Introduction

Geochronological studies within Dronning Maud Land (DML), Antarctica, have revealed that the crust formed at c. 1.1–1.0 Ga during the Mesoproterozoic, and was metamorphosed and intruded by voluminous magmatic bodies in the late Neoproterozoic-early Palaeozoic (Ohta et al. 1990; Moyes 1993; Mikhalsky et al. 1997; Jacobs et al. 1998; Paulsson and Austrheim 2003). The late Neoproterozoic-early Palaeozoic activity was related to the Pan–African orogeny and assembly of Gondwana forming the East–African–Antarctic orogen (e.g. Stern 1994; Meert 2003; Jacobs and Thomas 2004). The East–African–Antarctic orogen is up to 1,000 km wide and extends for more than 8,000 km along the eastern margin of Africa and into East Antarctica, and resulted from multiplate collision of various parts of East– and West–Gondwana.

Pan–African granulite-facies metamorphism is frequently reported from areas that were involved in the Gondwana assembly (e.g. Harley 2003). Within central DML, granulite facies metamorphism, dated at 625–515 Ma (Mikhalsky et al. 1997; Jacobs et al. 1998; Henjes-Kunst 2004), reached 800–900 °C (Bucher-Nurminen and Ohta 1993; Bisnath and Frimmel 2005; Piazolo and Markl 1999; Colombo and Talarico 2004; Elvevold and Engvik 2012). Structural studies in combination with geochronology and petrology have revealed an early compressional event followed by a later extensional phase (Jacobs et al. 2003a; Bauer et al. 2004). The later phase of the Pan–African event in central DML is characterised by a decompressional P–T history and extensional structures which indicate tectonic exhumation (Jacobs et al. 2003b; Engvik and Elvevold 2004). Based on the tectonometamorphic evolution of the Pan–African orogeny of Filchnerfjella (Fig. 1; Engvik and Elvevold 2004), 40Ar/39Ar geochronology is used to constrain the timing of the cooling history during unroofing. The results are discussed in the broader context of the early Palaeozoic evolution of East Antarctica. Application of 40Ar/39Ar-geochronology to hornblende, biotite and K-feldspar reveals a cooling history with ages representing temperatures from 550 to 650 °C down to c. 200–250 °C (Harrison 1981; Harrison and McDougall 1982; Villa and Puxeddu 1994; Villa 1998; McDougall and Harrison 1999; Allaz et al. 2011). The data are further used to deduce cooling rates for this part of the orogen.

Fig. 1
figure 1

Geological map of eastern Mühlig–Hofmann– and Filchnerfjella, central Dronning Maud Land, Antarctica. Sample locations are shown. HM = Heimefrontfjella, WM = Wohlthatmassiv, LH = Lützow Holmbukta, PB = Prydz Bay

Full size image

Geological setting and tectonometamorphic evolution

Central DML consists of a series of granitoid intrusive rocks, which are emplaced in granulite– and upper amphibolite facies metamorphic country rocks (e.g. Bucher-Nurminen and Ohta 1993; Piazolo and Markl 1999; Engvik and Elvevold 2004; Roland 2004; Bucher and Frost 2005; D’Souza et al. 2006). The igneous suite extends from 6 to 13°E and comprises charnockite, syenite, quartz syenite, granite and several generations of dykes ranging in composition from granitic to gabbroic/dioritic.

The nunataks of Mühlig–Hofmannfjella and Filchnerfjella (6–8°E) consist of banded gneiss, migmatites, and younger post-kinematic quartz syenite (Figs. 1 and 2). The metamorphic sequence includes metapelites, and leucocratic, intermediate and mafic igneous rocks (Elvevold and Engvik 2012). The different rock types form layers, which vary in thickness from <1 m up to several tens of meters. The granulite facies metamorphism in central DML is dated by 207Pb/206Pb and U-Pb on zircons to between 625 and 515 Ma (Mikhalsky et al. 1997; Jacobs et al. 1998; Henjes-Kunst 2004). The quartz syenite is a part of the igneous suite intruding the banded gneisses. The syenites of central DML are dated by U-Pb on zircons to between 540 and 510 Ma (Mikhalsky et al. 1997; Paulsson 2003; Markl and Henjes-Kunst 2004).

