. The occupancies follow a substitution scheme according to Pb q (Nd,Sm) r (Cr,Fe) z Al4+2v−z Ge2−2v O10−v+q+3/2r . With v = number of vacancies, such a mullite can be understood as a “stuffed mullite” derived from a related “open mullite” (no vacancies filled with large cations) of composition (Cr,Fe) z Al4+2v−z Ge2−2v O10−v and then “stuffed” with qPb2+ + r(Nd3+,Sm3+) formula units where concurrently the number of available O3-vacancies is reduced by q + 3/2r units of extra oxygen. Thus, charge compensation upon incorporation of Pb2+ and (Nd,Sm)3+ is achieved by adding the amount of oxygen corresponding to the oxidation state of divalent Pb2+ and trivalent rare-earth elements. Based on this description, the maximum number of large cations which can be stuffed into the mullite structure can directly be calculated from the v-value of the related “open mullite”. In contrast, the smaller cations Fe3+ and Cr3+ are directly substituting Al3+. In the stuffed mullites, Pb and (Nd,Sm) could not be distinguished and were refined with a mixed occupancy on the same site. In addition to structure analysis, refractive indices of (Pb,Nd)-mullite were determined by immersion methods using a micro-refractometer spindle stage yielding n x = 1.697(3), n y = 1.708(3), and n z = 1.710(3) and 2V z = 122(4)°. The mean refractive index corresponds closely to the corresponding parameter calculated from the chemical composition whereas it would be significantly off if the extra cations were ignored, thus representing an independent evidence for the incorporation of Pb and Nd into the crystal structure.

Melt inclusions in zircon represent time capsules, which provide deep insights into igneous rock formation including timing, physicochemical conditions, and the compositional evolution of cogenetic magmas with respect to major, trace, and volatile elements. However, their full potential as petrogenetic indicators, in particular their usability and consistency as geothermometers, is poorly investigated. Therefore, we present new mineralogical and chemical data for recrystallized and homogenized melt inclusions and host zircon from different mafic and felsic rocks of the Bushveld Complex (BC), South Africa. Samples include rutile-bearing cumulate rocks of the Marginal and Critical zones, as well as rutile-free, magnetite-ilmenite-titanite-bearing diorites and granites of the Upper Zone, Rashoop Granophyre and Lebowa Granite Suite. All melt inclusions, irrespective of rock type, have rhyolitic compositions with SiO2 contents ranging from 65 to 78 wt.%, and H2O from 1.6 to 4.0 wt.%, whereas trace element contents differ systematically between rock types. In rutile-bearing mafic rocks, melt inclusions commonly show higher Ti contents (>800 ppm), higher Th/U ratios (up to 38), and lower REE contents (ƩREE 150 ppm). Liquidus temperatures of melt inclusions obtained from normative Qz-Ab-Or and rhyolite-MELTS modelling indicate melt entrapment mostly at 930–850 °C (at 200 MPa), tailing down in some samples to 700 °C. For rutile-bearing rocks, these temperatures overlap with those obtained by TiO2-in-melt and Ti-in-zircon geothermometry. For all rutile-free mafic and felsic rocks, reduced TiO2 activities of aTiO2 ∼ 0.3 are required for Ti-in-zircon geothermometry, and TiO2rutile ∼ 30 wt.% for TiO2 saturation geothermometry, to match temperatures from other geothermometers. Furthermore, partition coefficients obtained from melt inclusion - host zircon pairs are within error identical for mafic and felsic rocks and also reveal no systematic dependency on melt inclusion size, composition and entrapment temperature. The results of this study demonstrate that melt inclusions in zircon are a powerful tool for geothermometry and to constrain magma compositions, H2O contents, and TiO2 activities, which are critical for the understanding of magmatic processes shaping Earth’s crust.

