Allanite behaviour during incipient melting in the southern Central Alps

The response of allanite to incipient melting was investigated in migmatites from the Tertiary Barrovian-type sequence of the Central Alps (southern Switzerland, northern Italy). Inheritance and new mineral growth were recorded in composite allanite grains sampled from meta-granitoids and leucosomes. Ion microprobe (SHRIMP) dating of high Th/U allanite cores in meta-granitoids yield Permian ages consistent with magmatic crystallisation dating protolith intrusion. In contrast, low Th/U allanite overgrowths and weakly-zoned allanite in meta-granitoids and leucosomes yield Alpine U–Pb intercept ages between 30 ± 4 and 20 ± 5 Ma; these date allanite formation during the Barrovian cycle. Major and accessory mineral REE compositions suggest that Alpine allanite crystallised in the presence of a low-temperature melt. Whereas new zircon growth is rare in the migmatites, allanite readily recorded growth during the Alpine cycle. Allanite U–Th–Pb isotopes may therefore present a complementary approach to zircon for dating low-temperature partial melting, where the preservation of allanite is aided by low LREE solubility in hydrous granitic melt. The Th–Pb age is preferred to date high-Th magmatic allanite, however the U–Pb and Th–Pb ages of allanite overgrowths may differ (by up to 25%), and this demands a comparison of both U–Pb and Th–Pb isotopic systems to obtain a best estimate for the timing of low-Th allanite crystallisation. Protolith allanite preserves a substantial memory of its initial age in spite of upper amphibolite facies re-working during migmatisation (T = 620–700 °C), which places strong constraints on Pb closure temperature. Magmatic allanite contains <30% of initial (non-radiogenic) Pb, allanite in migmatites is characterised by an initial Pb of 20–70%, and subsolidus allanite has >60% of initial Pb. Therefore, the initial Pb may be useful as a sensor for the amount of melt present during allanite formation. The Pb isotope composition of allanite overgrowths indicates ¿5% inherited radiogenic Pb from precursor allanite, which suggests efficient redistribution and homogenisation of Pb isotopes during the Alpine partial melting period.




--------------------------------------------------------------------------------

1. Introduction

The epidote-group mineral allanite occurs most commonly as an accessory phase in metaluminous granites (Gromet and Silver, 1983) and in low- to medium-grade regional metamorphic terranes (Finger et al., 1998 and Wing et al., 2003), and is an important host of thorium (Th), uranium (U) and light rare earth elements (LREE) in both types of occurrences (Gromet and Silver, 1983 and Janots et al., 2008). Such features have made allanite a prospective U–Th–Pb chronometer in magmatic (Barth et al., 1994 and Oberli et al., 2004) and metamorphic rocks (Catlos et al., 2000, Parrish et al., 2006 and Gregory et al., 2009a) as well as an important mineral-scale tracer of geochemical processes involving the REEs (Hermann, 2002, Oberli et al., 2004 and Gregory et al., 2009a). As in other U-Th-bearing minerals, allanite can be heterogeneous in chemistry and texture and thus the development of high-resolution dating protocols specific to this mineral (Catlos et al., 2000 and Cox et al., 2003), in particular using the Sensitive High Resolution Ion Microprobe (SHRIMP, Gregory et al., 2007), has enhanced its use in geochronological studies by providing texturally-sensitive information and the ability to target individual mineral zones.

Dating of allanite has proven most useful for samples in which zircon, the most commonly used U–Pb chronometer, is rare or inherited (e.g. von Blanckenburg et al., 1992). For example, allanite forms readily during subsolidus metamorphism (Smith and Barreiro, 1990, Wing et al., 2003, Janots et al., 2008, Krenn and Finger, 2009 and Rasmussen and Muhling, 2009), and the phase relationships of allanite and monazite can be used to date prograde mineral reactions in Barrovian-type sequences (Janots et al., 2009). Allanite may also form as a high-pressure mineral replacing monazite (Janots et al., 2006 and Gabudianu et al., 2009) and zoisite (Tribuzio et al., 1996, Hermann, 2002 and Spandler et al., 2003), making it a target for dating subduction-related metamorphism (Parrish et al., 2006; Rubatto et al. 2008; Gabudianu et al., 2009). These studies have shown that under subsolidus conditions allanite may retain its isotopic signature with subsequent heating (Janots et al., 2009) or decompression (Parrish et al., 2006; Rubatto et al., 2008; Gabudianu et al., 2009). However, significant fluid interaction and deformation can expose allanite to open-system isotopic behaviour, even at low metamorphic temperatures (Gabudianu et al., 2009).

In comparison, allanite formation in high-grade rocks, and the response of U–Th–Pb isotope systematics within allanite to superimposed metamorphism and anatexis, is poorly understood. Allanite can newly grow during upper amphibolite facies events, providing direct ages of metamorphism (Gregory et al., 2009a). Although there are no diffusion experiments to provide a direct quantitative basis for the closure temperature of allanite in metamorphic rocks, estimates have been derived from other empirical data. Heaman and Parrish (1991) dated allanite from high grade gneisses by ID-TIMS and found that when metamorphosed to second sillimanite zone, protolith allanite still preserved a component of its original age. Parrish (2001) concluded based on U–Pb isotope systematics that the closure temperature of allanite is likely ~650 °C. Isotopic inheritance in the form of inherited cores has not been reported for allanite. Owing to its tendency to incorporate Pb on crystallisation, however, allanite may inherit radiogenic Pb from precursor minerals (Romer and Siegesmund, 2003), thereby complicating the correction for non-radiogenic Pb in U–Th–Pb isotope analyses (Gabudianu et al., 2009). In order to interpret ages obtained from allanite in high grade rocks, it is thus important to understand the mechanisms of isotopic resetting and new mineral growth, such as diffusion versus overgrowth processes that affect allanite. In particular, overgrowth processes can be problematic for geochronology unless microanalytical techniques are used (i.e. SHRIMP). Barth et al. (1994) observed that the Th–Pb and U–Pb age systems of magmatic allanite may differ, and suggested that these two geochronometers are decoupled from each other. There are few studies providing explanations of this feature (inherited Pb, Romer and Siegesmund, 2003; excess 206Pb, Oberli et al., 2004), or how these systems behave in metamorphic allanite with low Th/U (Gabudianu et al., 2009). All of these effects have significant implications for allanite geochronology.

This paper details the response of allanite to incipient melting and its performance as a chronometer by examining its occurrence, trace element composition, and age in migmatites from the Barrovian-type sequence of the Central Alps (southern Switzerland, northern Italy). The Central Alps have provided previous opportunities to study the isotope systematics of igneous and subsolidus allanite (Oberli et al., 2004 and Janots et al., 2009). Here they are the ideal setting to examine allanite in high-grade rocks whose protolith was allanite-bearing (Berger et al., 2008), and where the young age of Alpine metamorphism enables multiple allanite crystallization events to be recorded using SHRIMP and LA-ICPMS techniques (Gregory et al., 2007). Allanite has high and variable trace element contents and is therefore susceptible to matrix-induced variations in ion microprobe 206Pb+/238U+ and 208Pb+/232Th+ analyses (Catlos et al., 2000) similar to xenotime (Fletcher et al., 2004) and monazite (Stern and Berman, 2000 and Fletcher et al., 2010). Matrix effects are a potential problem in this study because the ages of the rocks (Permian and Tertiary) require 208Pb/232Th and 206Pb/238U ratios. Gregory et al. (2007) demonstrated that magmatic allanite can be dated accurately (2–3% ± 2s) for samples with LREE + Th of 0.5–0.9 cpfu. Potential complications for allanite dating caused by matrix effects, initial radioactive disequilibrium and the composition of non-radiogenic Pb are discussed in the context of both Th–Pb and U–Pb age systems in magmatic and metamorphic grains. The new allanite ages obtained in this study provide constraints on the closure temperature of allanite and enable the timing of migmatisation to be determined in samples for which new zircon growth is rare.

2. Geological setting

The Central Alpine orogen of southern Switzerland and northern Italy underwent Barrovian-type metamorphism in the mid-Tertiary following continental collision of Africa with Europe. Regional metamorphism in the Central Alpine Lepontine domain is delineated by Barrovian mineral zones, which indicate an increasing metamorphic grade to the south from sub-greenschist to upper-amphibolite facies conditions (Engi et al., 1995 and Todd and Engi, 1997). This Barrovian-type metamorphic sequence terminates at the Insubric fault line (Schmid et al., 1989), a major tectonic lineament through the Alpine orogen. This fault juxtaposes the high-grade metamorphic core of the Lepontine domain against the units of the Southern Alps, which underwent only weak metamorphism during the Alpine orogeny (Fig. 1). The current tectonic architecture is derived from the south-vergent subduction of the European continental margin beneath the Apulian plate and subsequent lithospheric uplift (Schmid et al., 1996).



Fig. 1. Schematic geological map of the Central Alps with the sample locations (after Engi et al., 2004). Solid lines are isograds determined from metamorphic mineral assemblages (Engi et al., 2004). Limit of migmatisation based on field observations by Burri et al. (2005). VA = Val d’Arbedo, BE = Bellinzona, GO = Golino, BZ = Berzona, N = Novate leucogranite, LM = Lago Maggiore.

Figure options




Peak metamorphic conditions of 685 ± 50 °C and 0.6–0.8 GPa, determined by amphibole-plagioclase thermobarometry on granitic leucosome from the Southern Steep Belt (SSB) (Burri et al., 2005), were reached immediately north of the Insubric Line (Fig. 1). The SSB is a highly deformed rock package comprising poly-deformed Variscan basement intruded by Permian granitoids; now metamorphosed granitic gneisses (Schaltegger and Gebauer, 1999). The pre-Alpine units are intercalated at the metre to kilometre-scale with Alpine units consisting of marbles, calc-silicate and metasedimentary rocks, mafic to ultramafic rocks (Engi et al., 1995, Burri et al., 2005 and Berger et al., 2005) and high-pressure relicts (Gebauer, 1996). Consequently, workers have interpreted the SSB to represent part of a tectonic mélange zone (Engi et al., 2001).

Partial melting associated with Alpine metamorphism was dominantly fluid-assisted, and limited to amphibolite-grade rocks of the SSB (Burri et al., 2005). In situ melting of granitic gneisses occurred in the mid-crust at temperatures close to the wet granite solidus (~630 °C) and in some cases produced up to ~30 vol.% leucosomes (Burri et al., 2005). At outcrop, leucosomes are variably deformed and heterogeneously distributed even within a common protolith (Burri et al., 2005). Only relatively small melt volumes were produced by hydrate-breakdown reactions in muscovite-bearing rocks (Burri et al., 2005). This led Burri et al., 2005 and Berger et al., 2008 to infer a causal relationship between mid-crustal deformation, fluid flow, and migmatisation within the SSB.

The nature of incipient Alpine melting and complex association of high-grade rocks in the SSB has complicated geochronology and the determination of Alpine ages (Romer et al., 1996). The long-accepted age for peak metamorphism in the SSB (~28 Ma, Engi et al., 1995) is broadly coincident with igneous activity related to the emplacement of the Bergell Pluton (Berger et al., 1996 and Oberli et al., 2004). Zircon and allanite in the western segment of the Bergell tonalite record a protracted magmatic history lasting from 33 to 28 Ma (Oberli et al., 2004). Accessory mineral ages from syn- and post-kinematic pegmatites and aplites within the SSB also span several million years, from 29 to 25 Ma (Schärer et al., 1996, Romer et al., 1996 and Liati et al., 2000). The age of related amphibolite-grade metamorphism, determined by Sm–Nd analysis of garnet, falls within this period (~27 Ma, Vance and O’Nions, 1992). In fact, Alpine igneous activity continued to at least ~24 Ma with the emplacement of the Novate leucogranite stocks (Liati et al., 2000). Together, these data suggest a long-lasting thermal history for the SSB. Recent ion microprobe dating of zircon in migmatites from the SSB indicates that the duration of melting in this area was from about 32–22 Ma (Rubatto et al., 2009).

The best evidence of Alpine in situ melting is preserved in granitic gneisses (Burri et al., 2005). The gneisses were derived from ~280 to 300 Ma calc-alkaline granitoids (Romer et al., 1996 and Schaltegger and Gebauer, 1999), many of which were originally allanite-bearing (Berger et al., 2008). They have been interpreted as only being metamorphosed during Alpine events. The migmatites studied for geochronology contain abundant allanite and zircon and occur in an east-west profile through the SSB (Fig. 1). The principal deformation structures in the migmatites are Alpine in origin (Berger et al., 2005 and Berger et al., 2008).

3. Sample description

Samples collected in the migmatite zone of the Southern Steep Belt (Burri et al., 2005 and Berger et al., 2008) represent leucosome (suffix “L”) referring to thick leucocratic segregations resulting from melt crystallization, and orthogneisses (suffix “M” for mesosome) referring to country rock containing relatively small amounts of leucosome. The country rocks have granitic to tonalitic protoliths and are thus meta-granites or meta-tonalites. Some of the samples studied here (VAM1-VAL1, VAM2-VAL2 and BEM1-BEL1) are the same as investigated in Rubatto et al. (2009) for zircon geochronology. Sample details are summarised in Table 1.



Table 1. Summary of samples containing allanite used for U–Th–Pb analysis.










Sample

Rock type

Location

Temp in °C

Assemblage
--------------------------------------------------------------------------------





VAM1

Metatonalite

Arbedo (725 100, 119 350)


pl-qtz-hbl-bt ± Kfs

zrn-ttn-aln-ap-mag



VAL1

Discordant leucosome

Arbedo (725 100, 119 350)

630–670

pl-qtz-hbl Kfs

zrn-aln-ttn



VAM2

Metatonalite

Arbedo (725 100, 119 350)

<700

pl-qtz-hbl-bt ± Kfs

zrn-ttn-aln-ap



VAL2

Discordant leucosome

Arbedo (725 100, 119 350)

630–680

pl-qtz-hbl Kfs

zrn-aln-ttn








BEM1

Metagranodiorite

Bellinzona (722 500, 118 700)

640–690

pl-qtz-Kfs-bt-hbl

zrn-aln-ap



BEL1

Leucosome

Bellinzona (722 500, 118 700)

610–670

pl-qtz-Kfs-bt-hbl

zrn-aln-ap








GOL03

Metagranite

Golino (701 621, 113 762)


Kfs-qtz-pl-bt-hbl

zrn-aln-ap



GOL06

Metadiorite

Golino (701 621, 113 762)


hbl-pl-qtz-Kfs-bt

zrn-aln-ap-ttn








BER1

Metagranite

Berzona (691 206, 117 274)


Kfs-qtz-pl-bt

zrn-aln-ap ± ttn


Temperature estimates taken from Ti-in-zircon thermometry (Rubatto et al., 2009).Mineral abbreviations are according to Bucher and Frey (1994).Swiss grid coordinates.