Fig. 2
figure 2

Oblique aerial photograph of Filchnerfjella showing the banded gneisses exposed in Klevekampen and Kubusfjellet, and the intrusive quartz syenite in Trollslottet. (Photo: Norwegian Polar Institute)

Full size image

The banded gneisses (Figs. 2 and 3a) typically comprise granulite facies mineralogy (Elvevold and Engvik 2012). Metapelitic rocks occur as garnet–biotite– and sillimanite–bearing gneisses and migmatites. Mafic and intermediate rocks contain orthopyroxene, amphibole, biotite, garnet and clinopyroxene in variable amounts. The banded gneisses experienced peak metamorphic conditions of about 800–900 °C at intermediate pressures of 0.5–1 GPa and migmatisation has affected large parts of the metamorphic sequence. Breakdown of Grt-Sill-Spl-bearing assemblages to Crd-bearing assemblages indicate that peak conditions were followed by near-isothermal decompression. The dominant E–W fabric illustrates transposition of migmatitic and leucocratic melts during the near-isothermal decompression. Extensional shear bands and shear zones evolved from the ductile partial melting stage through semiductile towards brittle conditions. Their occurrence shows that exhumation persisted towards brittle crustal conditions during tectonic W/SW-vergent extension (Engvik and Elvevold 2004). The quartz syenite (Figs. 2 and 3b) post-dates formation of the main fabric in the metamorphic sequence. The quartz syenite is mostly coarse-grained and contains megacrysts of mesoperthitic K-feldspar, with minor quartz, plagioclase and mafic minerals. The intrusion contains narrow semiductile to brittle shear zones, which suggest that the tectonic exhumation continued after their emplacement. A regional late-magmatic fluid infiltration event locally occurred along pegmatitic veins and is the last record of the Pan–African event affecting the area (Fig. 3c; Engvik et al. 2005; Engvik and Stöckhert 2007).

Fig. 3
figure 3

Field pictures of a) Granulite facies migmatitic gneiss, Klevekampen, Filchnerfjella. b) Quartz syenite, Mühlig–Hofmannfjella. c) Fluid infiltration zones constituted by alteration halos along pegmatitic veins, Trollslottet. The alteration causes bleaching of the dark rock due to hydration mineral reactions and a high density of microcracks in feldspar and quartz

Full size image

40Ar/39Ar geochronology

40Ar/39Ar analyses of mineral separates from selected samples were carried out at the Geological Survey of Norway. Mineral separates of K-feldspar, biotite and hornblende were analyzed by step-heating in a resistance furnace (AN29 and AHA182) or with a CO2 laser (AHA197 and AHA200). Detailed analytical procedures are presented in the Electronic supplement. An overview of the results is presented in Table 1 and complete data tables are reported in the Electronic supplement. The 40Ar/36Ar intercept ratios calculated from the inverse isochrons (Table 1) overlap the atmospheric value of 298.56 ± 0.31 (Lee et al. 2006) within 2σ in all cases, except for AHA200 Hornblende (for which the inverse isochron in any case has an unacceptably high MSWD). In our discussion of the age results we rely on the plateau ages.

Table 1 Summary table of 40Ar/39Ar-data. ‘Steps used’ refers to the step numbers used in the age calculation. %39Ar = the percentage of cumulative 39Ar gas represented by the indicated steps. 40Ar/36Ar is the intercept on the vertical of the inverse isochron diagram. MSWD, mean square weighted deviation. Prob., probability of fit e.g. Chidist(MSWD*df,df). T/OP, Temperature (oven samples) or Operating Power (CO2 laser). TFA = total fusion age. All uncertainties reported at 2σ
Full size table

Four samples were selected for analyses, each sample contains hornblende, biotite, and K-feldspar in the parageneses. Mineral chemical data for the dated phases are presented in Table 2. The quantitative microanalyses were performed using a Cameca SX100 electron microprobe equipped with 5 wave length-dispersive spectrometers (WDS) at the Institute of Geosciences, University of Oslo. The accelerating voltage was 15 kV and the counting time 10 s on peak using a beam current of 15 nA. A defocused beam (10 m) was used for K-feldspar. Standardization was made on a selection of synthetic and natural minerals and oxides. Data reduction was done by the PAP program (Pouchou and Pichoir 1984). Sample locations are shown in Fig. 1.