Geosciences only play a minor role in today’s German and Austrian school curricula, although being strongly related to important topics such as climate change or the sustainability of resources. The Mineralogical Science Kit (Mineralogischer Lehrkoffer “mileko”) aims at bringing back geoscientific and mineralogical contents into STEAM-fields (science, technology, engineering, art, mathematics) by linking mineralogy to regular teaching subjects. In 5 different modules, pupils can not only discover the world of rocks and minerals but also get an understanding of important principles and processes such as redox reactions by extracting copper from ore (chemistry), the density of rocks that can be linked to the structure of the Earth’s interior (physics, geography), or how the inner structure of materials affects their outer appearance by exploring natural crystalline bodies and their geometry (chemistry, mathematics). Since 2014, about 650 boxes were produced and sent to schools and museums all over Germany. This year marks the start of a new format in the “mileko” portfolio, which is the development of an online accessible video channel. This new digital format aims to expand the applicability of the Mineralogical Science Kit by making geoscientific contents directly accessible for a broad young target group, which not only complements former activities but will also allow us to attract the attention of a wider audience. In combination with activities such as workshops and school visits, we hope to enhance discussions between children, teachers and scientific experts resulting in knowledge transfer from the scientific community to the public.

Single crystals of ABW-type CsAlTiO4 (CAT), CsFeTiO4 (CFT), CsGaTiO4 (CGT), and ANA-type CsTi1.1Si1.9O6.5 (CST) were grown and characterized by electron microprobe analyses, single-crystal X-ray diffraction, thermal analyses, and spindle-stage optical investigations to determine the electronic polarizability of 4-coordinated Ti4+, α([4]Ti4+). The crystal structure of CAT was confirmed to crystallize in the highest possible topological symmetry Imma (a = 8.9677(2) Å, b = 5.7322(1) Å, c = 9.9612(3) Å) with tetrahedrally coordinated Al and Ti equally distributed on Wyckoff position 8i. Twinning by reticular merohedry with a twin index of 2 was observed for most of the crystals resulting in a hexagonal twin lattice (a = 11.487(3) Å, c = 8.968(2) Å) with Laue symmetry 6/mmm. Refractive indices measured by immersion methods on an untwinned specimen are n x = 1.716(5), n y = 1.725(2), and n z = 1.727(1) with 2V z = 127.1(6)°. The diffraction patterns of CFT and CGT clearly showed superstructure reflections causing a symmetry lowering of index 4 with a transformation according to 2a, b, c from Imma to Pmab with a = 18.3054(7) Å, b = 5.8083(2) Å, c = 9.9938(4) Å for CFT, and a = 18.2921(6) Å, b = 5.7636(2) Å, c = 9.9210(3) Å for CGT. Refractive indices for CGT are n x = 1.750(3), n y = 1.772(3), and n z = 1.776(2) with 2V z = 132(1)°. The crystal structure of the ANA-type CsTi1.1Si1.9O6.5 was confirmed to crystallize in space group Ia 3 ¯ $\overline{3}$ d (a = 13.8333(4) Å). The extra 0.5 O atoms are needed for charge compensation and to allow the sum of electronic polarizabilities to give a total electronic polarizability calculated from the refractive index n = 1.718(4). The electronic polarizability of [4]Ti4+ was calculated from the difference between the observed total polarizabilities (derived from the mean refractive indices of CAT and CGT) and the sum of electronic polarizabilities of cations and anions omitting the polarizability of Ti resulting in α([4]Ti4+) = 5.15(5) Å3.

The diffusive exchange of 30 trace elements during the interaction of natural mafic and silicic alkaline melts was experimentally studied at conditions relevant to shallow magmatic systems. In detail, a set of 12 diffusion couple experiments have been performed between natural shoshonitic and rhyolitic melts from the Vulcano Island (Aeolian archipelago, Italy) at a temperature of 1200 {\deg}C, pressures from 50 to 500 MPa, and water contents ranging from nominally dry to ca. 2 wt. %. Concentration-distance profiles, measured by Laser Ablation ICP-MS, highlight different behaviours, and trace elements were divided into two groups: (1) elements with normal diffusion profiles (13 elements, mainly low field strength and transition elements), and (2) elements showing uphill diffusion (17 elements including Y, Zr, Nb, Pb and rare earth elements, except Eu). For the elements showing normal diffusion profiles, chemical diffusion coefficients were estimated using a concentration-dependent evaluation method, and values are given at four intermediate compositions (SiO2 equal to 58, 62, 66 and 70 wt. %, respectively). A general coupling of diffusion coefficients to silica diffusivity is observed, and variations in systematics are observed between mafic and silicic compositions. Results show that water plays a decisive role on diffusive rates in the studied conditions, producing an enhancement between 0.4 and 0.7 log units per 1 wt.% of added H2O. Particularly notable is the behaviour of the trivalent-only REEs (La to Nd and Gd to Lu), with strong uphill diffusion minima, diminishing from light to heavy REEs. Modelling of REE profiles by a modified effective binary diffusion model indicates that activity gradients induced by the SiO2 concentration contrast are responsible for their development, inducing a transient partitioning of REEs towards the shoshonitic melt., Comment: 57 pages, 12 figures, 5 tables

