The fate of zircon during polyphase granulite facies metamorphism in Rogaland, South Norway

Get Complete Project Material File(s) Now! »

Monazite zoning and micro-chemistry

Monazite crystals were studied both in thin section, which preserves their textural context, and in mineral separates recovered by standard procedures, including crushing, heavy liquids, and magnetic separation. They were examined using transmitted and reflected light microscopy, followed by Scanning Electron Microscopy (SEM) using a Jeol 6360LV equipped with a Bruker XFlash 5010 SDD detector (EDS), which allowed for quantitative phase mapping. Back Scattered Electron (BSE) images were taken at 20 kV with the Jeol 6360LV SEM to guide in-situ chemical analysis acquired using a Cameca SX-FIVE electron probe micro-analyser (EPMA) hosted at the Raimond Castaing micro-characterization centre (Toulouse). The EPMA was operated with a focused beam at standard operating conditions of 15 kV and 20 nA using a TAP–LLIF–PET– LPET configuration for point analyses. Selected grains were subsequently mapped for their minor elements, including Si–Th–Ca–S and Y–U–Th–Pb using a TAP–LPET–PET–LPET configuration. Element maps were generated at 15 kV and 200 nA with a step–size of 1 μm, and a dwell time of 1 s. ZAF matrix corrections were calculated using fixed REE and P content, which were chosen from point analysis. The following X–rays and mineral standards were used: Y L 4, Th M 2, S K 4, Si K La L -free (REE)PO4, Pb M 2P2O7 U M 2.
Background positions were carefully selected after acquiring a Wavelength Dispersive Spectrometer (WDS) intensity spectrum on the area of interest. In particular, the S K was carefully chosen at (-1800; +500) in order to avoid any interference (Fig. 2). A selected 95% confidence level in analytical errors for point analyses, calculated following the formula of Ancey et al. (1977), are as follows: 0.08 wt% for SO2, 0.06 wt% for CaO, 0.7 wt% for ThO2, and 0.25 wt% for both SiO2 and Y2O3. Monazite formula were calculated in the system 2(REE)PO4 – CaTh(PO4)2 –2ThSiO4, utilizing the following sequence: (1) Ca + Th + U are assigned to the cheralite [CaTh(PO4)2] component, (2) any remaining Th is combined with Si are assigned to the huttonite [ThSiO4] component, (3) Y and HREE (Gd–Lu) are assigned to the xenotime [(HREE,Y)PO4] component, and (4) the light rare earth elements (La–Eu) are assigned to the monazite [(LREE)PO4] component.

Monazite texture, composition and inclusions

Garnet-rich layers in sample ALR 13-58 contain large (up to 900 μm) monazite grains clouded with solid inclusions. The monazite grains make up to ~ 1 % of the volume of the mineral assemblage (Fig. 3b,d). These grains are anhedral and usually found in the matrix or partly included within peak metamorphic minerals such as orthopyroxene or green spinel and are eventually overgrown by retrograde garnet II rims (Fig. 3b). In addition, BSE mapping of thin sections reveals small monazite grains (< 15 μm) hosted in garnet I. These inclusions were too small to be analysed by LA–ICP–MS and are not discussed further.

Nano-characterization of monazite by TEM

Three FIB foils were cut in one large monazite crystal (1029-48) in the three compositional domains (D1, D2, D3) for detailed inspection. The foil in D1 was cut in the S-rich core that is free of inclusions. Imaging in STEM mode and dark-field (with the HAADF detector) were used to reveal density contrasts. Low magnification images (Fig. 10a) show black dots, representing a negative density contrast, homogenously distributed all over the foil, that represent a negative density contrast in monazite. These dots are 5–10 nm large with an equant shape (Fig.10b–c) and distributed in a short period modulation of ~15–25 nm. Energy dispersive spectroscopy (EDS) point or line scans acquisitions across the black dots reveal that they are enriched in Ca + S and depleted in Ce + P compared with the host monazite (Fig. 10b). Bright-field imaging (Fig. 10d) qualitatively show little or no lattice misorientation between the (Ca + S)-rich nanoclusters and the host monazite. Inspection of the foil cut in SO2-bearing D2 (Fig. 10e) did not reveals (Ca + S)-rich nanoclusters but contains a polymineralic inclusion. Examination of this inclusion by TEM coupled with EDS mapping shows that euhedral pyrite is associated with iron oxide (hematite), apatite, and a phyllosilicate tentatively identified as celadonite (Fig. 10f). The boundary between the monazite and the phyllosilicate is irregular. The foil from the SO2-free D3 is homogenous, i.e. devoid of (Ca + S)-rich nanoclusters and other inclusions. Only some porosity is observable (not shown).