Table options





Two structurally controlled sample pairs of meta-tonalite and leucosome (VAM1-VAL1 and VAM2-VAL2) were collected in Val d’Arbedo (Fig. 1) from a road cut that exposes a cross section of the sequence. Samples VAM1 and VAM2 are derived from meta-tonalite and contain diffuse leucosome with a weak subvertical fabric (Fig. 2a). They have a shared mineral assemblage of plagioclase, quartz, green amphibole and biotite, with abundant accessory zircon, apatite, titanite and allanite. The amphiboles form porphyroblasts that are commonly poikioblastic and include plagioclase; K-feldspar is rare or absent. Leucosomes VAL1 and VAL2 are discordant to the main fabric in the meta-tonalites (Fig. 2a). Although structurally younger, they preserve a weak foliation defined by subordinate biotite and VAL1 is openly folded (Fig. 2b). The leucosomes are coarse-grained and contain large green amphibole (Fig. 2b), abundant zircon, but only minor titanite and allanite.



Fig. 2. Field occurrence of migmatites in the SSB (a) Val d’Arbedo migmatite with deformed leucosome VAL1 and country rock with thin, layer-parallel leucosomes (VAM1). (b) Detail of leucosome VAL1 showing large poikilitic amphibole grains and a weak foliation. (c) Typical outcrop appearance of migmatites at the locality of Bellinzona. Note the different generations of cross-cutting leucosomes. (d) Deformed migmatite at the locality of Golino. (e) Berzona migmatite with layer-parallel leucosome and biotite selvages.

Figure options




A migmatite (BEM1 and BEL1) was sampled from road outcrop near Bellinzona, between Carasso and Gorduno (Fig. 1; see also Berger et al., 2008 and Berger et al., 2009). The migmatites are intensely folded with a penetrative subvertical foliation and contain distinct leucosomes of different structural ages (i.e. stromatic to discordant) (Fig. 2c). BEM1 comprises alternating mm- to cm-size bands of biotite and leucocratic material with abundant allanite, apatite and zircon. Allanite is large (up to 1 mm), zoned and commonly associated with biotite. Leucosome BEL1 formed axial planar to isoclinal folding and is plagioclase-rich with allanite, apatite and zircon, rare green amphibole and minor biotite indicating a weak foliation. The leucosome contains accessory allanite, apatite and multiple generations of metamorphic zircon (Rubatto et al., 2009).

The Melezza river between Golino and Intragna exposes large outcrops of banded, biotite-rich migmatite. These migmatites are intensely deformed. Sample GOL03 is granitic in composition with layer-parallel leucosomes and a structurally late, plagioclase-rich leucosome (Fig. 2d). The migmatites contain quartz, plagioclase, K-feldspar, biotite, and rare green amphibole. Zircon and apatite are common in leucocratic zones, whereas allanite is often in contact with biotite. Allanite grains are 200–700 µm in length and show core to rim zonation. Small (50–200 µm) idiomorphic grains are weakly zoned in comparison. Sample GOL06 comes from a metre-sized boudin of mafic gneiss within the migmatite (most likely meta-diorite). It contains diffuse leucosome and a foliation marked by green amphibole. Allanite is less common in the meta-diorite but is texturally similar to allanite in GOL03.

Sample BER1 was collected from a biotite-rich migmatite outcropping in the Bordione River near Berzona (Fig. 1). The migmatite contains minimum melt composition leucosomes concordant to a penetrative fabric marked by biotite (Fig. 2e). Rare discordant, plagioclase-rich veins were observed to truncate the main fabric. BER1 derives from a granitic protolith and contains thin bands of leucosome rimmed by biotite selvages (Fig. 2e). The accessory mineral assemblage includes apatite, zircon, allanite and minor titanite. Allanite is commonly found with biotite.

4. Analytical methods

Allanite grains in mounts and thin section were imaged by scanning electron microscope in backscattered electron mode (SEM-BSE), and analysed for major and minor elements using a wavelength dispersive electron microprobe (EMP), prior to isotopic analysis. SEM-BSE imaging revealed the internal structure of allanite grains defined primarily by rare earth element (REE) zoning, and EMP analysis provided a quantitative chemical characterisation of the allanite samples, used to assess the effects of compositional variation on ion microprobe analyses, and to calibrate the LA-ICPMS geochemical data. Major and trace element analyses on selected bulk rock samples were done by wavelength dispersive X-ray fluorescence (XRF). Details of the instrument parameters and analytical procedures used for SEM-BSE imaging, EMP and XRF analysis are provided in the Electronic annex supplement.

4.1. Sensitive High Resolution Ion Microprobe

U–Th–Pb dating of allanite was conducted over five separate sessions using the SHRIMP II and SHRIMP RG (Reverse Geometry) ion microprobes at the Research School of Earth Sciences, ANU. Analyses were performed on allanite in polished grain mounts with a 2.0–3.5 nA, 10 kV primary beam focused through a ~100 µm aperture to form a ~20–25 µm diameter spot. Operating procedure was broadly similar to that used for zircon (Williams, 1998), except that the post-collector retardation lens on SHRIMP II was fully activated during analysis to suppress low-energy scattered ions. At a mass resolution of >5100 (M/¿M at 1% peak height) the Pb, Th and U isotopes were resolved from all major interferences. Data acquisition for allanite followed that of Gregory et al. (2007) and each analysis consisted of six scans through the mass stations. Allanite standards were cast in the same grain mount as the samples and analysed after every three unknowns.

U–Th–Pb ratios were corrected by using 2-dimensional power law Pb/U–UO/U and Pb/Th–ThO/Th relationships (Gregory et al., 2007) and normalised to reference allanite (CAP, ID-TIMS mean Th–Pb age = 275.5 ± 1.5 Ma, n = 4; mean U–Pb age = 278.4 ± 5.8 Ma, n = 3; Barth et al., 1994). The U–Pb system of the CAP allanite is affected by uncertainties associated with initial radioactive disequilibrium due to 230Th-derived excess 206Pb (estimated at <4 Ma for U–Pb ages, Barth et al., 1994), and by the partial loss of U-derived Pb isotopes in some high Th/U crystal domains (Barth et al., 1994). Data sets for the CAP allanite were vetted for low 206Pb/238U outliers (±2s), which typically constitute ~10% of standard measurements. Minimum 1s external precisions of 3.0% for the U–Pb system and 2.0% for the Th–Pb system were assigned to the allanite standard. No corrections were made for potential matrix effects in 206Pb/238U and 208Pb/232Th measured from allanite.

To monitor the accuracy of the standard calibration, four secondary allanite standards, including Siss (ID-TIMS Th–Pb age = 31.5 ± 0.4 Ma, von Blanckenburg et al., 1992), Bona (ID-TIMS Th–Pb age = 30.1 ± 0.3 Ma, von Blanckenburg et al., 1992), BC (’98–19’ ID-TIMS zircon U–Pb age = 90.8 ± 1.0 Ma, (Butler et al., 2002) and Tara (SHRIMP Th–Pb age = 415 ± 3 Ma, Gregory et al., 2007) were analysed with the sample unknowns in two of the sessions. The results are presented in the Electronic annex supplement. The 208Pb/232Th ages obtained by SHRIMP are identical within error of their reference ages. The elevated SHRIMP 206Pb/238U ages for Siss, Bona and 98–19 allanites, however, reflect radioactive disequilibrium-related bias in form of excess 206Pb (von Blanckenburg et al., 1992 and Oberli et al., 2004).

The dependence of ion microprobe 206Pb+/U+ and 208Pb+/Th+ analyses on sample chemistry can compromise the use of this technique for dating chemically complex minerals if suitable matrix-matched standards are unavailable. Whereas the investigated allanites display a range of REE and Th contents (LREE + Th = 0.2–0.8 cations per formula unit) (see Sections 5 and 6), the magmatic standards have typically high LREE and Th contents of 0.5–0.9 cpfu (Gregory et al., 2007). Such compositional differences can offset standard and sample / and / (n = 0–2) used for Th–Pb and U–Pb calibrations, respectively (e.g. allanite, Catlos et al., 2000; monazite, Fletcher et al., 2010). We monitored matrix effects by comparing standard and sample ln[ThO+/Th+] and ln[UO+/U+] measurements from the same session (listed with the U–Th–Pb isotope data in the Electronic annex supplement). Matrix effects on Th–Pb isotope ratios from metamorphic allanite were identified (i.e. different ln[ThO+/Th+] values) and are discussed in Section 9.2.

All raw data were processed using SQUID-2 (rev. 2.50) software (Ludwig, 2009), which reproduces the procedure described by Gregory et al. (2007) for allanite. Listed uncertainties for radiogenic isotope ratios and ages are ±1s and include all components of statistical precision, including uncertainty on the estimation of non-radiogenic Pb and uncertainty on the Pb/U and Pb/Th calibrations, which range from 0.33% to 1.2% and 0.23% to 0.52% (1s), respectively. The 1s external reproducibility of Th–Pb and U–Pb ages measured from the allanite standard were quadratically propagated into the final uncertainties on all single spot ages and ratios, since there is no a priori reason to believe all sample allanites are a homogeneous population. The final age uncertainties do not include potential unquantified uncertainties in 206Pb/238U and 208Pb/232Th arising from matrix effects in allanite analyses. To compare ages from both the U–Pb and Th–Pb systems in allanite, the final 206Pb/238U and 208Pb/232Th ratios of each analysis are plotted on 2-dimensional U–Th–Pb concordia diagrams (Section 6). Age calculations and plots were made using Isoplot/Ex (Ludwig, 2003). Pooled ages are cited at 95% confidence limits (cl.) unless otherwise stated.

4.2. Laser ablation ICP-MS

Large (<500 µm) allanite grains observed in polished thin section from BER1 meta-granite were dated in situ (in thin section) using laser ablation ICPMS (LA-ICPMS) techniques. The LA-ICPMS method was used initially as a reconnaissance tool, to provide a rapid assessment of the age of allanite overgrowths in a migmatite sample that contained no new zircon growth. U–Th–Pb analyses were performed using a 193 nm ArF Excimer laser system couple to an Agilent 7500S quadruple ICP-MS housed at RSES, ANU. Instrument parameters were generally as described by Eggins et al. (1998) and data acquisition and reduction followed that of Gregory et al. (2007). The laser was focused to a spot size of 32 µm using a laser pulse rate of 5 Hz and laser irradiance of approximately 10–12 J/cm2. The analyte was ablated into a mixed He-Ar (1:3) carrier gas (gas flow ~1.2 L/min). Each isotope analysis was of 65 s duration in time-resolved (peak hopping) analysis mode, including 40 s of ablation and 25 s monitoring gas blank. The depth of laser drilling was ~20–25 µm per analysis. A post-plasma oxide was used to monitor the production of molecular compounds (ThO/Th <0.5%) and lead hydride interferences were checked on a pure Pb sample. Torch position (~5.6 mm) and lens tuning were adjusted to maximise sensitivity for the Pb isotopes, Th and U; 204Pb was not measured due to a systemic 204Hg interference.

Raw data was processed offline using an in-house macro-based EXCEL spreadsheet. Each analysis was corrected for background gas blank and for laser-induced element fractionation processes, which occur during stationary laser sampling (e.g. Eggins et al., 1998). External calibration of 208Pb/232Th was done against an allanite standard (AVC, ID-TIMS Th–Pb age = 276.3 ± 2.2 Ma, Barth et al., 1994). Element (Th, U, P, Si, Ca and REE) concentrations were referenced directly to a NIST SRM 610 glass. Both materials were re-analysed once after every 8 sample analyses. Internal precision on the matrix normalisation factor (F = 208Pb/232Threference ÷ 208Pb/232Thmeasured) determined from the allanite standard is 0.6% (1s) and the external reproducibility based on repeat standard analyses is 1.25% (1s) for AVC (n = 10). The CAP allanite was analysed as a secondary standard and used to calibrate AVC for inter-comparison of reference materials. Eleven analyses of CAP yielded a Th–Pb isochron age of 289 ± 12 Ma (MSWD = 0.36, n = 11/11) and 10 analyses of AVC yielded an age of 281 ± 23 Ma (MSWD = 0.13, n = 10/10) (Electronic annex Table 5).

Th–Pb ages were calculated from LA-ICPMS analyses using a 2-dimensional isochron plot, a method suited to analyses of Th-rich minerals for which accurate determination of non-radiogenic Pb is particularly crucial. The Th–Pb isochrons are constructed using the non-radiogenic portion of 206Pb, instead of 204Pb, as the reference isotope. The measured 232Th/206Pb and 208Pb/206Pb are adjusted accordingly: 232Th/206Pbinitial = 232Th/206Pbmeasured ÷ f-206 (Gregory et al., 2007). The uncertainties in the measured ratios, the calculation of non-radiogenic Pb and the external precision of the allanite standard are added in quadrature into the final uncertainty on all isotope ratios. Calculation of isochron ages (quoted at 95% cl.) was done using Isoplot/Ex (Ludwig, 2003).

Rare earth element (REE) and trace element concentrations of silicate and accessory minerals were determined by LA-ICPMS, using similar working conditions to U–Th–Pb isotope analysis, and laser spot sizes of 24–84 µm. The analytical protocol for REE analysis is outlined in the Electronic annex supplement.

4.3. Estimation of initial Pb in allanite from 207Pb abundance

Allanite can have high contents of initial Pb (Pb included at mineral formation) that make calculated 206Pb/238U and 208Pb/232Th ages sensitive to uncertainty in the choice of non-radiogenic Pb correction (all Pb components, except in situ accumulated radiogenic Pb, are thus dominated by initial Pb and include a minor or irrelevant amount of Pb from surface contamination). In this study, the non-radiogenic Pb content of an analysis is given as a fraction or percentage of total measured 206Pb (f-206) or 208Pb (f-208), an expression routinely used by the ion probe community for the purpose of data interrogation.

Isotopic analyses by both SHRIMP and LA-ICPMS are burdened with isobaric interferences on the mass 204 that could be neither avoided with the current techniques, nor accurately subtracted (Gregory et al., 2007). In addition, the uncertainty associated with measuring the very small 204 peak can be significant for geologically “young” minerals. These problems can be circumvented by the estimation of initial Pb from the abundance of 207Pb rather than 204Pb (Gregory et al., 2007). The f-206 (and f-208) is thus calculated from the measured 207Pb/206Pb, assuming U–Pb concordance (Williams, 1998). The equations used to calculate f-206 and f-208 are stated in the Electronic supplement.

The 207Pb-based approach is model-dependent and is therefore potentially less accurate than the “classical” approach based on measurements of all Pb isotopes in minerals with low U/204Pb and Th/204Pb. For example, the f-208 and f-206 estimates depend on assumed concordance and may be inaccurate if the systems are discordant. The 207Pb correction method is generally suited to “young” Phanerozoic samples for which the range of potential radiogenic 207Pb/206Pb is small and it is often valid to assume near concordance (Williams, 1998). For the purpose of this study, we have compared the f-206 and f-208 values calculated using 204Pb-corrected data in some samples with the 207Pb-corrected data (see Electronic supplement), in order to evaluate the affect of potential isotopic discordance and/or unquantified spectral interferences on the 204 peak during ion probe analysis (cf. Stern and Berman, 2000). The estimates of initial Pb based on 204Pb are generally comparable to the model-dependent 207Pb corrected data for most of the analysed allanites.

