The peralkaline F-rich igneous pluton of the Serra de Monchique complex in southern Portugal shows various REE- and HFSE-bearing mineral assemblages in variously evolved and hydrothermally overprinted rock units. Based on mineral compositions and field relations, the differentiation history of the complex will be elucidated with special focus on partitioning of halogens, REEs and HFSEs.

The complex consists of two main units: an inner core unit of homogeneous nepheline to sodalite syenite and an outer roof unit of heterogeneous nepheline and sodalite to quartz-(monzo)syenites and breccias. Faults separate the two main units. The complex is crosscut by lamprophyric and phonolitic dyke rocks.

Alkali feldspar, nepheline and sodalite are the dominant felsic minerals in all units. Plagioclase occurs sparsely. Clinopyroxene following the diopside-aegirine trend (Di₈₀Hd₁₂Ae₀₅ – Di₀₂Hd₀₉Ae₈₉) is the dominant mafic mineral together with biotite, locally occurring amphibole, magnetite and ilmenite. Titanite (up to 0.158 LREE³⁺ apfu) and fluorapatite (Fap₄₉Hap₄₇Clap₀₄ – Fap₁₀₀Hap₀Clap₀ with up to 0.664 LREE³⁺ apfu) occur as accessory phases throughout the complex, whereas fluorite and zircon occur as traces. Late magmatic phases include HFSE-bearing silicates of the wöhlerite- and rinkite-groups, documenting a transition from miaskitic to agpaitic phase assemblages during differentiation.

Alteration by hydrothermal fluids affected the whole complex. Aqueous fluids resulted in significant growth of zeolites, carbonate-bearing fluids formed calcite and cancrinite, mostly at the expense of nepheline. Breakdown of HFSE- and F-bearing silicates by carbonate-bearing fluids led to the growth of secondary fluorite, calcite and zircon, whereas breakdown of titanite caused formation of calcite and anatase.

Due to low-degree partial melting of metasomatized mantle sources, the northern Tanzanian segment of the East African Rift hosts volcanoes that produced diverse rock series, ranging from basaltic to trachytic, and from nephelinitic ±melilititic to phonolitic compositions. Some of the latter group also produced carbonatites (Oldoinyo Lengai, Kerimasi, Mosonik, Shombole), while others (Sadiman and Burko) did not.

We present a petrological study of alkaline silicate rocks and associated carbonatites from Kerimasi, Mosonik, and Shombole, and compare them with the nearby Sadiman, Burko, and Oldoinyo Lengai volcanoes. The silicate rocks include olivine-melilitites, various nephelinites and phonolites, and their plutonic counterparts. Whole-rock X_Mg (Mg/(Mg+Fe)) values decrease from ~0.7 in melilitites via 0.6–0.5 in nephelinites to <0.2 in phonolites, indicating progressive magma evolution. Simultaneously, CaO, FeO, TiO₂, and P₂O₅ decrease, while alkalis, Al₂O₃, and SiO₂ increase; in some cases, peralkaline compositions ((Na+K)/Al >1) are reached. Mineral assemblages change from åkermanite +forsterite +perovskite +Fe-Ti oxides in primitive rocks, via diopside +andradite +nepheline in intermediate rocks, to aegirine-augite +nepheline +alkali feldspar +titanite ±sodalite-aenigmatite-eudialyte in evolved rocks. Carbonatites are mostly calciocarbonatites that differ in amount and composition of silicate phases. Apatite, magnetite, mica, and clinopyroxene are present in all volcanoes. Olivine, monticellite, perovskite, periclase occur only at Kerimasi, while Mosonik and Shombole typically contain nepheline, titanite, K-feldspar.

These textural, chemical, and mineralogical variations reflect different storage conditions, cooling histories, as well as heterogeneous mantle sources and partial melting processes, which we aim to constrain by estimating temperature, pressure, redox state, silica activity, and CaO activity.

Plutonic bodies are solidified magma reservoirs, and their chemical and isotopic characteristics represent the integration of magmatic and sub- solidus processes operating across spatial and temporal scales during pluton construction, crystallization, and cooling. Disentangling these processes and understanding where chemical and isotopic signatures were acquired requires the combination of multiple tools trac­ing processes at different time and length scales.