Cumulate rocks of the Upper Main Zone and Upper Zone (UUMZ) of the Bushveld Complex, South Africa, contain the world's major resources of Fe-Ti-V ±, hosted in Ti-magnetite and apatite, and are commonly considered as having crystallized from the last major injection of magma into the magma chamber. In this study, we present the petrography, modal proportions, whole-rock major element chemistry (260 samples), electron microprobe data (∼10 000 analyses for plagioclase, olivine, and pyroxene), and compiled analyses of Cr in magnetite (239 samples) for the UUMZ sampled over 2.1 km of the Bierkraal drill cores in the western limb of the Complex. The UUMZ section exhibits a broad normal fractionation trend upwards, but a series of reversals to more primitive anorthite contents in plagioclase, Mg# in pyroxenes and olivine, Cr in whole-rocks and Cr in magnetite separates are observed, accompanied by the appearance or disappearance of various minerals. Anorthosite or leucogabbro layers are closely linked to these reversals; the reversals in An % of plagioclase are used as boundaries to divide the UUMZ into 18 cycles. These cycles are interpreted as indications of magma chamber replenishment by plagioclase-laden magmas (up to 20 vol. % plagioclase) and are also marked by spikes in Cr content. In addition, abundant Fe-Ti oxide-bearing plagioclase-rich rocks are identified in the lower half of the UUMZ. These have crystallized from a hybrid melt produced by the mixing of a new plagioclase-bearing magma batch and the resident magma. Further crystallization of this hybrid liquid may lead to the formation of magnetite layers in the lower part of the UUMZ. The Bushveld UUMZ therefore grew by multiple emplacements of crystal-laden magmas coming from deep-seated chambers. Slow cooling in a shallow chamber explains the systematic bottom-up compositional evolution in the cumulate pile within individual cycles. The residual melt reached silicate liquid immiscibility soon after the saturation of apatite. Thereafter, segregation of conjugate Fe-rich and Si-rich melts and crystallization of the paired melts produces cumulates with a smooth upward decrease in Fe-Ti oxides, whereas plagioclase mode increases in each apatite-bearing cycle. A comparison of systematic geochemical analyses and a detailed lithological stratigraphy between the Bushveld limbs demonstrates the possible connectivity between the western and eastern Upper Zone but indicates notable differences from the Bellevue section of the northern limb.

At fast-spreading mid-ocean ridges (MORs), the horizon between the axial melt lens (AML) and the overlying sheeted dikes is characterized by extensive anatectic processes. The heat flux of the AML in combination with hydrothermal fluids from above causes high-grade contact metamorphism, which may result in anatexis of the roof rocks above the AML. The products of this process are silica-rich anatectic melts that have the potential to contaminate MOR basalts and residual hornfels. Here, we simulate the complex igneous and metamorphic processes occurring at the AML roof by hydrous partial melting experiments and provide corresponding trace element partition coefficients between melt and residues, which are useful to quantify those processes. We present trace element patterns from experimental anatectic felsic melts and the related residue produced by hydrous partial melting of various types of AML roof rocks. The starting materials used are sheeted dikes and hornfelses from Hole 1256D drilled by the Integrated Ocean Drilling Program. Results are compared with directly-related natural lithologies (i.e., felsic veins and granoblastic hornfels) from the same site. The trace element contents generally overlap with natural examples and experimental melts produced at low water activity (aH2O