The initial 207Pb/206Pb used was based on a model Pb isotopic composition (Stacey and Kramers, 1975) at the approximate age of the sample. When the initial 207Pb/206Pb approximates the modelled Pb isotope composition, a single spot correction method is applicable, however this cannot be assumed for Pb-rich minerals such as allanite. An independent check on the initial 207Pb/206Pb was done by regressing total 207Pb/206Pb versus 238U/206Pb data using an inverse concordia (Tera–Wasserburg) plot. A comparison of the initial 207Pb/206Pb in allanite determined by this method with the modelled composition is discussed in Sections 6 and 9. The relationship between f-206 and f-208 and the 207Pb-corrected 206Pb/238U and 208Pb/232Th ages in samples with relatively uniform Th/U contents is also used as a tool to assess the choice of initial Pb composition.

5. Allanite major element composition

The chemistry of allanite was determined prior to ion microprobe dating to assess the extent of compositional zoning in allanite grains used for U–Th–Pb isotope analysis. The major element compositions of allanite grains are given in Table 2. Element (CaO, Fetotal and Th2O) maps of representative allanite dated in this study from meta-tonalite VAM1 clearly illustrate a strong internal zonation (Fig. 3a). From core to overgrowth, Th2O content decreases dramatically and is almost absent in epidote, whereas CaO content increases overall and Fetotal decreases only slightly. The compositional variation between allanite cores and overgrowths define a solid solution between the allanite and epidote end-members (Fig. 3b) where REE are primarily incorporated into allanite via: Ca2+ + Fe3+ ¿ REE3+ + Fe2+ (Petrík et al., 1995 and Gieré and Sorenson, 2004). The cores and overgrowths from meta-granite samples fall within or slightly below the end-member chemical classification for allanite (REE + Th ¿ 0.5 cations per formula unit). In contrast, overgrowths from meta-tonalite samples are classified as REE-epidote (REE + Th < 0.2 cpfu).



Table 2. Average major element composition of allanite determined by EMP analysis.



















VAM1
--------------------------------------------------------------------------------


BEM1
--------------------------------------------------------------------------------


GOL03
--------------------------------------------------------------------------------


GOL06
--------------------------------------------------------------------------------


BER1
--------------------------------------------------------------------------------





c

ep

ov

ep

ov

c

ep

ov

c

ov

c

r

ov




n

2

5

11

4

4

4

2

11

6

7

1

4

16



SiO2

32.3

36.8

35.7

35.8

34.9

31.1

32.5

31.5

31.8

33.7

32.8

33.3

33.1



TiO2

0.31

0.19

0.19

0.14

0.17

1.06

0.17

0.21

1.26

0.36

0.17

0.14

0.13



Al2O3

18.2

24.7

23.4

24.1

22.8

14.1

19.1

18.2

16.3

21.6

20.4

21.9

21.5



MgO

0.91

0.18

0.30

0.26

0.35

0.94

0.41

0.50

1.08

0.58

0.85

0.46

0.38



CaO

14.2

23.4

21.8

22.1

20.9

11.2

15.9

14.5

12.4

17.5

12.3

15.1

15.4



MnO

0.39

0.23

0.29

0.28

0.27

0.86

1.18

1.20

0.33

0.25

0.39

0.52

0.78



FeO

12.2

9.76

10.1

9.72

10.4

14.8

13.0

12.9

11.9

8.95

10.5

10.8

11.7



La2O3

5.41

0.45

1.10

0.63

1.22

6.40

2.51

3.19

7.01

3.69

4.22

3.10

2.68



Ce2O3

8.92

0.89

2.22

1.29

2.61

11.1

4.87

5.93

10.7

6.52

9.24

6.51

5.71



Pr2O3

0.93

0.14

0.29

0.19

0.33

1.14

0.63

0.73

0.95

0.67

0.86

0.61

0.60



Nd2O3

2.73

0.53

1.13

0.80

1.34

3.43

2.05

2.52

2.59

2.03

3.07

2.25

2.21



Sm2O3

0.22

0.12

0.16

0.20

0.28

0.28

0.37

0.39

0.18

0.18

0.30

0.29

0.33



Gd2O3

0.09

0.06

0.11

0.33

0.30

0.16

0.39

0.43

0.09

0.07

0.18

0.16

0.31



Y2O3

0.05

0.15

0.10

0.79

0.51

0.13

1.32

1.17

0.07

0.05

0.07

0.27

0.48



SrO

0.02

0.20

0.08

0.09

0.08

0.02

0.01

0.02

0.01

0.04

0.21

0.15

0.04



ThO2

1.35

0.08

0.20

0.27

0.54

1.31

1.77

2.21

1.75

0.85

1.25

1.28

1.17








Total

98.2

97.9

97.2

97.0

97.0

98.0

96.2

95.6

98.4

97.0

96.8

96.8

96.5



LREE + Th

0.60

0.06

0.14

0.13

0.20

0.79

0.43

0.51

0.74

0.41

0.65

0.45

0.42


n = number of EMP analyses used for average; c = core; ep = epidote; ov = overgrowth; LREE + Th given in cations per formula unit on the basis of 12.5 oxygens.

Table options







Fig. 3. (a) Element (CaO, Fetotal and Th2O) maps determined by electron microprobe and BSE image of allanite grain from sample VAM1 metatonalite. (b) Summary of allanite EMP data from allanite dated in this study.

Figure options




6. Allanite ages and trace element composition

Allanite rare earth element data are provided in Table 3. Allanite U–Th–Pb data determined by SHRIMP and LA-ICPMS are compiled in tables in the Electronic annex and presented in Fig. 4, Fig. 5, Fig. 6, Fig. 7 and Fig. 8.



Table 3. Average trace element composition of allanite samples determined by LA-ICP-MS analysis.


















VAM1
--------------------------------------------------------------------------------


VAL1

VAM2
--------------------------------------------------------------------------------


VAL2

BEM1
--------------------------------------------------------------------------------





c

ep

ov

ov

c

ov

wz

ov

c

alc

ep

ov




Trace elements in ppm




n = 5

n = 8

n = 14

n = 4

n = 5

n = 2

n = 3

n = 3

n = 3

n = 1

n = 1

n = 2



P

118

238

212

163

139

129

129

130

174

61

68

53



Sc

51

147

146

185

34

123

23

130

91

133

538

301



V

514

398

378

353

293

387

218

384

711

556

373

366



Cr

69

30

38

38

160

137

122

335

100

217

131

109



Rb

0.15

0.26

0.43

0.15

–

0.21

–

0.23

0.19

0.20

0.52

0.37



Sr

603

1080

548

532

986

410

1290

389

118

155

293

301



Y

355

743

1000

1190

355

1111

197

1560

1070

2160

5560

3630



Zr

2

19

16

16

3

16

3

15

4

5

15

12



Nb

0.18

0.09

0.03

0.02

0.04

–

0.03

0.02

0.11

0.07

–

0.03



Ba

0.96

2.90

2.27

1.72

1.38

2.37

1.4

2.40

1.43

0.63

1.63

1.1



La

43,000

2980

6820

7500

32,900

12,500

24,900

14,000

59,200

33,000

5400

8340



Ce

73,200

6670

14,900

15,600

55,400

23,900

41,700

28,600

106,000

68,400

12,400

18,400



Pr

6540

792

1770

1770

4970

2470

3650

3230

9850

8000

1520

2170



Nd

18,800

3270

7220

6910

14,600

8550

10,500

12,200

29,270

30,000

6670

9090



Sm

1480

648

1310

1270

1360

1290

869

1960

2740

4380

1940

2280



Eu

106

145

269

248

96

289

66

367

174

333

397

385



Gd

546

470

851

856

557

824

326

1210

1120

2030

1790

1800



Tb

40

50

79

86

43

84

24

120

91

181

302

260



Dy

134

219

311

363

147

353

80

499

350

712

1570

1190



Ho

16

30

40

48

16

46

9

64

45

89

231

158



Er

28

55

72

87

26

84

14

120

89

165

394

256



Tm

3

5

7

8

2

8

1

12

10

17

31

20



Yb

13

22

32

39

10

40

5

60

55

87

120

89



Lu

1.8

3

4

5

1.3

5

1

7

7

11

10

8



Hf

0.11

0.77

0.70

0.77

0.15

0.79

0.15

0.82

0.29

0.25

0.91

0.62



Ta

0.02

–

–

–

0.02

–

–

–

0.02

0.01

–

–



Pb

258

36

28

29

122

33

39

32

405

529

31

32



Th

12,200

443

1090

1900

8900

3980

6370

3360

13,900

18,600

2330

3780



U

67

65

145

201

63

257

38

516

76

438

689

548



LaN/LuN

2580

122

175

154

2630

252

4090

207

859

327

58

105



Eu/Eu*

0.36

0.80

0.78

0.73

0.34

0.86

0.38

0.73

0.30

0.34

0.65

0.58



Th/U

222

8

8

9

178

15

168

7

187

42

3

8









BEL1

GOL03

GOL06

BER1




ep

ov

c

ep

ov

c

ep

ov

c

r

ov



Trace elements in ppm













n = 3

n = 4

n = 3

n = 1

n = 4

n = 3

n = 2

n = 3

n = 1

n = 1

n = 4



P

144

111

93

69

76

122

76

72

47

53

57



Sc

269

284

116

257

241

37

76

74

38

17

125



V

433

443

463

207

191

676

474

455

164

83

138



Cr

356

324

18

39

28

154

180

160

9

33

20



Rb

0.59

5

3

1.9

1.8

0.13

0.21

0.94

0.13

0.15

1.3



Sr

292

260

57

133

120

118

245

183

3550

3120

984



Y

4301

4920

1350

12,610

11,300

550

608

515

545

1040

2830



Zr

15

13

5

3

3

4

10

9

4

4

4



Nb

0.03

0.31

0.44

0.12

0.15

1.96

0.05

0.03

0.09

0.03

0.47



Ba

4

17

0.48

0.18

0.22

0.70

1.4

4

21

26

7



La

11,400

14,000

60,700

24,300

26,700

60,900

25,600

29,400

43,600

31,900

24,900



Ce

24,400

29,500

110,000

49,800

55,100

89,700

53,900

55,100

81,800

55,500

48,000



Pr

2790

3380

10,300

5430

5990

6900

5440

5100

7990

4960

4900



Nd

10,900

13,200

31,400

19,500

21,400

18,100

17,500

15,500

25,700

14,800

16,700



Sm

2180

2650

3250

3810

4130

1500

1700

1370

2640

1440

2630



Eu

430

455

87

396

375

120

226

186

80

196

304



Gd

1790

2180

1420

3110

3280

651

764

617

928

658

1550



Tb

222

267

118

454

458

52

56

46

66

70

186



Dy

1130

1320

446

2630

2540

188

196

162

198

288

799



Ho

171

196

57

462

420

24

25

21

23

39

114



Er

351

403

120

1180

1030

45

47

40

43

81

232



Tm

36

41

14

152

131

4

4

4

5

9

26



Yb

175

209

87

920

805

24

23

20

33

58

147



Lu

20

25

12

112

99

3

3

3

4

7

17



Hf

0.80

0.75

0.25

0.32

0.24

0.22

0.48

0.43

0.10

0.24

0.18



Ta

0.02

0.34

0.04

0.05

0.05

0.11

–

0.02

0.02

–

0.03



Pb

37

36

364

74

75

313

44

44

367

358

116



Th

3390

4380

12,400

18,400

20,900

12,000

5480

6620

12,100

8270

9130



U

687

1140

69

1360

1010

193

410

326

422

1540

964



LaN/LuN

59

60

533

23

28

2062

863

1120

1090

518

157



Eu/Eu*

0.67

0.58

0.12

0.35

0.31

0.37

0.61

0.62

0.16

0.62

0.46



Th/U

5

4

180

14

22

62

13

20

29

5

11


c = core; ep = epidote; ov = overgrowth; wz = cloudy to weak oscillatory zoning; alc = altered core; LaN = chondrite normalised La; Eu/Eu* = europium anomaly.
Full-size table
Table options







Fig. 4. Allanite age and trace element data for Val d’Arbedo. (a) SEM-BSE image of composite allanite grain from VAM1 meta-tonalite. Dashed and solid circles indicate LA-ICPMS and SHRIMP analysis locations, respectively. Ep = epidote, ov = overgrowth, incl = inclusion (apatite) and alt core = secondary alteration of core. White bar is 100 µm. (b) Chondrite-normalised REE patterns of allanite cores from VAM1 and REE-epidote overgrowths from VAM1 and VAL1 leucosome. (c) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for allanite cores in VAM2 meta-tonalite and VAM1 meta-tonalite. Quoted ages are weighted mean ages. (d) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for REE-epidote overgrowths in VAM1 meta-tonalite and VAL1 leucosome. Quoted ages are weighted mean ages. (e) Inverse concordia of uncorrected allanite U–Pb analyses from VAM1. Error crosses are 1 sigma for all plots.

Figure options






Fig. 5. Allanite age and trace element data for Val d’Arbedo. (a) SEM-BSE images of REE-epidote overgrowths in VAL2 leucosome and VAM2 meta-tonalite and cloudy/weak oscillatory zoned allanite in VAM2. Dashed and solid circles indicate LA-ICPMS and SHRIMP analysis locations, respectively. Ep = epidote and ov = overgrowth. White bar is 100 µm. (b) Chondrite-normalised REE patterns of allanite cores from VAM2, and REE-epidote overgrowths and weakly zoned allanite from VAM2 and VAL2. (c) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for REE-epidote overgrowths in VAL2 and weakly zoned allanite in VAM2. Quoted ages are weighted mean ages. (d) Inverse concordia of uncorrected allanite U–Pb analyses from VAM2. Error crosses and bars are 1 sigma for all plots.

Figure options






Fig. 6. Allanite age and trace element data for Bellinzona. (a) SEM-BSE image of epidote grain with allanite overgrowth in BEL1 leucosome. Dashed and solid circles indicate LA-ICPMS and SHRIMP analysis locations, respectively. Ep = epidote. White bar is 50 µm. (b) SEM-BSE image of composite allanite grain from BEM1 meta-granodiorite. Alt core = alteration of core (patchy, irregular zoning). White bar is 100 µm. (c) Chondrite-normalised REE patterns of allanite cores from BEM1 and REE-epidote and allanite overgrowths from BEM1 and BEL1. (d) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for allanite overgrowths in BEL1 leucosome. Quoted ages are weighted mean ages. (e) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for allanite cores in BEM1 meta-tonalite. Quoted ages are weighted mean ages. (f) Age versus non-radiogenic Pb portion of total measured Pb for both U–Pb and Th–Pb systems in allanite from BEL1. Error bars and crosses are 1 sigma for all plots.