Table 2 Mineral chemical data of amphiboles, biotites and K-feldspar
Full size table

Sample descriptions

Samples AN29 and AHA182 were collected from migmatite at Klevekampen and Kubusfjellet, respectively. The orthopyroxene-bearing migmatite is heterogranular, fine– to medium–grained. Its characteristic dark brown colour is typical for the granulite facies rocks in the area. Anhedral quartz, plagioclase and K-feldspar grains, 0.5–2 mm in diameter, are predominant. K-feldspars (Or85-91Ab9-15) are present as microperthite. Hornblende (Mg/(Mg + Fe) = 0.19–0.23; K = 0.33–0.36 p.f.u.; Ca = 1.86–1.87; Cl = 0.03–0.04) and biotite (Mg/(Mg + Fe) = 0.26–0.28; Cl = 0.03–0.04) are concentrated in thin layers together with orthopyroxene (Fig. 4a). Amphibole shows no zoning except for a small rimward lowering in Na up to 0.08 p.f.u. Ilmenite, apatite, monazite and zircon are present as accessory minerals.

Fig. 4
figure 4

Photomicrographs (abbreviations after Whitney and Evans 2010). a) Crystals of biotite and hornblende in Opx–bearing migmatitic gneiss. Sample AN29, width of image 5.3 mm. b) Hornblende and biotite crystals in quartz syenites. Part of microperthite lath is showed in lower right corner. Sample AHA197, width of image 5.3 mm

Full size image

Samples AHA197 and AHA200 were collected at Trollslottet. The quartz syenite, which makes up Trollslottet nunatak, is dominated by microperthitic K-feldspar (Or83-84Ab16-17), quartz and plagioclase. The microtexture is strongly heterogranular, with microperthite as subhedral laths up to 5 cm, set in a medium grained matrix of quartz, plagioclase and mafic minerals. Myrmekite is common. 5–10 % mafic minerals are present, mostly as hornblende (Mg/(Mg + Fe) = 0.33; K = 0.28 p.f.u.; Ca = 1.90; Cl = 0.03) and biotite (Mg/(Mg + Fe) = 0.34; Cl = 0.04) (Fig. 4b). Orthopyroxene is replaced by fine-grained iddingsite, or often altered to grunerite. Apatite, ilmenite, zircon, allanite and hematite are accessory minerals. Minor and local alteration of microperthite is observed along micro fractures. Sample AHA200, which is sampled 1.5 m away from sample AHA197, is strongly altered due to late-magmatic fluid infiltration. K-feldspar of the altered quartz syenite is Or89Ab12 and shows replacement to microcline. AHA200 also is the only sample where sericite and chlorite are observed in thin section in feldspar and biotite respectively. Orthopyroxene and grunerite are absent, while biotite and hornblende have recrystallized into fine grained aggregates. Biotite is lower in Mg/(Mg + Fe), with a ratio of 0.29, and higher in Cl (0.13–0.14 p.f.u.) compared to biotite in the brown quartz syenite. The hornblende has Mg/(Mg + Fe) = 0.21–0.22, and a small rimward increase of Ca to 1.95 p.f.u. and K to 0.37 p.f.u., and a Cl-content of up to 0.17 p.f.u.