A crystal of BaCaBO3F was grown by the Czochralski method. It shows a phase transition at 242.7 °C upon cooling where the hexagonal high-temperature modification is transformed to a monoclinic low-temperature form (space group C2, a = 17.8779(8) A, b = 9.0596(4) A, c = 12.9735(5) A, β = 118.875(13)°, V = 1840.03(22) A3, Z = 18). The monoclinic phase has a symmetry lowered by an index of 18 relative to the hexagonal supergroup. This represents a higher symmetry than the rhombohedral model which has a symmetry of index 36 lower than the hexagonal form determined in a previous study (Li and Zeng, J. Cryst. Growth 2013, 382, 47). The monoclinic system is confirmed by optical investigations with a polarizing microscope showing biaxial negative character and refractive indices nx = 1.615(1), ny = 1.664(1), and nz = 1.665(1), 2Vx = 10.0(2)° as measured on a microrefractometer spindle stage in immersion liquids. Electron microprobe analysis yielded a slight excess of Ba resulting in the composition Ba1.04Ca0.96BO...

© 2016 Elsevier B.V.. The Bushveld Complex (South Africa) is the largest layered intrusion on Earth and plays a considerable role in our understanding of magmatic differentiation and ore-forming processes. In this study, we present new geochemical data for apatite-hosted multiphase inclusions in gabbroic cumulates from the Bushveld Upper Zone. Inclusions re-homogenized at high-temperature (1060-1100 °C) display a range of compositions in each rock sample, from iron-rich (35 wt.% FeOtot; 28 wt.% SiO2) to silica-rich (5 wt.% FeOtot; 65 wt.% SiO2). This trend is best explained by an immiscible process and trapping of contrasted melts in apatite crystals during progressive cooling along the binodal of a two-liquid field. The coexistence of both Si-rich and Fe-rich immiscible melts in single apatite grains is used to discuss the ability of immiscible melts to segregate from each other, and the implications for mineral and bulk cumulate compositions. We argue that complete separation of immiscible liquids did not occur, resulting in crystallization of similar phases from both melts but in different proportions. However, partial segregation in a crystal mush and the production of contrasting phase proportions from the Fe-rich melt and the Si-rich melt can be responsible for the cyclic evolution from melanocratic (Fe-Ti-P-rich) to leucocratic (plagioclase-rich) gabbros which is commonly observed in the Upper Zone of the Bushveld Complex where it occurs at a vertical scale of 50 to 200 m. ispartof: EARTH AND PLANETARY SCIENCE LETTERS vol:443 pages:108-117 status: published

The Bushveld Complex in South Africa is the largest layered intrusion on Earth. Its upper part is known for huge resources of iron, titanium, vanadium and phosphorus. Associated with the layered character of the rocks, these resources are enriched at certain levels of the intrusion, which makes it important to understand the formation processes of those layers. In this paper we give an introduction and overview of recent debates and challenges.

A mineral intermediate between sillimanite and mullite, tentatively designated as “sillimullite,” was studied by electron microprobe analyses and single-crystal X-ray diffraction methods. The chemical compositions derived from the microprobe results and the crystal-structure refinement are Al 7.84 Fe 0.18 Ti 0.03 Mg 0.03 Si 3.92 O 19.96 and Al 8.28 Fe 0.20 Si 3.52 O 19.76 (Fe is Fe 3+ ) corresponding to x -values of 0.02 and 0.12, respectively, in the solid-solution series Al 8+4x Si 4–4x O 20–2x assigning Fe 3+ , Ti, and Mg to the Al site. The composition derived from microprobe analysis is very close to a stoichiometric sillimanite (with Fe 3+ ,Ti, and Mg assigned to Al sites), while the composition derived from diffraction data is midway between sillimanite and Si-rich mullites. The discrepancy is assumed to be caused by the occurrence of amorphous nano-sized SiO 2 inclusions in the aluminosilicate phase not affecting the diffraction data but detected in the microprobe analysis. “Sillimullite” crystallizes in the orthorhombic space group Pnam with a = 7.5127(4), b = 7.6823(4), c = 5.785(3) A, V = 333.88(4) A 3 , Z = 1. It has a complete Si/Al ordering at tetrahedral sites like sillimanite but with neighboring double chains of SiO 4 and AlO 4 tetrahedra being offset by ½ unit cell parallel to c relative to each other causing the change of the space-group setting from Pbnm (sillimanite) to Pnam . Difference Fourier calculations and refinements with anisotropic displacement parameters revealed the formation of oxygen vacancies and triclusters as known in the crystal structures of mullite. Final refinements converged at R 1 = 5.9% for 1024 unique reflections with F o > 4σ( F o ). Fe was found to reside predominantly in the octahedral site and with minor amounts in one of the T* sites. Mg and Ti were not considered in the refinements. The crystal studied here is considered to represent a new mineral intermediate between sillimanite and mullite, named “sillimullite.”