Figure options






Fig. 7. Allanite age and trace element data for Golino. (a) SEM-BSE image of composite allanite grain from GOL03 meta-granite. Dashed and solid circles indicate LA-ICPMS and SHRIMP analysis locations, respectively. Ov = overgrowth. White bar is 100 µm. (b) Chondrite-normalised REE patterns of allanite cores and overgrowths from GOL03 and allanite overgrowths from GOL06 meta-diorite. (c) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for allanite cores in GOL03 and GOL06. Quoted ages are weighted mean ages. (d) 206Pb/238U versus 208Pb/232Th diagram of 207Pb-corrected single-spot ages determined by SHRIMP for allanite overgrowths in GOL03 and GOL06. Quoted ages are weighted mean ages. (e) Age versus non-radiogenic Pb portion of total measured Pb for both U–Pb and Th–Pb systems in allanite overgrowths in GOL03. (f) Age versus non-radiogenic Pb portion of total measured Pb for both U–Pb and Th–Pb systems in allanite overgrowths in GOL06. Error bars and crosses are 1 sigma for all plots.

Figure options






Fig. 8. Allanite age and trace element data for Berzona. (a) Photomicrograph of composite allanite grain in BER1 meta-granite (PPL); bt = biotite, plag = plagioclase. Circles in allanite are LA-ICPMS analysis pits (b) SEM-BSE image of grain in (a) showing bright-BSE, partially recrystallised core and darker-BSE overgrowth. White bar is 200 µm. (c) Chondrite-normalised REE patterns of allanite cores and overgrowths from BER1. (d) Th–Pb isochron of uncorrected allanite Th–Pb data determined by LA-ICPMS for BER1; Pbc denotes the non-radiogenic Pb component. Error bars and crosses are 1s for all plots.

Figure options




6.1. Val d’Arbedo

Zoned allanites <400 µm in length were identified in the VAM1 and VAM2 meta-tonalites by optical analysis and backscattered electron (BSE) imaging. They consist of a high-BSE intensity core, a low-BSE intensity rim and an unzoned, moderate-BSE intensity overgrowth (Fig. 4a). The different zones correspond to allanite (av. LREE + Th = 0.6 cpfu), epidote (<0.1 cpfu) and REE-epidote (~0.14 cpfu) (Table 2). In VAL1 leucosome, REE-epidote occurs as unzoned overgrowths on epidote grains; the large allanite cores are absent. Some allanite cores in VAM1 and all cores in VAM2 have patchy and irregular zoning that proceeds from core boundaries and along fractures (Fig. 4a). These zones are low-BSE intensity and therefore low-REE and are interpreted as secondary alteration. The allanite cores are characterised by highly fractionated REE patterns (av. LaN/LuN ~2800), moderate negative Eu anomalies (Eu/Eu* ~0.4) and high Th/U (~100–200) (Fig. 4b, Table 3). In contrast, the REE-epidote overgrowths in VAM1 and VAL1 have low La/Lu (LaN/LuN = 150–170), small Eu anomalies (Eu/Eu* ~0.7–0.8) and low Th/U (~7–9).

Eleven U–Th–Pb analyses of allanite cores from VAM1 meta-tonalite were obtained by SHRIMP. They have high radiogenic 208Pb contents (>96% of the total 208Pb) compared to 206Pb (68–81% of the total 206Pb) (Electronic annex Table 1). From 11 analyses a weighted mean 208Pb/232Th age of 286 ± 5 Ma (MSWD = 0.7) and a 206Pb/238U age of 276 ± 7 Ma (MSWD = 0.9) can be calculated (Fig. 4c). The U–Pb age is lower than but within uncertainty of the Th–Pb age. Four SHRIMP analyses were made on altered allanite cores from VAM2 meta-tonalite. The analyses do not define a statistical population and scatter along U–Th–Pb concordia from ~282 to ~163 Ma (Fig. 4c). Nonetheless, they confirm that allanite cores in this sample are pre-Alpine in age. Two analyses (VAM2-10.2, 13.1) of alteration zones have relatively low Th/U (~37) and high non-radiogenic Pb contents (Electronic annex Table 1).

Thirteen SHRIMP analyses were obtained on REE-epidote overgrowths from VAM1. The calculated proportion of radiogenic Pb for each analysis is extremely low, particularly for the U–206Pb system (~3–14% of the total 206Pb) (Electronic annex Table 1). The 13 analyses calculate a weighted mean 208Pb/232Th age of 23.1 ± 1.3 Ma (MSWD = 1.8). The corresponding 206Pb/238U age is lower than the 208Pb/232Th age and thus, the analyses are shifted to the left of U–Th–Pb concordia (Fig. 4d). The same analyses yield lower intercept U–Pb age of 24.1 ± 2.4 Ma and an initial 207Pb/206Pb composition of 0.86 ± 0.10 Ma (2s, MSWD = 1.3) by extrapolating the total 238U/206Pb–207Pb/206Pb data along a non-radiogenic Pb mixing line (Fig. 4e). Although less precise, the intercept age is preferred over the single-spot 206Pb/238U ages because it does not rely on the choice of non-radiogenic Pb correction, which is significant for this sample. The U–Pb intercept age is within uncertainty of the Th–Pb age and both indicate that the REE-epidote overgrowths in meta-tonalite are of Alpine-age and therefore younger than the allanite cores.

Eight U–Th–Pb analyses of REE-epidote overgrowths from VAL1 leucosome yield a weighted mean 208Pb/232Th age of 29.1 ± 2.0 Ma (MSWD = 2.4) and a significantly lower 206Pb/238U age of 22.6 ± 0.9 Ma (MSWD = 1.0) (Fig. 4d). The Th–Pb and U–Pb age systems have similarly low amounts of radiogenic Pb to overgrowths in VAM1 (~12–21% of the total 206Pb and 208Pb) (Electronic annex Table 1). The ages indicate, at least, that the REE-epidote overgrowths in leucosome are of Alpine origin.

REE-epidote overgrowths are present in VAM2 meta-tonalite and VAL2 leucosome (Fig. 5a). The overgrowths are similar in composition to those in VAM1 and VAL1, including low La/Lu (LaN/LuN = 200–250), small negative Eu anomalies (Eu/Eu* = 0.74–0.86) and low Th/U (av. ~15 and ~7) (Fig. 5b, Table 3). Allanite grains characterised by cloudy to weak-oscillatory zoning also occur in the meta-tonalite (Fig. 5a). The allanite REE patterns are HREE-depleted with respect to L-MREEs (av. LaN/LuN = 4090) and have moderate negative Eu anomalies (av. Eu/Eu* = 0.38) (Fig. 5b). The weakly zoned allanites have variable but high Th/U compositions (av. Th/U = 168) similar to allanite cores (Table 3).

Seven SHRIMP analyses of REE-epidote overgrowths from VAL2 leucosome yield a weighted mean 208Pb/232Th age of 28.1 ± 0.8 Ma (MSWD = 0.8) and a significantly lower 206Pb/238U age of 22.6 ± 1.2 Ma (MSWD = 2.0) (Fig. 5c). The ages are identical to those measured from overgrowths in VAL1 (Fig. 4d) and the Th–Pb and U–Pb analyses have similarly low amounts of radiogenic Pb (~27–34% of the total 206Pb and 208Pb) (Electronic annex Table 1). Two analyses (VAM2.10-1 and 7.1) of low Th/U, REE-epidote overgrowths from VAM2 give high 208Pb/232Th ages of ~30 and ~32 Ma, and low 206Pb/238U ages of ~22 and ~26 Ma (Fig. 5c).

Nine analyses were obtained on weak oscillatory-zoned allanite from VAM2 meta-tonalite. One analysis (VAM2.8-1) is almost free of radiogenic Pb and was not considered in age calculations (outlier in Fig. 5c). The proportion of of radiogenic Pb in each analysis is low and variable accounting for 1–16% of the total 206Pb and 16–67% of the total 208Pb (Electronic annex Table 1). The eight analyses calculate a weighted mean 208Pb/232Th age of 35 ± 2 Ma (MSWD = 2.7) and a 206Pb/238U age of 30 ± 3 Ma (MSWD = 0.6) (Fig. 5c). The same data give a lower intercept U–Pb age of 30 ± 4 Ma and a tightly constrained initial 207Pb/206Pb composition of 0.839 ± 0.004 (2s, MSWD = 0.8) (Fig. 5d), which is identical to the model value of Stacey and Kramers (1975) at the assumed sample age (~30 Ma).

6.2. Bellinzona

Allanite in BEL1 leucosome occurs as thin (<50 µm) unzoned overgrowths on idiomorphic epidote crystals (Fig. 6a). In BEM1 meta-granodiorite, allanite occurs as large (<1 mm) composite grains comprised of high-BSE intensity cores and low-BSE intensity epidote rims (Fig. 6b). Rare overgrowths of REE-epidote (av. REE + Th = 0.20 cpfu) are found on composite grains, similar to allanite overgrowths in BEL1. Secondary alteration of allanite cores is observed, indicated by regions of patchy, low-BSE intensity zoning at core boundaries and adjacent to fine fractures (Fig. 6b). The high-BSE intensity cores are LREE-enriched (av. LaN/LuN = 860), have moderate negative Eu anomalies (av. Eu/Eu* = 0.3) and high Th/U (av. ~185) (Table 3, Fig. 6c). The overgrowths in BEL1 and BEM1 are less LREE-enriched (av. LaN/LuN = 60 and 105, respectively) with small negative Eu anomalies (Eu/Eu* = 0.6) and low Th/U (av. <10).

Six U–Th–Pb analyses were obtained by SHRIMP on low Th/U allanite overgrowths in BEL1. The radiogenic 206Pb and 208Pb contents of each analysis are similar: 38–53% of the total 206Pb and 30–34% of the total 208Pb (Electronic annex Table 2). A total 238U/206Pb–207Pb/206Pb plot of the six analyses yields an imprecise initial 207Pb/206Pb composition of 0.78 ± 0.11 (95% cl.) and lower intercept age of 20 ± 5 Ma (MSWD = 3.9, 95% cl., Electronic annex Fig. 1). The initial Pb composition determined from allanite overgrowths is lower than but within uncertainty of the model Pb isotope composition (Stacey and Kramers, 1975). The lack of correlation between the corrected 206Pb/238U ages and the amount of non-radiogenic Pb in each analysis (Fig. 6d), suggests that the choice of non-radiogenic Pb correction used for the allanite overgrowths is robust. The six analyses lie along U–Th–Pb concordia (Fig. 6e) and yield a weighted mean 208Pb/232Th age of 22.3 ± 0.7 Ma (MSWD = 1.0) and an identical 206Pb/238U age of 22.2 ± 0.8 Ma (MSWD = 0.9).

Thirteen analyses were obtained on high Th/U allanite cores in sample BEM1 meta-granodiorite. They contain >97% radiogenic 208Pb of the total 208Pb compared to 53–73% radiogenic 206Pb (Electronic annex Table 2). Twelve of the 13 analyses calculate a weighted mean 208Pb/232Th age of 297 ± 7 Ma (MSWD = 0.5), which agrees well with the weighted mean 206Pb/238U age of 293 ± 4 Ma (MSWD = 0.8) (Fig. 6f). One analysis (BEM.3-2) was made on an alteration zone and was omitted from age calculation (Electronic annex Table 2). The alteration zone has low Th/U (~60) compared to unaltered cores and gives low 208Pb/232Th and 206Pb/238U ages (~268 and ~238 Ma, respectively), which are offset from U–Th–Pb concordia (Fig. 6f).

6.3. Golino

Allanites in GOL03 meta-granite and GOL06 meta-diorite occur as unzoned overgrowths on epidote and as large (<500 µm) grains comprising a high-BSE intensity core (av. REE + Th = 0.7–0.8 cpfu) and a low-BSE intensity overgrowth or rim (av. REE + Th > 0.4–0.5 cpfu) (Fig. 7a, Table 3). Secondary alteration of allanite cores is indicated by patchy, low-BSE intensity zones that occur mainly at the core-rim boundary (Fig. 7a). Allanite cores in GOL03 and GOL06 have large to moderate negative Eu anomalies (av. Eu/Eu* = 0.1 and 0.4, respectively) and highly fractionated REE patterns showing LREE enrichment (av. LaN/LuN ~500 and ~2000) (Table 3, Fig. 7b). In contrast, allanite overgrowths in GOL03 have low LREE with respect to M- (av. LaN/LuN ~28), and smaller negative Eu anomalies (av. Eu/Eu* = 0.3) (Fig. 7b). REE patterns of allanite overgrowths in GOL06 are HREE-depleted (Fig. 7b), which may reflect the presence of HREE-bearing amphibole in this sample and/or the bulk rock composition of the meta-diorite. The cores and overgrowths have distinct Th/U compositions in both the meta-granite (core Th/U = 145–185, overgrowth Th/U = 14–40), and in the meta-diorite (core Th/U = 53–76, overgrowth Th/U = 17–38) (Electronic annex Table 3).

Seventeen U–Th–Pb analyses of allanite cores in GOL03 meta-granite were determined by SHRIMP. The proportion of radiogenic 208Pb (>94% of total 208Pb) in each analysis is higher than radiogenic 206Pb (>53% of total 206Pb) (Electronic annex Table 3). Fourteen of 17 SHRIMP analyses from sample GOL03 yield a weighted mean 208Pb/232Th age of 281 ± 5 Ma (MSWD = 2.1) and a 206Pb/238U age of 286 ± 11 Ma (MSWD = 1.6). One analysis (GOL03.11-2) overlapped core and overgrowth zones (outlier in Fig. 7c), and two analyses (GOL03.21-1 and 24.2) were measured on alteration zones and thus omitted from age calculation. The alteration zones yield U–Pb and Th–Pb ages lower than the weighted averages (Electronic annex Table 3). Eight analyses were obtained on allanite cores in GOL06 that displayed varying degrees of secondary alteration. Grouped together, they yield a weighted mean 208Pb/232Th age of 268 ± 10 Ma (MSWD 1.6), identical to the 206Pb/238U age of 268 ± 7 Ma (MSWD = 0.6) (Fig. 7c).

Twelve analyses of allanite overgrowths in GOL03 and GOL06 were obtained by SHRIMP. The calculated proportion of radiogenic Pb in each analysis is 69–79% (of total 208Pb) and 31–53% (of total 206Pb) for GOL03 and 41–56% (208Pb) and 17–27% (206Pb) for GOL06 (Electronic annex Table 3). From 12 analyses a weighted mean 208Pb/232Th age of 28.9 ± 0.8 Ma (MSWD = 1.9) and a lower 206Pb/238U age of 26.8 ± 0.7 Ma (MSWD = 1.2) can be calculated for sample GOL03 (Fig. 7d). Similarly, nine of 12 analyses for GOL06 calculate a weighted mean 208Pb/232Th age of 28.8 ± 1.2 Ma (MSWD = 2.3) and a lower 206Pb/238U age of 26.4 ± 0.8 Ma (MSWD = 1.9). Three analyses (GOL06.10-1, 5-1 and 8.1) overlapped the core-rim boundary and were omitted (Fig. 7c). For both samples, the low U–Pb ages relative to Th–Pb shift the data to the right of U–Th–Pb concordia (Fig. 7d). There is no correlation between the 207Pb-corrected single-spot ages and the amount of non-radiogenic 206Pb (and 208Pb) in each analysis (Fig. 7e and f), which suggests that initial Pb isotope composition of the sample does not deviate significantly from the model Pb composition used (see Section 9.1). A total 238U/206Pb–207Pb/206Pb plot of the six analyses yields an initial 207Pb/206Pb composition of 0.818 ± 0.017 (95% cl.) and a lower intercept age of 26 ± 3 Ma (MSWD = 2.3, 95% cl., Electronic annex Fig. 2). The U–Pb intercept age although less precise, is identical to the weighted mean U–Pb age.