Mechanism of S incorporation in monazite

Investigation of spatial distribution of S and Ca in monazite (Fig. 4) complemented by EPMA point analyses in S-bearing domains D1 and D2 (Fig. 6) indicate that S is accommodated as sulphate through the anhydrite substitution mechanism Ca2+ + S6+ = REE3+ + P5+. This substitution vector has been already proposed by Kukharenko et al. (1961) and later confirmed by several workers on the basis of EPMA analyses (Chakhmouradian and Mitchell 1999; Ondrejka et al. 2007; Krenn et al. 2011). Because Sr is only present at the trace level, the Sr–Ca substitution in the anhydrite component is negligible. High-resolution TEM investigations reveal 5–10 nm nanoclusters composed of CaSO4 with a coherent interface relative to the host monazite (in D1; Fig. 10c–d). In a theoretical perspective, the presence of clino-anhydrite in monazite is not surprising as CaSO4 crystallizes in the monazite-type structure (P21/n) at high pressure (> 2 GPa ; Crichton 2005; Ma et al. 2007; Bradbury and Williams 2009). Moreover, size wise, [SO4]2- and [ClO4]- closely resemble [PO4]3-, and Ca2+ is the cation closest to Ce3+ (Shannon 1976). Two contrasting interpretations of such nanoclusters may be proposed. The first one involves the presence of primary heterogeneities incorporated during the crystallization of monazite. Such nanoclusters heterogeneities (c. 5–50 nm) are frequently observed during mineral synthesis when annealing duration is too short (e.g. in Ca–Pb fluoro-vanadinite apatites, Dong and White 2004), and reflect disequilibrium conditions and/or heterogeneous crystalizing medium. The second possibility is the presence of exsolution of nanophases (here CaSO4) due to homogeneous nucleation. Indeed, the regularity in shape, the short period modulation (~ 15–25 nm; Fig. 10a) of the CaSO4 nanoclusters, together with a coherent interface, rather point to exsolution by homogeneous nucleation. Such phase separation is documented for instance in apatite where nanoclusters (5–10 nm) of ellestadite (S-rich apatite) developed in short period modulation (Ferraris et al. 2005). The second mechanism is the preferred one for our monazite sample and implies the possible existence of a miscibility gap. Experimental data could bring useful geothermometric tools to unravel any temperature or pressure dependence on the incorporation of clino-anhydrite in monazite. With this respect, the absence of clino-anhydrite nanoclusters in the S-bearing D2 monazite investigated by TEM (SO2 ~ 1500 ppm) may be interpreted as a solubility limit.