The mixing of magmas is a fundamental process in the Earth system causing extreme compositional variations in igneous rocks. This process can develop with different intensities both in space and time, making the interpretation of compositional patterns in igneous rocks a petrological challenge. As a time-dependent process, magma mixing has been suggested to preserve information about the time elapsed between the injection of a new magma into sub-volcanic magma chambers and eruptions. This allowed the use of magma mixing as an additional volcanological tool to infer the mixing-to-eruption timescales. In spite of the potential of magma mixing processes to provide information about the timing of volcanic eruptions its statistical robustness is not yet established. This represents a prerequisite to apply reliably this conceptual model. Here, new chaotic magma mixing experiments were performed at different times using natural melts. The degree of reproducibility of experimental results was tested repeating one experiment at the same starting conditions and comparing the compositional variability. We further tested the robustness of the statistical analysis by randomly removing from the analysed dataset a progressively increasing number of samples. Results highlight the robustness of the method to derive empirical relationships linking the efficiency of chemical exchanges and mixing time. These empirical relationships remain valid by removing up to 80% of the analytical determinations. Experimental results were applied to constrain the homogenization time of chemical heterogeneities in natural magmatic system during mixing. The calculations show that, when the mixing dynamics generate millimetre thick filaments, homogenization timescales of the order of a few minutes are to be expected.

International audience; At oceanic spreading centers, interactions between magma and hydrothermal convecting systems trigger major physical, thermal, and chemical exchanges. The two-pyroxene hornfels recovered from the base of the sheeted dike sequence at Integrated Ocean Drilling Program (IODP) Site 1256 (equatorial Eastern Pacific) are interpreted as a conducting boundary layer between the underlying axial melt lens and the hydrothermally cooled sheeted dikes. They are cut by numerous small, felsic veins, which were recently interpreted as a product of hydrous partial melting of sheeted dikes. Here, we present trace element compositions of products (melts and residues) of hydrous partial melting experiments using basalts and hornfels from IODP Site 1256 as starting material. The experimental products generated between 910 °C and 970 °C match the natural lithologies from Site 1256 in terms of major and trace element compositions. The compositions of the anatectic melts correspond to the compositions of the felsic veins, while the residual minerals match the compositions of the two-pyroxene hornfels, evidencing that hydrous partial melting is an important magmatic process in the gabbro/dike transition of fast-spreading mid-oceanic ridges. Our results complement previous experimental studies on anatectic processes occurring at the roof of the magma chambers from fast-spreading mid-ocean ridges. Moreover, calculations of mixing and assimilation fractional crystallization using the experimental partial melts as contaminant/assimilant showed that anatectic melts can only be a minor contributor to the contamination process.

International audience; Replenished axial melt lenses at fast-spreading mid-oceanic ridges may move upward and intrude into the overlying hydrothermally altered sheeted dikes, resulting in high-grade contact metamorphism with the potential to trigger anatexis in the roof rocks. Assumed products of this process are anatectic melts of felsic composition and granoblastic, two-pyroxene hornfels, representing the residue after partial melting. Integrated Ocean Drilling Program Expeditions 309, 312, and 335 at Site 1256 (eastern equatorial Pacific) sampled such a fossilized oceanic magma chamber. In this study, we simulated magma chamber roof rock anatectic processes by performing partial melting experiments using six different protoliths from the Site 1256 sheeted dike complex, spanning a lithological range from poorly to strongly altered basalts to partially or fully recrystallized granoblastic hornfels. Results show that extensively altered starting material lacking primary magmatic minerals cannot reproduce the chemistry of natural felsic rocks recovered in ridge environments, especially elements sensitive to hydrothermal alteration (e.g., K, Cl). Natural geochemical trends are reproduced through partial melting of moderately altered basalts from the lower sheeted dikes. Two-pyroxene hornfels, the assumed residue, were reproduced only at low melting degrees (