6.4. Berzona

Five allanite grains from BER1 meta-granite were analysed in situ from polished thin section (Fig. 8a). In thin section, allanite is unzoned, with the exception of a single, large (~500 µm) grain comprised of a core (av. REE + Th = 0.65 cpfu) and overgrowth (av. REE + Th > 0.4 cpfu) (Fig. 8a). The high-BSE intensity core (Fig. 8b) has high LREE contents (av. LaN/LuN = 1090) relative to the low-BSE intensity overgrowth and other analysed grains (av. LaN/LuN ~518 and ~157) (Table 3; Fig. 8c). The REE composition of the allanite core is also distinguished from the overgrowth by a large negative Eu anomaly (av. Eu/Eu* = 0.16 for core and 0.62 for overgrowth) (Table 3). All allanite types in BER1 have low Th/U (Th/U = 14–18) (Table 3), in contrast to the high Th/U cores from previous samples. Thirty-one U–Th–Pb analyses of allanite were obtained by LA-ICPMS, including 6 core analyses and 25 overgrowth analyses (Electronic annex Table 4). The calculated proportion of radiogenic 206Pb for each analysis is 24–54% of total 206Pb (Electronic annex Table 4). Excluding the core analyses, the uncorrected data yield a Th–Pb isochron age of 25.0 ± 2.4 Ma (MSWD = 0.6) (Fig. 8d), which is considered a best estimate for the timing of Alpine allanite crystallization in this sample. The six core analyses lie off this isochron (Fig. 8d) and give an imprecise Th–Pb isochron age of 32 ± 34 Ma (MSWD = 0.2) (Electronic annex Fig. 3).

6.5. Age summary

High Th/U allanite cores in meta-tonalite, meta-granodiorite and meta-granite yield pre-Alpine 208Pb/232Th ages of 286 ± 5 Ma (VAM1), 297 ± 7 Ma (BEM1), 281 ± 5 Ma (GOL03) and 268 ± 10 Ma (GOL06). Notably, no pre-Alpine ages were reported from allanite in the leucosome samples. Alteration zones in allanite cores give low 208Pb/232Th and 206Pb/238U ages (down to ~163 Ma) and may have low Th/U compositions and high non-radiogenic Pb contents compared to unaltered cores. In contrast, allanite overgrowths and weak oscillatory-zoned allanite in the meta-granitoids and leucosomes give a range of Alpine ages from 20 ± 5 Ma (BEL1) to 30 ± 4 Ma (VAM2). Whereas the 208Pb/232Th and 206Pb/238U ages of pre-Alpine allanite generally agree within uncertainty, the Th–Pb and U–Pb age systems within allanite overgrowths differ in most samples (i.e. high 208Pb/232Th ages and low 206Pb/238U ages from the same analysis).

7. Bulk-rock trace element geochemistry

The bulk major and trace element compositions of two pairs of leucosome and meta-granitoid from Val d’Arbedo were investigated and are presented in Table 4. Leucosome samples VAL1 and VAL2 have major element compositions typical of granite. The REE patterns are enriched in LREE and flat for HREE (Fig. 9a). Sample VAL1 displays a small negative Eu-anomaly whereas VAL2 is characterised by a small positive Eu anomaly. The overall low LREE content together with low Zr content are in agreement with fluid-assisted melting at 650–700 °C as the main process for the formation of the leucosomes (Berger et al., 2008 and Rubatto et al., 2009).



Table 4. Bulk rock major and trace elements determined by LA-ICP-MS and XRF.










VAL1

VAL2

VAM2

VAM1




SiO2

71.6

69.4

53.5

50.8



Al2O3

15.7

17.3

18.7

19.6



Fe2O3

2.30

1.49

8.05

9.33



CaO

4.52

5.22

6.94

7.41



MgO

0.80

0.69

4.30

3.54



MnO

0.04

0.03

0.13

0.15



Na2O

3.08

4.68

3.53

3.28



K2O

2.27

0.70

2.50

2.73



TiO2

0.27

0.11

1.34

1.86



P2O5

0.10

0.06

0.26

0.62



F (ppm)

601

141

<1600

1830



Cl (ppm)

50

45

51

132








Total

100.7

99.7

99.5

99.7



Trace elements (average in ppm)




n = 3

n = 3

n = 3

n = 3



Sc

10

10

21

23



V

84

69

207

232



Cr

138

22

53

45



Rb

71

10

74

100



Sr

445

416

360

464



Y

4.8

5.2

23

37



Zr

56

33

171

342



Nb

2.6

1.4

14

24



Ba

404

164

512

700



La

6.4

4.0

24

56



Ce

13

8.1

49

110



Pr

1.5

0.98

6.2

13



Nd

6.3

4.0

26

52



Sm

1.3

0.90

5.3

10



Eu

0.35

0.40

1.5

2.3



Gd

1.3

0.90

4.9

8.8



Tb

0.16

0.14

0.70

1.2



Dy

0.95

1.0

4.5

7.3



Ho

0.19

0.19

0.88

1.4



Er

0.51

0.56

2.6

3.7



Tm

0.07

0.08

0.36

0.52



Yb

0.45

0.57

2.5

3.4



Lu

0.07

0.09

0.35

0.48



Hf

1.5

0.91

4.4

8.2



Ta

0.16

1.6

1.9

2.9



Pb

8.9

12

9.2

10



Th

1.6

0.95

4.3

11



U

0.26

0.30

0.92

1.3



Th/U

6.2

3.2

4.6

8.4


Table options







Fig. 9. Mineral and bulk rock trace element data determined by LA-ICPMS. (a) Chondrite-normalised REE-plots for country rock-leucosome pairs sampled at Val d’Arbedo (VAM1/VAL1 and VAM2/VAL2). (b) REE patterns of titanite, hornblende and plagioclase from sample VAM1. Note the very low LREE content in plagioclase and small negative Eu anomalies in titanite and hornblende. (c) REE-plots for zircon (Rubatto et al., 2009) and apatite. (d) Trace element distribution or budget for sample VAM1 meta-tonalite; see text for details.

Figure options




Meta-granitoid samples VAM1 and VAM2 have higher FeO and MgO contents than the leucosomes and have high total alkali contents (5–6 wt.%), suggesting tonalite as a possible protolith. However, the rocks show lower than expected SiO2 content (51 and 53.5 wt.%), which is attributed to melt extraction from the orthogneiss during anatexis. REE contents are generally higher in meta-tonalite compared to leucosome (Fig. 9a), reflecting the higher amount of trace element-rich phases such as titanite, allanite, apatite and amphibole in this rock type (see Section 8).

8. Mineral trace element compositions in meta-tonalite

Trace element compositions for silicate and accessory phases in VAM1 (Table 5) were obtained to examine the role of allanite as a trace element host during fluid-assisted partial melting. Chondrite-normalised rare earth element (REE) plots of each mineral are shown in Fig. 9b and c. Major mineral chemistry is provided in the Electronic annex Table 6 and only the key trace element features of minerals are discussed here. Using the mineral and bulk rock compositions from VAM1, it is possible to perform a mass balance in order to evaluate the major hosts for trace elements in meta-tonalite. The methodology for the mass balance calculations is detailed in the Electronic annex supplement.



Table 5. Average trace element compositions of VAM1 minerals determined by LA-ICP-MS.















Hornblende
--------------------------------------------------------------------------------


Plagioclase

Biotite

Titanite
--------------------------------------------------------------------------------


Apatite

Zircon
--------------------------------------------------------------------------------





Core

Rim

Core


Core

Rim


Core

Rim




Trace elements in ppm




n = 5

n = 4

n = 17

n = 13

n = 8

n = 7

n = 8

n = 2

n = 2



Li

17

15

0.16

85

0.46

0.84

0.89





Be

3.2

3.2

3.3

0.16

0.02

0.05

0.05





Sc

108

116

1.4

17

4.5

5.1

1.4





V

499

493

1.4

457

564

694

7





Cr

20

16

1.1

31

6.2

8.2

1.3





Rb

8.5

7.7

2.5

429

0.15

5.5

0.50





Sr

52

48

819

3.0

31

40

287

1.7

0.25



Y

45

20

0.04

0.12

504

2070

20

2840

400



Zr

18

17

<0.01

0.33

130

169

0.40

11

2.0



Nb

4.4

5.5

<0.02

3.5

772

880

5.5





Cs

0.01

0.01

<0.03

9.0

0.01

0.15

0.02





Ba

64

55

84

2400

0.20

29

2.2





La

0.15

0.05

0.08

0.01

1.5

35

2.1





Ce

0.93

0.24

0.14

0.01

10

207

7.8

24

6.0



Pr

0.27

0.06

0.02

<0.01

2.5

52

1.5

0.09

<0.01



Nd

2.4

0.49

0.07

0.01

20

380

10

1.5

0.21



Sm

2.0

0.44

<0.01

0.01

16

227

3.6

5.3

0.51



Eu

0.84

0.24

0.11

0.02

6.9

78

0.91

1.1

0.17



Gd

4.5

1.1

<0.01

0.06

34

345

5.2

41

4.5



Tb

1.0

0.29

<0.02

<0.01

8.7

62

0.61

17

2.0



Dy

7.7

2.8

<0.03

0.01

77

416

3.4

234

29



Ho

1.7

0.73

<0.04

<0.01

19

80

0.70

91

12



Er

5.0

2.6

<0.05

<0.02

63

215

1.9

448

67



Tm

0.67

0.39

<0.06

<0.03

10

28

0.23

101

17



Yb

4.2

2.7

<0.07

0.01

67

176

1.5

937

185



Lu

0.60

0.42

<0.08

<0.01

8.6

23

0.30

169

38



Hf

1.0

0.89

<0.09

0.05

8.5

11

0.05

10,400

1160



Ta

0.03

0.04

<0.10

0.08

81

39

0.47

5.7

2.5



Pb

6.6

6.1

23

3.8

1.9

2.6

3.6

16

2.0



Th

<0.01

<0.01

<0.01

<0.01

0.17

5

0.04

663

73



U

0.01

0.01

<0.01

0.01

2

29

0.38

2420

562



Eu/Eu*

0.85

1.1

–

–

0.91

0.86

0.64

0.16

0.22



GdN/LaN

36

29

–

–

26

12

3

–

–



Th/U

–

–

–

–

0.07

0.16

0.09

0.27

0.13


Table options





A key feature of the mineral trace element compositions in VAM1 is the LREE-depleted REE patterns of phases co-existing with allanite. Hornblende and titanite REE patterns are LREE-depleted with respect to M-HREE (GdN/LaN = 36 and 26, respectively), with a general increase in trace element concentration (including LREE) from core to rim (Fig. 9b). It is possible that some cores of titanite grains are of magmatic origin, but their trace element contents are significantly lower than that documented for igneous titanite in metaluminous granites (Table 5) (Bea, 1996). Apatite and plagioclase in VAM1 are also REE-depleted (e.g. subchondritic LREE in plagioclase, Fig. 9c) compared to REE contents of igneous plagioclase and apatite in metaluminous granites (Bea, 1996). The REE patterns of hornblende, titanite, allanite and zircon rims are characterised by small negative Eu anomalies (Fig. 9b and c, Table 5), despite their crystallization in a feldspar-bearing rock. The Th/U composition of titanite (av. Th/U <0.2) and apatite (av. Th/U <0.1) is notably low (Table 5). Similarly, zircon overgrowths in VAL1 leucosome and rare zircon rims in VAM1 are Th-depleted (Th/U <0.1–0.001, Rubatto et al., 2009) compared to inherited zircon cores in the meta-tonalite (Table 5).

The mass balance results for VAM1 are given in Fig. 9d, and the relative contribution of each mineral to the calculated bulk rock composition is shown. The diagram shows metamorphic allanite to be the principal host of LREE and Th in the meta-tonalite, in line with the observed LREE-depletions in mineral REE patterns. The contribution of allanite to the REE budget is reduced with increasing atomic number of the REE. Titanite is the principal host of HREE and contributes to MREE, whereas amphibole hosts about 20% of the HREE. Although plagioclase REE patterns display an accentuated positive Eu anomaly (Fig. 9b), it contributes only 2% to the Eu budget of the bulk rock (Fig. 9d). Apatite and zircon budget P, and Zr and Hf, respectively, while Rb and Ba are primarily hosted in biotite.

9. Discussion

9.1. Behaviour of initial Pb in allanite

Allanite may incorporate large and variable amounts of Pb into its structure on formation (“initial Pb”), effectively reducing the precision of U–Pb ages, and to a lesser degree, Th–Pb ages. The proportion of non-radiogenic 208Pb in pre-Alpine allanite is <3% of the total 208Pb. Consequently, the choice of initial Pb isotopic composition only had a slight effect on calculated Th–Pb ages and the use of a model composition (Stacey and Kramers, 1975) is considered appropriate.

The proportion of non-radiogenic 208Pb and 206Pb in Alpine allanite ranges from 20% to 100%. The amount of non-radiogenic Pb changes with bulk rock composition: allanite overgrowths in meta-tonalite (e.g. VAM/L1, VAM/L2) have typically higher non-radiogenic Pb contents and lower Th/U than in meta-granodiorite (e.g. BEM/L1, GOL06) and meta-granite (e.g. GOL03, BER1). Protolith composition is thus an important consideration in sample selection for allanite dating. For allanites with such high non-radiogenic Pb contents, particularly those with >50% initial Pb, the suitability of a model Pb isotope composition must be based on independent evidence. The initial Pb can be considerably radiogenic relative to a bulk-crust model value (Stacey and Kramers, 1975) if allanite incorporates at formation radiogenic Pb from a Th- or U-bearing precursor mineral (Romer and Siegesmund, 2003). Such a situation may be favoured for solid-state and incomplete metamorphic reactions or mineral replacements because element transfer is limited to a local environment (Gabudianu et al., 2009). The measured isotopic composition of a coexisting low-U phase (K-feldspar) can be used to reduce such data but this assumes isotopic equilibration was achieved prior to closure of Pb diffusion in allanite, and in most complex metamorphic rocks this is likely to be problematic. In the investigated samples, two generations of epidote/allanite formed during metamorphism and it is unclear which one is in equilibrium with plagioclase. Additionally, feldspar Pb composition may re-equilibrate during retrograde, subsolidus metamorphism, whereas allanite is unaffected by low-grade overprints. Alternatively, concordia and isochron plots are useful as they avoid the need for choosing an initial Pb isotopic composition (Aleinikoff et al., 2002). A potential caveat is obtaining a sufficient spread of data to define a statistically meaningful regression (Ludwig, 2003).