Mineral composition and phase equilibria modelling

Phase equilibria of metamorphic assemblages were modelled using P–T and T–X pseudosections in the Na2O–CaO–K2O–FeO–MgO–Al2O3–SiO2–H2O–TiO2–O2 chemical system using Perple_X (version 6.7.2; Connolly 2009). All samples were modelled using the internally consistent thermodynamic database of Holland and Powell (2011) together with activity– composition models of White et al. (2014) except for the osumilite gneiss (ALR 13-58). For this sample we used the Kelsey et al. (2004) database for UHT rocks in conjunction with updated models of spinel from White et al. (2007) and osumilite from Holland et al. (1996). The phases under consideration are garnet, silicate melt, plagioclase, K-feldspar, sillimanite, spinel, magnetite, ilmenite–hematite, rutile, orthopyroxene, sapphirine, cordierite, biotite, muscovite, quartz with additional osumilite in the Kelsey et al. (2004) database.
The bulk rock composition have been determined by ICP–OES on the rock chips left over after thin section preparation at the CRPG (Nancy), following standard procedure described in Carignan et al. (2001). FeO has been measured by titration to estimate the oxidation state of the whole rock. Oxidation ratio measured from the bulk rock was tested with T–X sections and found suitable to model the observed paragenesis for all samples. Modelled H2O content were constrained using T–X sections considering all loss on ignition (LOI) as H2O and H2O = 0.1 mol %. Because our aim is to model the evolution of the rock at the peak T conditions, we chose H2O value so-that the solidus is the closer to the interpreted peak field. A value of 1 mol. % H2O was found suitable for all samples. Chemical composition used for modelling are reported as oxides weight percentage along with measured compositions in Tab. 1.
Quantitative analysis of silicate and oxides were collected with the Cameca SX-Five microprobe with standard operating conditions of 15 kV and 20 nA. Representative mineral analyses are reported for sample ALR 13-64, 13-05, 13-06, 13-22, 13-58; 14-19 in supplementary material S3-1, S3-2, S3-3, S3-4, S3-5 and S3-6 respectively The mineral formula of orthopyroxene (Opx) was recalculated on the basis of 4 cations with Fe3+ estimated by normalizing the analyses to 6 O. Garnet (Grt) and feldspar formula were calculated on the basis of 8 cations. Minerals belonging to the magnetite–ulvospinel–spinel–hercynite solid solution were calculated on the basis of 3 cations and 4 oxygens. Biotite (Bt) analyses were recalculated on the basis of 11 O. Cordierite (Crd) formula was calculated on the basis of 18 O. Osumilite (Osm) was calculated on the basis of 30 cations following Das et al. (2001). Sapphirine (Spr) formula was calculated on the basis of 7 cations and normalized to 10 O to evaluate Fe3+ substitution. All mineral abbreviations follow Whitney and Evans (2010).

READ  Atmospheric composition from absorptive VUV stellar occultations

Monazite–xenotime Y and REE thermometry

Thermometry based on the partitioning of Y and REE between monazite and xenotime was applied in one monazite–xenotime bearing sample (ALR 13-22). Because the original experimental calibration of the thermometer, performed in the simple CePO4–YPO4 system revealed only small pressure dependence (Gratz and Heinrich 1997), it has been neglected in the present case study. However, Seydoux-Guillaume et al. (2002b) showed that incorporation of Th through the huttonite (ThSiO4) substitution enhances Y solubility in CePO4 and proposed an improved calibration curve together with a phase diagram in the CePO4–YPO4–ThSiO4 system. For that reason, the temperature calculated with the Gratz and Heinrich (1997) calibrations should be treated as maximum temperature. Conversely, the empirical calibrations of Pyle et al. (2001) and Heinrich et al. (1997) have the advantage to integrate the full range of REE distribution for typical monazite in metapelitic rocks and a moderate content of ThO2 (< 10 Wt. %) but are restricted to T < 750 °C and garnet-present samples. We performed temperature calculation on the basis of monazite EPMA analyses (Tab. 2), which were acquired prior to laser ablation U–Th–Pb analyses. The four calibrations of Gratz and Heinrich (1997), Heinrich et al. (1997), Pyle et al. (2001) and Seydoux-Guillaume et al. (2002b) are presented in Tab. 2. The calibration of Seydoux-Guillaume et al. (2002b) is preferred for comparison purpose and geological interpretation since the associated phase diagrams explicitly take into account the Th-component in monazite and because the experiments were performed up to 1000°C, i.e. at UHT conditions.