Hornblende age data

Age spectra and inverse isochron plots for hornblende separates are displayed in Fig. 5. Hornblende separated from sample AN29 yielded a statistically valid plateau age (481.14 ± 5.62 Ma at 2σ; probability-of-fit 0.84) for 85.07 % of the cumulative 39Ar. The three other hornblende spectra indicate perturbation of the K/Ar system. Not surprisingly this is most apparent for the strongly altered sample, AHA200. The Cl/K ratio for this sample is higher than for the other hornblende separates, as confirmed by mineral chemical data (Table 2; Cl/K = 0.4 for both rim and core). No systematic variation of step ages with Ca/K or Cl/K ratio is apparent however for hornblende samples AHA182, AHA197 or AHA200. Inverse isochron ages for AHA197 hornblende (475.93 ± 9.78 Ma at 2σ) and AHA182 Hornblende (495.84 ± 10.43 Ma at 2σ) have probability-of-fit values at the minimally acceptable value (0.05). However, this is for portions of the spectra (44.25 % and 49.73 % of the cumulative 39Ar, respectively) that are too small for them to be considered robust ages. Despite their more complex spectra, the weighted mean ages for parts of the hornblende spectra for samples AHA182 and AHA197 indicate they likely recorded a similar event as AN29 hornblende (Fig. 5).

Fig. 5
figure 5

40Ar/39Ar release spectra (top panels) and inverse isochron diagrams (bottom panels) for hornblende separates. Grey boxes in the spectra represent steps included in the spectrum age calculation. Red squares in the inverse isochron diagrams represent steps included in the isochron calculation. WMA, weighted mean age. a) AN29. b) AHA182. c) AHA197. d) AHA200. All uncertainties and error envelopes at 2σ

Full size image

Biotite age data

Age spectra and inverse isochron plots for biotite separates are plotted in Fig. 6. All four samples analyzed yielded stable spectra, with statistically valid plateaus and inverse isochrons with 40Ar/36Ar ratio overlapping the atmospheric value for each sample. The biotite plateau ages for samples AN29 and AHA182 are virtually identical at 464.48 ± 3.76 Ma (2σ; probability-of-fit 0.34) and 465.80 ± 3.66 Ma (2σ; probability-of-fit 0.95) respectively. For sample AHA197 the plateau age is younger by about 14 Ma at 451.75 ± 4.69 Ma (2σ; probability-of-fit 0.12), with sample AHA200 yielding the youngest age of 434.92 ± 3.88 Ma (2σ; probability-of-fit 0.59). As for AHA200 Hornblende, chemical data (Table 2) indicate a higher Cl/K ratio for AHA200 biotite compared to the other three biotite separates, again confirming the strongly altered nature of this sample.

Fig. 6
figure 6

40Ar/39Ar release spectra (top panels) and inverse isochron diagrams (bottom panels) for biotite separates. Grey boxes in the spectra represent steps included in the spectrum age calculation. Red squares in the inverse isochron diagrams represent steps included in the isochron calculation. a) AN29. b) AHA182. c) AHA197. d) AHA200. All uncertainties and error envelopes at 2σ

Full size image

K-feldspar age data

Age spectra and inverse isochron plots for K-feldspar samples are plotted in Fig. 7. For sample AHA182 a statistically valid plateau (436.90 ± 3.39 Ma at 2σ; probability of fit 0.64) was obtained for 68.35 % of the cumulative 39Ar. K-feldspar spectra for samples AHA197 and AHA200 indicate perturbed systems for which no statistically valid plateau or isochron ages can be calculated. Nevertheless, weighted mean ages for large parts of these two spectra overlap with the K-feldspar age for sample AHA182. Samples AHA182 and AN29 had nearly identical biotite ages, but as for AHA197 and AHA200 the AN29 K-feldspar the spectrum is disturbed. Figure 7a displays the age spectrum for AN29 K-feldspar together with the Ca/K ratio and the percentage of radiogenic 40Ar. In the second part of the spectrum, from ~40 % cumulative 39Ar and onwards, the percentage of radiogenic 40Ar correlates positively with the step ages and negatively with the Ca/K content. This suggests that the disturbance of the AN29 K-feldspar spectrum may partly be compositionally controlled. The high temperature steps in the AN29 K-feldspar spectrum overlap in age with the statistically robust age for AHA182 K-feldspar.