The Bushveld Complex in South Africa is the largest layered intrusion on Earth. Its upper part is known for huge resources of iron, titanium, vanadium and phosphorus. Associated with the layered character of the rocks, these economically valuable elements are enriched at certain levels of the intrusion. Thus it is important to understand the formation processes of those layers. Using samples from the Bierkraal drill cores, representing the entire Upper and Upper Main Zone in the Western Limb, this thesis gives detailed insights into magmatic processes and the parental magma as well as the prevailing conditions forming the Upper and Upper Main Zone of the Bushveld Complex. A detailed study of modal proportions of the Bierkraal drill cores, contributed to a better understanding of the top part of the intrusion. The observed compositional cyclicity of the Upper and Upper Main Zone was recently explained by several magma injections with a plagioclase-laden magma. Anorthosite layers are produced by crystal settling of the transported plagioclase crystals and magnetitite layers crystallized from a hybrid melt, produced by mixing of resident and injected magma. However, further discussion of the prevailing magma conditions affecting the mineral compositions as well as detailed description of the sub-magmatic system is missing. In this study, the investigation of melt inclusions in apatite from the Upper Zone showed a compositional range from Fe-rich to Si-rich liquids best explained by silicate liquid immiscibility. During late-stage magmatic evolution the liquid line of descent of the Upper Zone reaches the two-liquid field and immiscible melts start to segregate. The continuous range of melt inclusion compositions observed in this study is the result of cooling. Thus, the liquid evolves along the binodal of the two-liquid field, producing more and more contrasting compositions. The dense, low-viscous, Fe-rich liquid percolates downwards and the cumulates from this Fe- and P-rich immiscible melt form the nelsonite layers observed in the Upper Zone. Trace-element distribution in titanomagnetite and clinopyroxene as well as major-element composition of clinopyroxene revealed two major compositional shifts towards more evolved signatures within the Upper Zone stratigraphy. These distinct changes in mineral composition are in correlation with bulk-rock vanadium concentrations. They can be explained by the injection of a more evolved magma. Compared to previously proposed cyclical shifts incomposition, these two events are much more pronounced, arguing against a continuously evolving staging chamber. In fact, this supports the hypothesis of a sub-compartmentalized Bushveld staging chamber as a source for the magma injections. Prevailing oxygen fugacity (fO2) conditions in the Upper Zone were estimated using the V partitioning between titanomagnetite and clinopyroxene. The results show that fO2 variations are minor in the Upper Zone and are approximately around FMQ - 2, which is 1.5 - 2 log units lower than previously assumed. To discuss potential parental magmas of the Upper and Upper Main Zone, crystallization experiments were performed in an internally heated pressure vessel, aiming to reproduce the mineral assemblage above the Pyroxenite Marker. Previously proposed and new calculated compositions were tested at conditions relevant to the Bushveld Complex (2 kbar; 1080°C - 1140°C; ~FMQ -2). In contrast to previously proposed basaltic compositions, the experimental results show that the parental magma to the Upper and Upper Main Zone of the Bushveld complex must be andesitic. Moreover, experimentally produced mineral compositions revealed, that the Upper and Upper Main Zone parental magma contained a residual liquid from the underlying zones high in Mg and Ca also small amounts of H2O (, Der Bushveld Komplex in Südafrika ist die größte Lagenintrusion der Erde. Dessen oberer Teil ist bekannt für seine großen Vorkommen an Eisen, Titan, Vanadium und Phosphor. Dem lagigen Charakter der Gesteine