Fig. 10 demonstrates that model Pb isotope compositions predicted by Stacey and Kramers (1975) change little from the approximate time of granitoid intrusion (~280 Ma, 207Pb/206Pb = 0.864) to high grade Alpine metamorphism (~30 Ma, 207Pb/206Pb = 0.838). In contrast, intake at growth of radiogenic Pb inherited from the precursor pre-Alpine allanite lowers the initial 207Pb/206Pb significantly (Fig. 10). In fact, only 5% of such “inherited radiogenic Pb” (207Pb/206Pb at 280 Ma = 0.05176, Stacey and Kramers, 1975) lowers the initial 207Pb/206Pb to 0.80 (Fig. 10). The use of a model composition would underestimate the amount of non-radiogenic Pb in allanite and produce older apparent ages that increase with the proportion of initial Pb calculated (Gabudianu et al., 2009). For a radiogenic 207Pb/206Pb composition approximating that of magmatic allanite, the concordia method allows for detecting above 1% inherited radiogenic Pb in favourable cases (see data for sample VAM2, initial 207Pb/206Pb = 0.839 ± 0.005 2s, Fig. 5d). Independent estimates of initial 207Pb/206Pb from concordia intercepts ranged from 0.78 ± 0.11 (BEL1) to 0.86 ± 0.10 (VAM1) and agree, within error, with the model Pb isotope composition used ( Fig. 4, Fig. 5 and Fig. 10). For some samples the 207Pb/206Pb intercept has a large uncertainty. The 207Pb/206Pb value of allanite in sample BEL1 indicates that more than 5% of inherited Variscan Pb is possible (Fig. 10). Application of the concordia 207Pb/206Pb value to calculate the amount of non-radiogenic Pb in BEL1 allanite lowers the single spot ages by around 10%. The presence of some 5% or more “inherited radiogenic Pb” would therefore result in a spread of apparent ages positively correlated with non-radiogenic Pb content, in samples with relatively uniform Th/U compositions. This trend is not observed for any of the samples dated by SHRIMP (Figs. 6c, 7e–f). The SHRIMP data indicate only a limited transfer of radiogenic Pb (¿5%) between inherited-magmatic and newly-grown allanite during Alpine metamorphism. Although the application of a model Pb isotope composition is considered appropriate in this case, the U–Pb concordia ages are used in preference to single spot 206Pb/238U ages for the interpretation of allanite U–Pb data. For sample BER1, the Th–Pb isochron 208Pb/206Pb intercept at 2.58 ± 0.29 (2s) is higher than the model Pb value (208Pb/206Pb at ~30 Ma = 2.07, Stacey and Kramers, 1975). Given that most crustal rocks have ratios around 2.1 (Stacey and Kramers, 1975) the enhanced 208Pb/206Pb value may represent evidence for inheritance in this sample.



Fig. 10. Inverse Concordia plot of uncorrected U-Pb analyses for allanite overgrowths GOL03 and BEL1 and indicators showing the% effect of inherited radiogenic Pb from precursor allanite (207Pb/206Pb = 0.05176 at ~280, Stacey and Kramers, 1975) on the initial 207Pb/206Pb isotope composition of newly grown allanite. Error crosses are 1s.

Figure options




To assess the applicability of allanite dating in different geological settings, the non-radiogenic Pb contribution in allanite determined by ion microprobe was compiled for a range of magmatic, migmatitic and metamorphic rocks of Phanerozoic age (Fig. 11). In each case, the origin of allanite was inferred from P–T determinations, trace element compositions and petrography. The amount of non-radiogenic Pb, given in Fig. 11 as a proportion of the total measured thoranogenic Pb, is dependent on sample age and Th/U. The sample ages are ~12–290 Ma for magmatic allanite, ~30–550 Ma for migmatitic allanite and ~30 – 550 Ma for subsolidus metamorphic allanite (Fig. 11 and references therein). The allanite Th/U ratios are ~30–250 (magmatic), ~3–290 (migmatitic) and ~2–45 (metamorphic). The data in Fig. 11 show a relationship between allanite origin and non-radiogenic Pb content, irrespective of sample age.



Fig. 11. Compilation of the fraction of non-radiogenic Pb of the total measured 208Pb (f-208) in allanite from different geological settings. Samples analysed in this study are in bold. Other compositions are from allanite described in the literature. Magmatic samples: CAP allanite (276 Ma, ID-TIMS Th–Pb age, Barth et al., 1994); Tara allanite (415 Ma, SHRIMP Th–Pb age, Gregory et al., 2007); AVC allanite (276 Ma, ID-TIMS Th–Pb age, Barth et al., 1994); Siss (Bergell) allanite (31.5 Ma, ID-TIMS Th–Pb age, von Blanckenburg et al., 1992); Diabosatsu allanite (12 Ma, SHRIMP Th–Pb age, Gregory, 2009); Bona (Bergell) allanite (30.1 Ma, ID-TIMS Th–Pb age, von Blanckenburg et al., 1992); BC allanite (92 Ma, SHRIMP Th–Pb age, Gregory et al., 2007); PE13 allanite (550 Ma, SHRIMP Th–Pb age, Gregory et al., 2009a and Gregory et al., 2009b); PE13 allanite (559 Ma, SHRIMP Th–Pb age, Gregory et al., 2009a and Gregory et al., 2009b); MF161 allanite (29 Ma, SHRIMP Th–Pb age, Janots et al., 2009); APi0413 allanite (31 Ma, SHRIMP Th–Pb age, Janots et al., 2009); WS2 (31 Ma, SHRIMP U–Pb age, Gabudianu et al., 2009); La-VdT-2 allanite (47 Ma, SHRIMP Th–Pb age, Rubatto et al., 2008), TAW1 allanite (~30 Ma, SHRIMP U–Pb age, Gregory, 2009).

Figure options




Igneous allanites with high Th/U contain relatively small amounts of non-radiogenic 208Pb (~1–35%). In comparison, allanites that form under subsolidus conditions have typically high non-radiogenic 208Pb (>50%) and low Th/U. Certainly Th/U is responsible, in part, for the low non-radiogenic Pb contents in magmatic allanite and high non-radiogenic Pb contents in metamorphic allanite shown in Fig. 11. It is worth noting, however, that migmatitic allanite with high Th/U (e.g. ~60–290, sample VAM2) can contain much higher non-radiogenic 208Pb contents (89–100%) than magmatic allanite of equivalent age and Th/U composition (e.g. the Bergell allanites). Thus, the amount of initial Pb may depend also on the process/reaction by which allanite formed.

Whilst non-radiogenic Pb is limiting for dating, particularly for metamorphic allanite containing relatively low Th and/or U, it may be useful l as a petrological tool. Fig. 11 suggests that the amount of non-radiogenic Pb can be used as an indicator for subsolidus versus suprasolidus allanite. In addition, the composition of the initial Pb (e.g. presence of “inherited radiogenic Pb”) can potentially be used as a tracer for the recrystallization processes of allanite. The fact that no appreciable amount of “inherited radiogenic Pb” entered metamorphic allanite at formation indicates that the Pb isotope composition had been homogenised during the partial melting process. In this sense, metamorphic allanite may be used as a Pb isotope sensor.

9.2. U–Pb versus Th–Pb ages in allanite

A number of studies have demonstrated the importance of the Th–Pb system for accurate dating of Th-enriched minerals, such as allanite (von Blanckenburg et al., 1992, Barth et al., 1994, Oberli et al., 2004 and Gregory et al., 2007). The Th–Pb ages of high Th/U allanite are typically more precise and less sensitive to non-radiogenic Pb correction than U–Pb (208Pb radiogenic contents exceed 206Pb). Furthermore, the Th–Pb age is not affected by initial radioactive disequilibrium (excess 206Pb from unsupported 230Th), which may yield 206Pb/238U ages older than the 208Pb/232Th age on the same mineral (Oberli et al., 2004). We therefore prefer the Th–Pb system to date the high Th/U allanite cores. Excess 206Pb contributions are not detected in pre-Alpine allanite: U–Pb and Th–Pb ages from the same grain are consistent at the level of precision of the dating technique (±2–3%, 1s) (Fig. 12). The similar trace element compositions of the allanite samples and standards used to normalise ion microprobe data (LREE + Th = 0.6–0.8 cpfu; Gregory et al., 2007) minimized potential bias in 206Pb*/238U+ and 208Pb*/232Th+ caused by matrix mis-matching.



Fig. 12. Probability distribution diagram of single-spot 206Pb/238Pb ages (grey line) and 208Pb/232Th ages (black line) determined by SHRIMP for allanite samples VAM/L1, VAM/L2, BEM/L1, GOL03 and GOL06. Alpine ages are low Th/U allanite overgrowths and high Th/U weak oscillatory zoned allanite, and pre-Alpine ages are high Th/U allanite cores. Inset: enlargement of 206Pb/238Pb and 208Pb/232Th ages of allanite and REE-epidote overgrowths only.

Figure options




In contrast, allanite and REE-epidote overgrowths are low in Th and Th/U (Table 3) and may have equal amounts of radiogenic 208Pb and 206Pb (Electronic annex Tables 1 and 2). Valuable geochronological information may thus be gained from both the U–Pb and Th–Pb isotopic systems in low Th/U allanite. A comparison of the two geochronometers in Alpine allanite reveals that the U–Pb and Th–Pb ages on the same grain may differ by up to 25% with systematically low U–Pb ages compared to Th–Pb (Fig. 12). The decoupling of U–Pb and Th–Pb geochronometers in allanite may be explained by isotope systematics, such as initial isotopic heterogeneity of Pb (Romer and Siegesmund, 2003 and Oberli et al., 2004) and isotopic resetting (Barth et al., 1994), or by analytical complications due to matrix effects (Catlos et al., 2000). We demonstrated above that only a small amount of “inherited radiogenic Pb” (¿5%) is possible in the allanite overgrowths and would not account for the observed difference between U–Pb and Th–Pb ages. Excess 206Pb is not relevant for low Th/U allanite, especially for the Val d’Arbedo and Bellinzona samples (Th/U = 3–20, av. 8), and if present, would produce older U–Pb ages, which are not observed.

SHRIMP dating is highly matrix dependent due to the effect of mineral chemistry on secondary ionisation efficiency (Williams, 1998). To achieve high accuracy in U–Th–Pb isotope analysis SHRIMP analyses must be calibrated against a matrix-matched standard (Stern and Berman, 2000, Fletcher et al., 2004 and Fletcher et al., 2010). The Alpine overgrowths dated in this study have much lower trace element contents (REE-epidote LREE + Th = 0.2 cpfu, allanite = >0.2–0.5 cpfu) than pre-Alpine allanite and the standard CAP (Gregory et al., 2007). Such compositional differences produced a systematic offset in standard and sample ThO+/Th+ but did not noticeably affect UO+/U+ (see Electronic annex Tables 1–4). The offset in ThO+/Th+ is probably due to the different relative proportion of complex molecules produced by different allanite matrices during the sputtering process. The low oxide production of allanite and REE-epidote overgrowths (i.e. low ThO+/Th+) compared to that of CAP allanite can be explained by the different LREE + Th contents of the standard and samples because REE + Th incorporation represents the primary element substitution in allanite (Gieré and Sorenson, 2004). This offset places a greater dependence on the accuracy of the 2-dimensional calibration slope, thereby increasing the uncertainty on the standard calibration. Based on the good agreement of standard and sample UO+/U+ values we conclude that U–Pb (concordia) ages for low Th/U allanite are more robust than Th–Pb ages, in this case. It is clear that the apparent decoupling of U–Pb and Th–Pb ages in allanite, either by isotope systematics or by analytical effects, requires more systematic investigation of both geochronometers. The development of a low-REE-Th standard for the accurate calibration of ion microprobe 206Pb/238U and 208Pb/232Th analyses is highly desirable for SHRIMP dating of young, metamorphic allanite.

9.3. Isotopic resetting in allanite

High spatial resolution dating isolated two main stages of allanite growth. The allanite cores from meta-granitoids have trace element compositions typical of magmatic allanite in tonalite and granodiorite (Gregory et al., 2009b): high La/Sm and Th/U and marked negative Eu anomalies (Fig. 13a–b). The ages of allanite cores (from 268 ± 10 to 297 ± 7 Ma) are consistent with the ages of inherited zircon (Rubatto et al., 2009) and the timing of Permian granitoid intrusion in the Lepontine Domain (Romer et al., 1996 and Schaltegger and Gebauer, 1999). We conclude that the allanite cores are inherited from the igneous protolith. A later generation of allanite overgrowths and new grains in meta-granitoids and leucosomes are characterised by low La/Lu and Th/U and small negative Eu anomalies (Fig. 13a–b). These yield a range of ages from 30 ± 4 to 20 ± 5 Ma (Fig. 14), which fall within the period recorded for Alpine collisional orogeny (Romer et al., 1996, Berger et al., 2005 and Janots et al., 2009), and also for anatectic zircons in the same area (Rubatto et al., 2009).



Fig. 13. (a) La/Sm versus Th/U plot for allanite and bulk rock samples. (b) La/Sm versus Eu/Eu* plot for allanite and bulk rock samples.

Figure options






Fig. 14. Summary of allanite ages for each sample (including 1s uncertainties): white squares = allanite in country rock (meta-granitoid); white circles = allanite in leucosome. Ages given for samples VAL1, VAL2 and GOL06 are weighted mean U–Pb ages, sample ages for VAM1, VAM2, BEL1 and GOL03 are U–Pb intercept ages and sample BER1 is a Th–Pb isochron age. Pre-Alpine ages are weighted mean Th–Pb ages.

Figure options




The detection of inherited cores in composite grains has implications for isotopic resetting of allanite in high-grade rocks. Given the low solubilities of LREE and Th in hydrous granitic melts (Montel, 1993), it is predicted that high LREE and Th allanite will remain as a residual phase during incipient partial melting. This prediction is consistent with the preservation of inherited cores in meta-granitoids and indicates that the conditions of partial melting (T = 620–700 °C from Ti-in-zircon thermometry, Rubatto et al., 2009) were insufficient to completely dissolve protolith allanite. Instead, protolith allanite preserves a substantial memory of its initial age in spite of upper amphibolite facies re-working, which places strong constraints on closure temperature. For allanite grains >100 µm (typical grain size of inherited cores) the closure temperature of Pb diffusion is above that of the wet granite solidus (i.e. T > 630 °C). The robustness of allanite is even more significant in samples from Bellinzona and Val d’Arbedo where there is good evidence of prolonged high temperatures from zircon dating and Ti-in-zircon thermometry (Rubatto et al., 2009). This observation is consistent with, and improves on other empirical estimates of Pb closure temperature in allanite from high-grade gneisses (~650 °C, Heaman and Parrish, 1991 and Parrish, 2001) and calc-alkaline igneous rocks (700–800 °C, von Blanckenburg et al., 1992 and Oberli et al., 2004).