Fig. 7
figure 7

40Ar/39Ar release spectra (top panels) and inverse isochron diagrams (bottom panels) for K-feldspar separates. Grey boxes in the spectra represent steps included in the spectrum age calculation. Red squares in the inverse isochron diagrams represent steps included in the isochron calculation. WMA, weighted mean age. MSWD, mean square weighted deviation. a) AN29. b) AHA182. c) AHA197. d) AHA200. All uncertainties and error envelopes at 2σ

Full size image

The statistically valid plateau and isochron ages for AHA182 K-feldspar provide a well-constrained timing for closure of the K/Ar system in K-feldspar. Despite the noted disturbances, the K-feldspar spectra for two other samples (AHA197, AHA200) apparently recorded the same cooling event.

Discussion

Cooling history

Robust 40Ar/39Ar plateau ages on hornblende (481.14 ± 5.62 at 2σ), biotite (465.80 ± 3.66 and 464.48 ± 3.76 at 2σ) and K-feldspar (436.90 ± 3.39 at 2σ) are available for migmatite from Filchnerfjella. For the fresh quartz syenite (AHA197), a robust biotite plateau age (451.75 ± 4.69 at 2σ) and a disturbed K-feldspar spectrum that nevertheless is compatible with the K-feldspar age of the migmatite was obtained. The strongly altered quartz syenite has an even younger biotite age of 434.92 ± 3.88 Ma (2σ), which overlaps with the robust K-feldspar age of the migmatite. The younger biotite ages of the quartz syenite can be explained by a slower/later cooling of the large intrusion than the host gneisses, while the crustal temperature distribution has equilibrated by the time reflected by K-feldspar ages. The altered quartz syenite sample (AHA200) was taken with only 1.5 m separation from the dark, fresh quartz syenite (sample AHA197). The relatively poor constraints on the hornblende age of the altered quartz syenite might be interpreted as an effect of the late-magmatic fluid infiltration, explaining the relatively high Cl/K ratio (Table 2) of this sample. The fluid infiltration caused local heating and partial replacement of the mineralogy of the host rock (Engvik et al. 2009). A similar effect of fluid infiltration on the age distribution is shown by Markl and Henjes-Kunst (2004).

The Pan–African granulite facies metamorphism in central DML is dated between 625 and 515 Ma, while the syenites and charnockites intruded between 540 and 510 Ma (Mikhalsky et al. 1997; Jacobs et al. 1998; Markl and Henjes-Kunst 2004; Henjes-Kunst 2004). The gneisses in Filchnerfjella experienced peak metamorphic conditions of 800–900 °C at intermediate pressures followed by near-isothermal decompression (Engvik and Elvevold 2004). This near-isothermal decompression was accompanied by extension and magmatic intrusions, followed by cooling. The magmatic activity was associated with a slightly raised (30–40 °C/km) geothermal gradient. The late– and post–Pan–African cooling ages reflects crustal levels about 0.2–0.4 GPa (Engvik et al. 2005). By that time, the crust had relaxed to a stable continental geotherm of ~20 °C/km.

The 40Ar/39Ar-ages demonstrate that the cooling took place during the later stage of the Pan–African orogeny. The cooling history (Fig. 8) is based on estimates of ‘closure-temperatures’ from literature data. For hornblende, estimates are in the range of 500–550 °C (Harrison 1981) to 550–650 °C (Villa 1998), based on laboratory experiments and geological data, respectively. For biotite a commonly cited figure based on hydrothermal laboratory experiments is 300–350 °C (Grove and Harrison 1996). Validation of similar laboratory determinations against fluid inclusion thermometry on natural samples however suggests the laboratory based determinations could be too low by as much as ~100 °C (Villa and Puxeddu 1994), implying the geologically relevant biotite closure temperature is more likely 400–450 °C. A range of 200–250 °C has been reported for the closure temperature of K-feldspar (Harrison and McDougall 1982).