entsprechend, sind diese wirtschaftlich bedeutsamen Elemente in bestimmten Lagen der Intrusion angereichert. Aus diesem Grund ist es wichtig, die Prozesse, die diese Lagen bilden genau zu verstehen. Diese Arbeit gibt einen detaillierten Einblick in die magmatischen Prozesse und die Zusammensetzung des Stamm-Magmas der Upper und Upper Main Zone (UUMZ) und die vorherrschenden Bedingungen bei der Kristallisation. Dafür wurden Proben aus den Bierkraal Bohrkernen genommen, die die gesamte UUMZ im westlichen Teil der Intrusion repräsentieren. Eine detaillierte Studie über den Modalbestand der Gesteinsproben aus den Bierkraal Bohrkernen, hat zu einem besseren Verständnis des oberen Teils der Intrusion beigetragen. Das von Yuan et al. (2017) entwickelte Entstehungsmodell erklärt die Zyklizität, die in Mineralzusammensetzungen beobachtet werden kann, mit mehreren Magma-Injektionen in die Bushveld Magmakammer. Dabei transportiert das injizierte Magma auch Plagioklas-Kristalle, die in der Magmakammer nach unten sinken und so Anorthosit-Lagen bilden. Die ökonomisch wichtigen Magnetit-Lagen kristallisieren aus einer Hybrid-Schmelze aus, die durch Mischung zwischen dem vorhandenden und injizierten Magma entsteht. Schmelzeinschlüsse in Apatit aus der Upper Zone zeigen Zusammensetzungen von Fe-reichen bis Si-reichen Schmelzen. Diese Spanne von Zusammensetzungen kann man am besten mit der Entmischung zweier silikatischer Schmelzen erklären. Während der Kristallisation in der Upper Zone entwickelt sich die Schmelze hin zu einer Mischungslücke hin, in der sich zwei nicht-mischbare Schmelzen voneinander trennen. Die kontinueirliche Reihe der Zusammensetzungen der Schmelzeinschlüssen entsteht während der voranschreitenden Abkühlung, bei der sich die beiden nicht-mischbaren Schmelzen entlang der Binodalen der Mischungslücke entwickeln, und so immer starker gegensätzliche Zusammensetzungen erhalten. Die dichtere, niedrig-viskose, Fe-reiche Schmelze sickert nach unten. Kumulate aus diesen Fe- und P-reichen Schmelzen bilden dann die Nelsonit-Lagen in der Upper Zone. Spurenelemente in Ti-Magnetiten und Klinopyroxenen sowie die Hauptelement-Zusammensetzung der Klinopyroxene zeigen zwei abrupte Veränderungen zu mehr entwickelten Zusammensetzungen im stratigraphischen Verlauf der Upper Zone. Diese markanten Sprünge in der Mineralzusammensetzung korrelieren mit Vanadium Konzentrationen im Gesamtgestein. Im Vergleich zu den bisher gezeigten Änderungen in Mineralzusammensetzungen die mit der Magma Injektion einhergehen (Yuan et al.,2017), sind diese viel ausgeprägter und sprechen gegen eine sich kontinuierlich entwickelte sub-Bushveld Magmakammer. Vielmehr spricht die Veränderung zu starker entwickelten Mineralzusammensetzungen für eine weiter unterteilte sub-Bushveld Magmakammer als Quelle für die Magma-Injektionen. Die Sauerstofffugazität (fO2) und deren Entwicklung in der Upper Zone wurde über die Vanadium-Verteilung zwischen Magnetit und Klinopyroxen bestimmt. Die Ergebnisse zeigen, dass fO2 in der Upper Zone nur gering variiert und sich im Bereich von FMQ -2 bewegt. Dies ist 1.5 bis 2 Log-Einheiten unterhalb der bisher angenommen Sauerstofffugazitäten für die Upper Zone. Potentielle Stamm-Magmen der UUMZ wurden über Kristallisationsexperimente in intern beheizten Gasdruck Anlagen getestet. Das Ziel dieser Experimente war es, die ersten Kumulate der UUMZ direct oberhalb des Pyroxenit Markers zu reproduzieren. Hierzu wurden publizierte und neu berechnete Zusammensetzungen bei für den Bushveld Komplex relevanten Bedingungen getestet (2 kbar; 1080°C - 1140°C; ~FMQ -2). Im Gegensatz zu den bisherigen Annahmen, dass das Stamm-Magma eine basaltische Zusammensetzung hat, zeigen die Experimente, dass nur eine andesitische Zusammensetzung die Mineralvergesellschaftung oberhalb des Pyroxenit Markers reproduziert. Die Zusammensetzung der experimentellen Mineralphasen zeigt ausßerdem, dass das Stamm-Magma der UUMZ anteilig aus der Restschmelze der darunter liegenden Zone bestehen muss und zudem geringe Mengen an H2O enthält (< 1wt%).