The low and scattered apparent ages (~280 to ~163 Ma) of alteration zones within inherited cores, however, indicate partial resetting of protolith allanite. Here, alteration is defined by irregular domains of patchy, low-BSE intensity zoning that proceed from fractures and core boundaries (Fig. 6b). The texture and U–Th–Pb isotopes points to partial loss of (radiogenic) Pb from metamict zones, possibly facilitated by fluid along fracture pathways that lead to geologically meaningless ages. Some analyses with low Th/U, including the low Th/U allanite core in sample BER1, may reflect partial recrystallisation of protolith allanite at higher metamorphic temperatures, although similar features (i.e. low-BSE intensity, low Th contents and high non-radiogenic Pb) have also been described for the alteration of magmatic allanite at low temperatures (~200 °C, Poitrasson, 2002). The agreement of U–Pb and Th–Pb ages (within ±1s) obtained from altered allanite suggests that the U–Pb and Th–Pb age systems in the analysed areas behaved similarly during secondary alteration, and were not significantly decoupled by secondary processes (e.g. Barth et al., 1994).

The identification of inherited cores partially replaced by new allanite overgrowths is of general importance for the interpretation of allanite ages in metamorphic rocks. The survival of igneous allanite grains to upper amphibolite facies and even anatexis coupled with the high reactivity of allanite during metamorphism suggests that allanite ages are likely to be reset by new mineral growth and recrystallization rather than by diffusion. The importance of mineral growth/recrystallization over diffusive resetting is supported by the fact that only Alpine ages were measured from newly-formed, low Th/U allanite, which indicates that growth/recrystallization processes were the dominant mechanism for new allanite ages in the migmatites.

9.4. REE distribution during incipient melting: constraints on allanite petrogenesis

Allanite is chemically complex, has a large stability field, and may form by several processes/reactions (Gieré and Sorenson, 2004). Trace element variations in allanite have proven valuable in relating this mineral to an evolving bulk silicate assemblage. For example, negligible Eu anomalies and high Sr content in allanite are proxies for its formation above the stability field of plagioclase in high pressure rocks (Rubatto et al., 2008), while changes in allanite HREE content have been used to infer the involvement of garnet (Gregory et al., 2009a).

In this study, the composition and origin of allanite are coupled. Magmatic allanite is characterised by higher Th and LREE contents than metamorphic allanite (Fig. 4b). As allanite is the main carrier of Th in these rocks (Fig. 9d), this distribution indicates that the modal abundance of allanite was smaller in the magmatic protholith than in the recrystallised meta-granitoid. Magmatic allanite has higher Th/U, generally displays a steeper LREE pattern (higher La/Sm) and has a more pronounced negative Eu anomaly (lower Eu/Eu*) than metamorphic allanite (Fig. 10a–b). These changes are best explained by a change in trace element budgets in the coexisting magmatic and metamorphic phases. Gregory et al. (2009b) showed that in the Bergell tonalite magmatic titanite has high trace element contents with normalized MREE ¿ LREE. Higher modal abundance of titanite with respect to allanite results in a LREE pattern of allanite that is steeper than that of the bulk rock. In such magmatic rocks, plagioclase is the most important host of Eu and thus crystallization of plagioclase produces a negative Eu anomaly in all co-existing minerals. These trends are observed in the magmatic allanite studied here. The La/Sm of allanite is higher than the measured bulk rock whereas Eu/Eu* is smaller (Fig. 10a–b). The situation is significantly different during the metamorphic growth or recrystallization. Allanite is the main host for LREE and Sm and thus the La/Sm of allanite will be determined by the bulk composition. This is confirmed for the Val d’Arbedo samples (Fig. 10b). Interestingly, our mass balance shows that although plagioclase is the most abundant phase in the meta-granitoid, it hosts only 2% of the Eu (Fig. 9d) and thus is not able to impose a significant negative Eu anomaly in the coexisting phases. However, metamorphic allanite influences the trace element patterns of other metamorphic phases. The LREE depletion of amphibole and titanite (Fig. 8b) as well as the very low Th/U of metamorphic zircon and titanite are likely related to coexisting allanite.

The anomalous HREE-depletions in weakly oscillatory-zoned allanite from VAM2 meta-tonalite compared to unzoned overgrowths in the same sample (Fig. 5b), indicates that the weakly zoned allanite formed in the presence of stable phase that sequestered HREE, such as garnet or zircon. Early (~32 Ma) metamorphic zircon in leucosome VAL2 is also HREE-depleted (Rubatto et al., 2009). Titanite and hornblende contain significant HREE but they would additionally result in MREE depletion (Fig. 8d). Whilst garnet is not observed in samples VAM2 and VAL2, it is documented as a residual phase in migmatites from Val d’Arbedo and Bellinzona (Berger et al., 2008 and Rubatto et al., 2009). Mineral equilibria modelling of Berger et al. (2008) shows that garnet can form at P–T conditions appropriate to the studied samples (i.e. between 650 and 750 °C at 0.8 GPa) in an average migmatite composition, and even small changes in pressure or bulk rock composition (e.g. Ca content) may render garnet metastable. Therefore, we suggest that garnet was likely stable in the sample when early, low HREE allanite formed at ~30 Ma.

9.5. Behaviour of mineral chronometers during incipient melting

Zircon is most commonly used chronometer to date migmatites (Vavra et al., 1996, Rubatto et al., 2001 and Hokada and Harley, 2004), whereas allanite and titanite are more likely to record low- to medium-grade events in metamorphic rocks (Aleinikoff et al., 2002 and Janots et al., 2009). Here we compare the behaviour of allanite, titanite and zircon (Rubatto et al., 2009) in the migmatites to assess their role as chronometers of incipient partial melting.

New zircon growth in migmatites occurred almost exclusively in leucosomes and was triggered by repeated injections of melt into the system (Rubatto et al., 2009). Zircon in country rocks (meta-granitoids) either failed to re-equilibrate (e.g. sample VAM1) or produced rare Alpine overgrowths (e.g. samples VAM2 and BEM1), probably owing to efficient melt segregation (Rubatto et al., 2009). Even in migmatites containing moderate amounts of leucosome (sample GOL03, Fig. 2d), metamorphic zircon is scarce. As a result, zircon dating was not attempted for similar migmatites at Berzona (sample BER1, Fig. 2e). The occurrence of mainly inherited zircon in country rocks indicates that metamorphic zircon growth during Barrovian metamorphism in the Central Alps was limited to areas of intense migmatisation (Val d’Arbedo–Bellinzona, Rubatto et al., 2009), where hydrate-breakdown melting was additionally reported in muscovite-bearing rocks (Fig. 1) (Burri et al., 2005).

In contrast, metamorphic allanite and titanite are more reactive and may form by different processes to metamorphic zircon. Alpine allanite formed in country rocks containing small leucosome volumes, and in segregated leucosomes with zircon (Fig. 14). Similarly, titanite with low trace element contents (interpreted as metamorphic titanite) occurred in both country rocks and leucosomes (Table 1). Berger et al. (2008) interpreted allanite textures in migmatite at Bellinzona to have formed by partial resorption of epidote (either prograde or magmatic) during incongruent melting followed by the precipitation of allanite overgrowths (+hornblende and accessory phases) from a melt (Mogk, 1992). Irregular core-rim boundaries of some allanite grains (Fig. 4, Fig. 5, Fig. 6, Fig. 7 and Fig. 8) however, make it difficult to distinguish between new growth on magmatic cores or recrystallization of the outerpart of a magmatic core. Newly grown allanite is abundant in migmatites that contain limited or rare metamorphic zircon, such as at Golino and Berzona, and is the principal U–Pb chronometer in these rocks (Fig. 14). The comparison of allanite and zircon in country rock–leucosome pairs in this study and in Rubatto et al. (2009) highlights the difficulty of dating amphibolite facies anatectic rocks using only zircon. Whereas new zircon may form in leucosomes, allanite (and titanite) could be better chronometers in the country rock (restite).

Allanite and titanite are found in many metamorphosed granitic rocks and are abundant in metaluminous rocks of intermediate composition, such as the meta-tonalites and meta-granodiorites investigated in this study. They are stable in metamorphic rocks to amphibolite facies, and the closure temperatures of Pb diffusion in both minerals lie at the upper limit of amphibolite facies, or possibly higher (T = >650–700 °C, this study, Aleinikoff et al., 2002). Allanite and titanite are therefore key minerals for dating medium to high temperature metamorphism. The utility of these chronometers may be limited by low U (and Th) contents, particularly for titanite, and high non-radiogenic Pb contents in metamorphic grains (Fig. 11) (Aleinikoff et al., 2002). The extremely low U and Th contents in titanite (from core to rim: 2–30 ppm U and 0.1–5 ppm Th) in the studied migmatites made it unsuitable for ion microprobe dating.

9.6. Implications of allanite ages for the Central Alps

New dating shows that allanite in meta-granitoids and leucosomes recorded different partial melt events during the Barrovian cycle between 30 ± 4 and 20 ± 5 Ma (Fig. 14). The distribution of allanite ages in different samples is in line with the episodic melt model of Rubatto et al. (2009), whereby fluid-assisted melting and new mineral growth in the migmatites is intermittent and heterogeneously distributed, due to the availability of localised fluids (Berger et al., 2008). The timing of allanite growth in migmatites at Golino and Berzona (25.0 ± 2.4 and 26 ± 3 Ma) compared to allanite ages from Val d’Arbedo and Bellinzona (from 20 ± 5 to 30 ± 4 Ma) suggests that melting ceased earlier in rocks away from the areas of intense migmatisation (Fig. 14).

Incipient partial melting in the Southern Steep Belt is contemporaneous with peak magmatism in the Central Alps, with the emplacement of the Bergell intrusive suite between 33 and 28 Ma (Oberli et al., 2004) and intrusion of the Novate leucogranite at 25 Ma (Liati et al., 2000). In addition, the widespread formation of late to post-kinematic ~29–25 Ma leucocratic dykes in the western part of the SSB (Romer et al., 1996 and Schärer et al., 1996) and the ages of monazite growth in similar migmatites at 28 and 30 Ma (Berger et al., 2009) indicate that crustal melting was occurring over a period of time until the late Oligocene. Repeated allanite formation from around 30 to 20 Ma in migmatites across the SSB is consistent with the age pattern of anatectic zircons in the same rocks (32–22 Ma at Val d’Arbedo–Bellinzona, Rubatto et al., 2009). The allanite U–Pb ages therefore support the fast exhumation and cooling history (100 ± 20 °C/Ma) of the migmatite belt proposed by Rubatto et al. (2009).

10. Conclusions

Migmatites of the Central Alps provide an excellent example of allanite chronology in amphibolite facies rocks. The complex history of allanite in the investigated samples demonstrates the value of microanalysis and the additional information which may be gained from this mineral using such a technique. Allanite was present in the migmatites as an igneous accessory and as a product of subsequent metamorphism and melting. The texture, chemistry and U–Th–Pb isotopes within allanite indicate that metamorphic overgrowths did not inherited appreciable amounts (¿5%) of radiogenic Pb from precursor magmatic allanite. Although protolith allanite underwent varying degrees of post-magmatic alteration, this study demonstrates that igneous allanite grains can survive to upper amphibolite facies and even anatexis and still retain a substantial component of the original age. Thus, in addition to its formation at lower metamorphic temperatures, allanite may provide robust metamorphic ages to at least upper amphibolite facies. The presence of non-radiogenic Pb in allanite, however, may limit dating applications in certain cases, particularly in low grade metamorphic rocks. The apparent decoupling of U–Pb and Th–Pb chronometers within low Th/U allanite requires further systematic investigation of allanite U–Pb and Th–Pb isotope systematics and the development of a low-Th-REE standard for ion microprobe analysis. In migmatite samples that lacked metamorphic zircon, allanite readily recorded the Alpine event. U–Th–Pb isotopes in allanite therefore present a complementary approach to zircon for dating incipient partial melting.



Acknowledgements

U. Troitzsch, S. Paxton, C. Allen and F. Brink are thanked for their technical support and advice. The Electron Microscopy Unit at The Australian National University provided access to SEM facilities. The thorough reviews and constructive comments of R. Parrish, J. Aleinikoff and an anonymous reviewer greatly improved the manuscript. The editorial handling of Y. Amelin is gratefully acknowledged. This work was funded by the Australian Research Council (DP0556700 to D.R. and J.H.).



Appendix A. Supplementary data




Supplementary data 1.

Help with DOC files

Options






References
Aleinikoff et al., 2002
J.N. Aleinikoff, R.P. Wintsch, C.M. Fanning, M.J. Dorais

U–Pb geochronology of zircon and polygenetic titanite from the Glastonbury Complex, Connecticut, USA: an integrated SEM EMPA, TIMS and SHRIMP study


Chem. Geol., 188 (2002), pp. 125–147



Barth et al., 1994
S. Barth, F. Oberli, M. Meier

Th–Pb versus U–Pb isotope systematics in allanite from co-genetic rhyolite and granodiorite: implications for geochronology


Earth Planet. Sci. Lett., 124 (1994), pp. 149–159



Bea, 1996
F. Bea

Residence of REE, Y, Th and U in granites and crustal protoliths; implications for the chemistry of crustal melts


J. Petrol., 37 (1996), pp. 521–552



Berger et al., 1996
A. Berger, C. Rosenberg, S.M. Schmid

Ascent, emplacement and exhumation of the Bergell pluton within the Southern Steep Belt of the Central Alps


Schweiz. Mineral. Petrogr. Mitt., 72 (1996), pp. 357–382



Berger et al., 2005
A. Berger, I. Mercolli, M. Engi

The Lepontine Alps: notes accompanying the tectonic and petrographic map sheet Sopra Ceneri (1:100,000)


Schweiz. Mineral. Petrogr. Mitt., 85 (2005), pp. 109–146



Berger et al., 2008
A. Berger, T. Burri, P. Alt-Epping, M. Engi

Tectonically controlled fluid flow and water-assisted melting in the middle crust: an example from the Central Alps


Lithos, 102 (2008), pp. 598–615



Berger et al., 2009
A. Berger, C. Rosenberg, U. Schaltegger

Stability and isotopic dating of monazite and allanite in partially molten rocks: examples from the Central Alps


Swiss J. Geosci., 102 (2009), pp. 15–29



Burri et al., 2005
T. Burri, A. Berger, M. Engi

Tertiary migmatites in the Central Alps: regional distribution, field relations, conditions of formation and tectonic implication


Schweiz. Mineral. Petrogr. Mitt., 85 (2005), pp. 215–232



Butler et al., 2002
R.F. Butler, G.E. Gehrels, S.L. Baldwin

Palaeomagnetism and geochronology of the Ecstall Pluton in the Coast Mountains of British Columbia: evidence for local deformation rather than large-scale transport


J. Geophys. Res., 107 (2002) EPM3-1–13



Catlos et al., 2000
E.J. Catlos, S.S. Sorensen, T.M. Harrison

Th–Pb ion-microprobe dating of allanite


Am. Mineral., 85 (2000), pp. 633–648



Cox et al., 2003
R.A. Cox, D.H.C. Wilton, J. Kôsler

Laser-ablation U–Th–Pb in situ dating of zircon and allanite: an example from the October Harbour Granite, central Coast Labrador, Canada