Fig. 8
figure 8

Cooling histories based on literature values for the closure temperature of hornblende (Villa 1998), biotite (Villa and Puxeddu 1994) and K-feldspar (Harrison and McDougall 1982). Robust plateau ages (solid dots) constrain the cooling history of Filchnerfjella migmatite. In the absence of robust K-feldspar ages for fresh and altered syenite at Trollslottet the timing of K-feldspar closure is based on more qualitative information from perturbed spectra (see text for discussion)

Full size image

Post-dating intrusion of the syenites, our best constraint on the timing of cooling below 550–650 °C (hornblende) for the migmatite is 481.14 ± 5.62 Ma (2σ). Subsequent cooling down to 400–450 °C is well-constrained by biotite ages of 464.48 ± 3.76 Ma (2σ) and 465.80 ± 3.66 Ma (2σ), indicating a cooling rate of ~10 °C/Myr. Cooling continued subsequently at rates of ~6 °C/Myr, with K-feldspar ages indicating temperatures of about 200–250 °C at 436.90 ± 3.39 Ma (2σ). Cooling rates based on closure temperatures derived from laboratory data (Harrison 1981; Grove and Harrison 1996) would differ somewhat, being ~12 °C/Myr (hornblende to biotite) and ~4 ° C/Myr (biotite to K-feldspar), respectively.

Regional comparison within East Antarctica

The 40Ar/39Ar hornblende– and biotite ages obtained in this study are similar to 40Ar/39Ar ages reported by Markl and Henjes-Kunst (2004) from the Eastern Mühlig–Hofmannfjella (7°E, 71°40′S, central DML). The 40Ar/39Ar amphibole– and biotite–ages from charnockite are 487.0 ± 2.4 and 466.6 ± 2.2 Ma, respectively. In the Schirmacheroase (11°30′E, 70°45′S, central DML), Henjes-Kunst (2004) achieved older amphibole ages of 590–575 Ma and biotite ages of 525–520 Ma from metamorphosed sills. From Heimefrontfjella (western DML), Jacobs et al. (1995) retrieved two tectonothermal events. The first event showed ages below 1,000 Ma, while a later Pan–African event was recorded in K/Ar– and 40Ar/39Ar–ages of biotite and muscovite to 533–476 Ma.

Late Neoproterozoic/early Palaeozoic high-grade metamorphism followed by exhumation and tectonic extension is reported over a wide area of East Antarctica. The Lützow–Holm Bay (e.g. Fraser et al. 2000) and Prydz Bay areas (e.g. Thost et al. 1994) have experienced an evolution very similar to the one revealed in DML. Fraser et al. (2000) achieved 40Ar/39Ar biotite and hornblende ages between 510 and 460 Ma in the Lützow–Holm Bay, following after Pan–African granulite facies metamorphism and near-isothermal decompression. Zhao et al. (1997) reported 40Ar/39Ar hornblende age of 514 ± 2 Ma, biotite ages between 500 and 490 Ma and K-feldspar ages around 455 Ma in the Prydz Bay, and reported cooling rates of 9–3 °C/Myr. Although the 40Ar/39Ar –geochronology ages of the Filchnerfjella show slightly younger ages, the data reported throughout the wider area of East Antarctica reflects the late Pan-African cooling history and show that the East Antarctica occurred as a large homogenous crustal fragment in the early Palaezoic. The younger ages of Filchnerfjella can be explained by a slightly later cooling after the intrusions of the voluminous magmatic bodies in the area.

After the late Pan-African cooling, East Antarctica underwent a longer time of relaxation before Gondwana breakup. The Gondwana breakup is reflected in apatite-fission-track data showing ages between 250 and 100 Ma throughout East Antarctica from Heimefrontfjella in the west to Prydz Bay in the east (Jacobs et al. 1995; Zhao et al. 1997; Meier et al. 2004).

Conclusions

40Ar/39Ar geochronological data combined with petrological evidence, constrain the timing of late-orogenic tectonometamorphic evolution of the high-grade metamorphic complex from the Mühlig–Hofmann– and Filchnerfjella in central DML. 40Ar/39Ar hornblende–, biotite– and K-feldspar cooling ages range from c. 481 to 437 Ma, and reflect late–Pan–African cooling during extension and exhumation. The ages indicate a cooling rate of ~ 10 °C/Myr between 481 and 465 Ma, followed by nearly 30 million years of cooling at ~ 6 °C/Myr. 40Ar/39Ar geochronological data reflecting the late Pan–African cooling history, reported throughout East Antarctica, are consistent and indicate that East Antarctica existed as a large homogenous crustal fragment in the early Palaezoic.