Here, we combine whole rock and mineral major, trace element, and isotopic compositions with high-precision U-Pb ID-TIMS zircon geochronology to evaluate differentiation processes and their timescales in the bimodal (gabbro and granite), shallow crustal Guadalupe Igneous Complex, Sierra Nevada, USA.

We show that the complex was constructed in ~300 kyr. Pluton-wide δ¹⁸O_{(whole-rock),} δ¹⁸O_(zircon), and Sr- Nd isotopic ranges are too large to be explained by in situ, closed- system differentiation, instead requiring open- system behavior at all scales. Low δ¹⁸O_(whole-rock) and δ¹⁸O_(zircon) values indicate assimilation of hydrothermally altered marine host rocks during ascent and/or emplacement. In-situ differentiation processes operated on a smaller scale (meters to tens of meters) for at least ~200 kyr via (1) percolation and segregation of chemically and isotopically diverse silicic interstitial melt from a heterogeneous gabbro mush; (2) crystal accumulation; and (3) sub- solidus, high- temperature, hydrothermal alteration at the shallow roof of the complex to modify the chemi­cal and isotopic characteristics.

Whole-rock and mineral chemistry in combination with geochronology allows deciphering open-system differentiation processes at the outcrop to pluton scale from magmatic to sub-solidus temperatures over time scales of hundreds of thousands to millions of years.

We present in-situ U-Pb and U-Th ages of zircon inclusions from core to rim in a single 8 cm-large K-Feldspar megacryst from the 32.9 ka dacite dome eruption of Taápaca volcanic complex (Central Andes, Chile). These large sanidine crystals formed in a rhyodacite magma and erupted in a hybrid dacite after magma mixing just prior to ascent. Zircon in the core (n = 13) are in U-Th secular equilibrium with U-Pb dates of 530±35 to 950±35 ka. Zircons in the mantle (n = 10) are again in U- Th secular equilibrium with U-Pb ages of 333±13 to 480±43 ka. Only four zircons yielded U-Th isochron ages of 302 (+55/-36) to 339 (+70/-42) ka (n = 4). Rim-hosted zircons (n = 16) are all significantly younger with U-Th ages of 58 (+12/-11) to 251 (+112/-54) ka. Thus, zircon continuously crystallized and became successively trapped during sanidine growth in the rhyodacite host magma over several hundreds of thousands of years (from 920 ky to 25 ky before eruption), with a maximum individual zircon growth time of 30–50 ky. These results agree with Ba-diffusion timescales for the host K-feldspar and other megacrysts from the same dome eruption. Our combined results indicate at least 500 ky of megacrystic growth at magmatic conditions of 740 to 820 °C and 1 to 2 kbar (thermobarometry on amphibole inclusions in the sanidine) and protracted “hot” magma storage at upper crustal levels in continental arc settings

The redox state of mantle-derived primary magmas and their derivative liquids has been extensively studied over the past decades, yet remains a challenging topic. Insights into magma redox conditions are primarily derived from analyses of iron speciation in pillow glasses or mineral phases. Among the various techniques available to determine Fe³⁺/Fe²⁺ ratios, the EPMA-based flank method stands out for offering high spatial resolution and accuracy at relatively low cost. Nevertheless, its routine application is still limited by the requirement of well-calibrated standards and a precision that is generally lower than that of XANES or Mössbauer spectroscopy.

In this contribution, we present further improvements of the flank method approach using a JEOL JXA-iHP200F EPMA. To address this, we synthesized a set of new dry and H₂O-bearing basaltic to andesitic glass standards ranging in Fe²⁺/∑Fe from ~0 to 1. In addition, we used natural boninitic and fore-arc basaltic pillow glasses (IODP EXP 352) as well as Shatsky Rise pillow glasses to test the beam stability of different sample types and investigate their performance.

We demonstrate that, while dry basaltic glasses can endure high beam energies, hydrous glasses only endure about a tenth of beam energies before alteration. The precision in Fe²⁺/∑Fe ratio obtained for dry basaltic glasses (~ ±0.05) is significantly lower compared to hydrous glasses (~ ±0.12) due to their higher stability allowing longer measurement times and stronger beam currents. Our flank method results for dry basaltic glasses generally agree with previously obtained XANES data of the same samples.