Can. Mineral., 41 (2003), pp. 273–291



Eggins et al., 1998
S.M. Eggins, R.L. Rudnick, W.F. McDonough

The composition of peridotites and their minerals: a laser ablation ICP-MS study


Earth Planet. Sci. Lett., 154 (1998), pp. 53–71



Engi et al., 1995
M. Engi, C.S. Todd, D.R. Schmatz

Tertiary metamorphic conditions in the eastern Lepontine Alps


Schweiz. Mineral. Petrogr. Mitt., 75 (1995), pp. 346–369



Engi et al., 2001
M. Engi, B. Berger, G. Roselle

The role of the tectonic accretion channel in collisional orogeny


Geology, 29 (2001), pp. 1143–1146



Engi et al., 2004
M. Engi, R. Bousquet, A. Berger

Explanatory notes of the map: metamorphic Structure of the Alps. Central Alps


Mitt. Öster. Min. Ges., 149 (2004), pp. 53–69



Finger et al., 1998
F. Finger, I. Broska, M.P. Roberts, A. Schermaier

Replacement of primary monazite by apatite-allanite-epidote coronas in amphibolite facies granite gneiss from the eastern Alps


Am. Mineral., 83 (1998), pp. 248–258



Fletcher et al., 2004
I.R. Fletcher, N.J. Mc Naughton, J.A. Aleinikoff, B. Rasmussen, S.L. Kamo

Improved calibration procedures and new standards for U–Pb and Th–Pb dating of Phanerozoic xenotime by ion microprobe


Chem. Geol., 209 (2004), pp. 295–314



Fletcher et al., 2010
I.R. Fletcher, N.J. McNaughton, W.J. Davis, B. Rasmussen

Matrix effects and calibration limitations in ion probe U–Pb and Th–Pb dating of monazite


Chem. Geol., 207 (2010), pp. 31–44



Gabudianu et al., 2009
I. Gabudianu, D. Rubatto, C. Gregory, R. Compagnoni

The age of HP metamorphism in the Gran Paradiso Massif, Western Alps: a petrologic and geochronologic study of a “silvery micaschist”


Lithos, 110 (2009), pp. 98–108



Gebauer, 1996
D. Gebauer

A P-T-t path for an (ultra)high-pressure ultra-mafic/mafic rock-association and its felsic country-rocks based on SHRIMP-daring of magmatic and metamorphic zircon domains. Example: Alpe Arami (Central Swiss Alps)


A. Basu, S.R. Hart (Eds.), Earth Processes: Reading the Isotopic Code, American Geophysical Union, Washington, DC (1996), pp. 309–328



Gieré and Sorenson, 2004
Gieré R., Sorenson S. (2004) Allanite and other REE-rich epidote-group minerals. In Epidotes (eds. F. Axel, F. Gerhard) . Reviews in Mineralogy and Geochemistry 56, pp. 431–493.


Gregory, 2009
Gregory C. J. (2009) Allanite chemistry and U–Th–Pb geochronology. Ph. D. thesis, Research School of Earth Sciences, The Australian National University Canberra, p. 271.


Gregory et al., 2007
C.J. Gregory, D. Rubatto, C. Allen, I.S. Williams, J. Hermann, T. Ireland

Allanite micro-geochronology: a LA-ICPMS and SHRIMP U–Th–Pb study


Chem. Geol., 245 (2007), pp. 162–182



Gregory et al., 2009a
C.J. Gregory, I.S. Buick, J. Hermann, D. Rubatto

Mineral-scale trace element and U–Th–Pb age constraints on metamorphism and melting during the Petermann Orogeny (central Australia)


J. Petrol., 50 (2009), pp. 251–287



Gregory et al., 2009b
C. Gregory, C. McFarlane, J. Hermann, D. Rubatto

Tracing the evolution of calc-alkaline magmas: in-situ Sm–Nd isotope studies of accessory minerals in the Bergell Igneous Complex, Italy


Chem. Geol., 260 (2009), pp. 73–86



Gromet and Silver, 1983
P.L. Gromet, L.T. Silver

Rare earth element distributions among minerals in a granodiorite and their petrographic implications


Geochim. Cosmochim. Acta, 47 (1983), pp. 925–939



Heaman and Parrish, 1991
Heaman L., Parrish R. R. (1991) U–Pb geochronology of accessory minerals. In Applications of radiogenic isotope systems to problems in geology (eds. L. Heaman, J. Ludden) Mineral Assoc Can Short-course Handbook 19, 59–102.


Hermann, 2002
J. Hermann

Allanite: thorium and light rare earth element carrier in subducted crust


Chem. Geol., 192 (2002), pp. 289–306



Hokada and Harley, 2004
T. Hokada, S.L. Harley

Zircon growth in UHT leucosome: constraints from zircon-garnet rare earth element (REE) relations in Napier Complex, East Antarctica


J. Mineral. Petrol. Sci., 99 (2004), pp. 180–190



Janots et al., 2006
E. Janots, F. Negro, F. Brunet, B. Goffe, M. Engi

Evolution of the REE mineralogy in HP-LT metapelites of the Sebtide complex, Rif, Morocco: monazite stability and geochronology


Lithos, 87 (2006), pp. 214–234



Janots et al., 2008
E. Janots, M. Engi, A. Berger, J. Allaz, J.-O. Schwarz, C. Spandler

Prograde metamorphic sequence of REE minerals in pelitic rocks of the Central Alps: implications for allanite–monazite–xenotime phase relations from 250 to 610 °C


J. Metamorph. Geol., 26 (2008), pp. 509–526



Janots et al., 2009
E. Janots, M. Engi, D. Rubatto, A. Berger, C. Gregory, M. Rahn

Metamorphic rates in collisional orogeny from in situ allanite and monazite dating


Geology, 37 (2009), pp. 11–14



Krenn and Finger, 2009
E. Krenn, F. Finger

Formation of monazite and rhadophane at the expense of allanite during Alpine low temperature retrogression of metapelitic basement rocks from Crete, Greece: microprobe data and geochronological implications


Lithos, 95 (2009), pp. 130–147



Liati et al., 2000
A. Liati, D. Gebauer, M. Fanning

U–Pb SHRIMP dating of zircon from the Novate granite (Bergell, Central Alps): evidence for Oligocene-Miocene magmatism, Jurassic/Cretaceous continental rifting and opening of the Valais trough


Schweiz. Mineral. Petrogr. Mitt., 80 (2000), pp. 305–316



Ludwig, 2003
K.R. Ludwig

Isoplot/Ex version 3. 0. A geochronological toolkit for Microsoft ExcelBerkeley Geochronological Centre Spec, Pub., Berkeley (2003) p. 70



Ludwig, 2009
K.R. Ludwig

SQUID 2 (rev. 2.50), A User’s ManualBerkeley Geochronology Ctr. Special Publication vol. 5 (2003) p. 104



Mogk, 1992
D.W. Mogk

Ductile shearing and migmatisation at mid crustal levels in an Archaean high grade gneiss belt, northern Gallatin range, Montana, USA


J. Metamorph. Geol., 10 (1992), pp. 427–438



Montel, 1993
J.-M. Montel

A model for monazite/melt equilibrium and application to the generation of granitic magmas


Chem. Geol., 119 (1993), pp. 127–146



Oberli et al., 2004
F. Oberli, M. Meier, A. Berger, C.L. Rosenberg, R. Giere

U–Th–Pb and 230Th/238U disequilibrium isotope systematics: precise accessory mineral chronology and melt evolution tracing in the Alpine Bergell intrustion


Geochim. Cosmochim. Acta, 68 (2004), pp. 2543–2560



Parrish, 2001
R.R. Parrish

The response of mineral chronometers to metamorphism and deformation in orogenic belts


J.A. Miller, R.E. Holdsworth, I.S. Buick, M. Hand (Eds.), Continental Reactivation and Reworking, 184Geol Soc Lon Spec Pub, Publishing house located in Bath, United Kingdom (2001), pp. 289–301



Parrish et al., 2006
R.R. Parrish, S.J. Gough, M.P. Searle, D.J. Waters

Plate velocity exhumation of ultrahigh-pressure eclogites in the Pakistan Himalaya


Geology, 34 (2006), pp. 989–992



Petrík et al., 1995
I. Petrík, I. Broska, J. Lipka, P. Siman

Granitoid allanite-(Ce) substitution relations, redox conditions and REE distributions (on example of I-type granitoids, Western Carpathians, Slovakia)


Geol. Carpathica, 46 (1995), pp. 79–94



Poitrasson, 2002
F. Poitrasson

In situ investigations of allanite hydrothermal alteration: examples from calc-alkaline and anorogenic granites of Corsica (southeast France)


Contrib. Mineral. Petrol., 142 (2002), pp. 485–500



Rasmussen and Muhling, 2009
B. Rasmussen, J. Muhling

Reactions destroying detrital monazite in greenschist-facies sandstones from the Witwatersrand basin, South Africa


Chem. Geol., 264 (2009), pp. 311–327



Romer and Siegesmund, 2003
R.L. Romer, S. Siegesmund

Why allanite may swindle about its true age


Contrib. Mineral. Petrol., 146 (2003), pp. 297–307



Romer et al., 1996
R.L. Romer, U. Schärer, A. Steck

Alpine and pre-Alpine magmatism in the root zone of the western Central Alps


Contrib. Mineral. Petrol., 123 (1996), pp. 138–158



Rubatto et al., 2001
D. Rubatto, I.S. Williams, I.S. Buick

Zircon and monazite response to prograde metamorphism in the Reynolds Range, central Australia


Contrib. Mineral. Petrol., 140 (2001), pp. 458–468



Rubatto et al., 2008
D. Rubatto, O. Muntener, A. Barnhoorn, C. Gregory

Dissolution-reprecipitation of zircon at low-temperature, high-pressure conditions (Lanzo Massif, Italy)


American Mineralogist, 93 (2008), pp. 1519–1529



Rubatto et al., 2009
D. Rubatto, J. Hermann, A. Berger, M. Engi

Protracted fluid-assisted melting during Barrovian metamorphism in the Central Alps


Contrib. Mineral. Petrol., 158 (2009), pp. 703–722



Schaltegger and Gebauer, 1999
U. Schaltegger, D. Gebauer

Pre-Alpine geochronology of the Central, Western and Southern Alps


Schweiz. Mineral. Petrogr. Mitt., 79 (1999), pp. 79–87



Schärer et al., 1996
U. Schärer, M. Cosca, A. Steck, J. Hunziker

Termination of major ductile strike-slip shear and differential cooling along the Insubric line (Central Alps): U–Pb, Rb–Sr and 40Ar/39Ar ages of cross-cutting pegmatites


Earth Planet. Sci. Lett., 142 (1996), pp. 331–351



Schmid et al., 1989
S.M. Schmid, H.R. Aebli, F. Heller, A. Zingg

The role of the Periadriatic line in the tectonic evolution of the Alps


D. Dietrich, M.P. Coward (Eds.), Alpine Tectonics, 45Geol Soc London Spec Pub (1989), pp. 153–171



Schmid et al., 1996
S.M. Schmid, A. Pfiffner, N. Froitzheim, G. Schönborn, N. Kissling

Geophysical-geological transect and tectonic evolution of the Swiss-Italian Alps


Tectonic, 15 (1996), pp. 1036–1064



Smith and Barreiro, 1990
H.A. Smith, B. Barreiro

Monazite U–Pb dating of staurolite grade metamorphism in pelitic schists


Contrib. Mineral. Petrol., 105 (1990), pp. 602–615



Spandler et al., 2003
C. Spandler, J. Hermann, R. Arculus, J. Mavrogenes

Redistribution of trace elements during prograde metamorphism from lawsonite blueschist to eclogite facies; implications for deep subduction-zone processes


Contrib. Mineral. Petrol., 146 (2003), pp. 205–222



Stacey and Kramers, 1975
J.S. Stacey, J.D. Kramers

Approximation of terrestrial lead evolution by a two-stage model


Earth Planet. Sci. Lett., 26 (1975), pp. 207–221



Stern and Berman, 2000
R.A. Stern, R.G. Berman

Monazite U–Pb and Th–Pb geochronology by ion microprobe, with an application to in situ dating of an Archean metasedimentary rock


Chem. Geol., 172 (2000), pp. 113–130



Todd and Engi, 1997
C.S. Todd, M. Engi

Metamorphic field gradients in the Central Alps


J. Metamorph. Geol., 15 (1997), pp. 513–530



Tribuzio et al., 1996
R. Tribuzio, B. Messiga, R. Vannucci, P. Bottazzi

Rare earth element redistribution during high-pressure-low-temperature metamorphism in ophiolitic Fe-gabbros (Liguria, northwestern Italy): Implications for light REE mobility in subduction zones


Geology, 24 (1996), pp. 711–714



Vance and O’Nions, 1992
D. Vance, R.K. O’Nions

Prograde and retrograde thermal histories from the central Swiss Alps


Earth Planet. Sci. Lett., 114 (1992), pp. 113–129



Vavra et al., 1996
G. Vavra, D. Gebauer, R. Schmidt, W. Compston

Multiple zircon growth and recrystallization during polyphase Late Carboniferous to Triassic metamorphism in granulites of the Ivrea Zone (Southern Alps): an ion microprobe (SHRIMP) study


Contrib. Mineral. Petrol., 122 (1996), pp. 337–358



von Blanckenburg et al., 1992
F. von Blanckenburg, G. Fruh-Green, K. Diethelm, P. Stille

Nd-, Sr-, O-isotopic and chemical evidence for a two-stage contamination history of mantle magma in the Central-Alpine Bergell intrusion


Contrib. Mineral. Petrol., 110 (1992), pp. 33–45



Williams, 1998
Williams I. S. (1998) U–Th–Pb geochronology by ion microprobe. In: Applications of microanalytical techniques to understanding mineralizing processes. (eds. M.A. McKibben, W.C. Shanks III, W.I. Ridley). Reviews in Economic Geology 7, pp 1–35.


Wing et al., 2003
B.A. Wing, J.M. Ferry, T.M. Harrison

Prograde destruction and formation of monazite and allanite during contact and regional metamorphism of pelites: petrology and geochronology


Contrib. Mineral. Petrol., 145 (2003), pp. 228–250




Corresponding author. Present address: Depatment of Applied Geology, Curtin University of Technology, GPO Box U1987, Perth 6845, WA, Australia.
Copyright © 2012 Elsevier Ltd. All rights reserved.






Bibliographic information




Citing and related articles




Related articles
U-Th-Pb and 230Th/238U disequilibrium isotope systematics: Precise accessory mineral chronology and melt evolution tracing in the Alpine Bergell intrusion
2004, Geochimica et Cosmochimica Acta

Show more information








.


Combined high-precision chronometry and geochemical tracing using accessory minerals: applied to the Central-Alpine Bergell intrusion (central Europe)
1992, Chemical Geology

Show more information








.


Prograde and retrograde thermal histories from the central Swiss Alps
1992, Earth and Planetary Science Letters

Show more information








.


View more articles »

Cited by in Scopus (2)


Related reference work articles

No articles found.





Applications and tools



























Share























Workspace