Table of contents :

Chapitre 1 Overview of the Sveconorwegian orogeny and Mesoproterozoic
evolution of Rogaland, S-Norway
Résumé
Framework of Rodinia assembly
The Sveconorwegian orogeny
1280–1150 Ma time interval (pre-Sveconorwegian)
1150–1080 Ma time interval (Arendal phase)
1050–1030 Ma time interval (Agder phase)
1030–1000 Ma time interval (Agder phase)
990–970 Ma time interval (Falkenberg phase)
970–950 Ma time interval (Dalane phase)
950–920 Ma time interval (Dalane phase)
Tectono-magmatic evolution of Rogaland
Magmatism
Metamorphism
Chapitre 2 Sulphate incorporation in monazite lattice and dating the cycle of sulphur in metamorphic belts
Résumé
Abstract
Introduction
Monazite crystal chemistry
Geological background
Analytical methods
Monazite zoning and micro-chemistry
Monazite U–Th–Pb geochronology
Transmission electron microscopic (TEM) imaging
Results
Sample petrology
Monazite texture, composition and inclusions
Nano-characterization of monazite by TEM
U–Th–Pb geochronology
Discussion
Mechanism of S incorporation in monazite
Significance of S-rich monazite
Metasomatic replacement of S-rich monazite
Dating S mobility in metamorphic belts
Conclusion
Acknowledgement
References
Supplementary materials
Chapitre 3 Two cycles of ultra-high temperature metamorphism in Rogaland, S. Norway: critical evidence from monazite Y-thermometry & U–Pb geochronology
Résumé
Abstract
Introduction
Geological setting.
Methods
Micro-chemistry
U–Th–Pb geochronology
Mineral composition and phase equilibria modelling
Monazite–xenotime Y and REE thermometry
Microstructures, petrography and mineral compositions
Orthopyroxene zone
Osumilite zone
Pigeonite zone
Phase equilibria modelling and textural interpretation
Orthopyroxene zone
Osumilite zone
Pigeonite zone
Monazite–xenotime chemistry and monazite U–Th–Pb geochronology
Orthopyroxene zone
Osumilite zone
Pigeonite zone
Discussion
A monazite based temperature–time path
Two phases of UHT metamorphism
Conclusion
References
Supplementary material
Chapitre 4 The fate of zircon during polyphase granulite facies metamorphism in Rogaland, South Norway
Résumé
Abstract
Introduction
Geological setting.
Analytical methods
U–Th–Pb geochronology
Chemical micro-analyses
Scanning ion imaging
Samples background
Results
Zircon zoning and microchemistry
Zircon U–Pb geochronology
Comparing monazite and zircon age record through time and space
Zircon oxygen isotopes
Scanning ion Imaging
Discussion
Response of O isotopes
Processes of zircon U–Pb partial resetting
Insight from Pb distribution
Tracking melt-present conditions in slow granulite
Conclusion
References
Supplementary material
S4-2: Phase equilibria modelling for sample ALR 13-69
S4-3: Monazite chemistry and U–Th–Pb geochronology for sample ALR 13-69
Chapitre 5 Discussion of temperaturetime evolution of Rogaland and plausible heat sources for UHT metamorphism
Résumé
Abstract
Introduction
Models for UHT metamorphism
P–T–t–D paths of the Rogaland
Geometrical relationships
Significance of the Opx-isograd
Timing of deformation and vertical movement
Interplay between magmatism and metamorphism
Timescale of AMC emplacement
Synthetic T–t diagram deduced from metamorphic rocks
Interplay between magmatism and metamorphism
Radiogenic heat production of the crust
Whole rock geochemistry
Redistribution of U and Th
Geodynamic speculation and conclusion
References
Supplementary materials
Conclusion et perspectives
Apports à la compréhension du comportement de la monazite au cours du métamorphisme de UHT
Apports à la compréhension du comportement du zircon au cours du métamorphisme de UHT
L’évolution du Rogaland et causes du métamorphisme de UHT
La monazite : traceur des minéralisations ?
Perspectives
Références

GET THE COMPLETE PROJECT

Related Posts