1. Introduction
Plausible geodynamic models of polymetamorphosed terranes require detailed knowledge of the magmatic and metamorphic processes that characterize the studied rock association. Crucial geochronological aspects are often poorly constrained and can only be revealed by application of a large number of different chronometric methods that specifically target different segments of the geological evolution. In spite of such efforts, the general picture may remain patchy, due to difficulties in linking age information to a distinct P–T stage, or because the last metamorphic overprint obscured traces of earlier events. This situation is encountered in the Orlica-Śnieżnik complex at the NE margin of the Bohemian Massif (Fig. 1a–c), where eclogites and granulites occur as isolated lenses within orthogneisses. Both rock types have attracted considerable attention due to (1) findings of presumed pseudomorphs after coesite and petrological considerations suggesting ultrahigh-pressure (UHP) metamorphic conditions (Bakun-Czubarow, Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb, Reference Bakun-Czubarow1992; Bröcker & Klemd, Reference Bröcker and Klemd1996; Kryza, Pin & Vielzeuf, Reference Kryza, Pin and Vielzeuf1996; Klemd & Bröcker, Reference Klemd and Bröcker1999), and (2) favourable mineral assemblages that allow application of various geochronological methods (e.g. Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a; Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007; Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009). Understanding of the petrological and geochronological record of these rocks plays a key role in unravelling the geodynamic evolution of the eastern Variscides. Available concepts strongly depend on Sm–Nd, Ar–Ar, Rb–Sr and Lu–Hf dating. These methods allow easier linking of age information with specific P–T–D stages than does U–Pb zircon geochronology. On the other hand, U–Pb zircon dating is a more promising tool to unravel protolith ages and to reveal age information related to earlier metamorphic stages that is not disclosed by other dating methods with higher susceptibility to overprinting. Although some U–Pb zircon data are available for the granulites (Klemd & Bröcker, Reference Klemd and Bröcker1999; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a; Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007), a detailed and systematic U–Pb geochronological and zircon trace element study of the (U)HP rocks has not yet been carried out. The main aim of the present study is to assess the potential of zircon to obtain further insight into the geochronological history of these rocks. Using secondary ion mass spectrometry (SIMS) and laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS), we have applied U–Pb zircon dating, trace element analysis and Ti-in-zircon thermometry to samples collected from some of the best preserved (U)HP occurrences of the Orlica-Śnieżnik complex.
Figure 1. (a) Location of the Orlica-Śnieżnik complex within the Variscan belt. RH – Rhenohercynian Zone; NP – Northern Phyllite Zone; MCH – Mid-German Crystalline High; SX – Saxothuringian Zone; MO – Moldanubian Zone; OFZ – Odra Fault Zone; ISF – Intra-Sudetic Fault Zone; EFZ – Elbe Fault Zone; MS – Moravo-Silesian Zone. (b) Simplified geological map of the Orlica-Śnieżnik complex and neighbouring units (modified after Don et al. Reference Don, Dumicz, Wojciechowska and Żelaźniewicz1990; Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000). (c) Location of the study area in the eastern part of the Orlica-Śnieżnik complex (modified after Perchuk et al. Reference Perchuk, Korchagina, Yapaskurt and Bakun-Czubarow2005).
2. Geological setting
The Orlica-Śnieżnik complex represents one of the major lithostratigraphic units of the West Sudetes and mainly consists of high-grade amphibolite-facies orthogneisses and schists (e.g. Don et al. Reference Don, Dumicz, Wojciechowska and Żelaźniewicz1990; Żelaźniewicz, Mazur & Szczepański, Reference Żelaźniewicz, Mazur and Szczepański2002). The orthogneisses have Cambrian protolith ages (c. 520–490 Ma; e.g. Oliver, Corfu & Krogh, Reference Oliver, Corfu and Krogh1993; Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000; Kröner et al. Reference Kröner, Jaeckel, Hegner and Opletal2001; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004) and record at least one episode of anatectic high-temperature (HT) metamorphism during Variscan time (e.g. Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000; Lange et al. Reference Lange, Bröcker, Mezger and Don2002, Reference Lange, Bröcker, Armstrong, Trapp and Mezger2005b; Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009), but possibly also migmatization in early Palaeozoic time (c. 515–485 Ma; Kröner et al. Reference Kröner, Jaeckel, Hegner and Opletal2001; Żelaźniewicz et al. Reference Żelaźniewicz, Nowak, Larionov and Presnyakov2006). Comprehensive overviews of the geology of the Orlica-Śnieżnik complex were presented by Don et al. (Reference Don, Dumicz, Wojciechowska and Żelaźniewicz1990), Żelaźniewicz, Mazur & Szczepański (Reference Żelaźniewicz, Mazur and Szczepański2002) and Lange et al. (Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005b). Here only a short summary of the features most relevant for the present study is given.
Eclogites locally occur as isolated blocks and lenses (up to tens of metres in size) within orthogneisses and are most common in a narrow NS-stretching zone (~ 3 km long) between the villages of Międzygórze and Nowa Wieś (e.g. Smulikowski, Reference Smulikowski1967; Bakun-Czubarow, Reference Bakun-Czubarow1968; Smulikowski & Smulikowski, Reference Smulikowski and Smulikowski1985; Bröcker & Klemd, Reference Bröcker and Klemd1996; Don, Reference Don2001). Granulites are restricted to a NE-trending zone (up to 2 km wide and ~ 12 km long) in the eastern part of the study area (Fig. 1c; Pouba, Paděra & Fiala, Reference Pouba, Paděra and Fiala1985; Bakun-Czubarow, Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb, Reference Bakun-Czubarow1992). Well-preserved outcrops are very rare. Several studies suggested that both eclogites and granulites have been tectonically incorporated into the orthogneisses (e.g. Don et al. Reference Don, Dumicz, Wojciechowska and Żelaźniewicz1990; Don, Reference Don2001). The most recent variants of such concepts propose that both rock types were transported to middle crustal levels by ductile vertical extrusion of a low viscosity matrix along relatively steep channels in front of a rigid backstop (e.g. Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Gordon et al. Reference Gordon, Schneider, Manecki and Holm2005; Schneider et al. Reference Schneider, Zahniser, Glascock, Gordon and Manecki2006; Pressler et al. Reference Pressler, Schneider, Petronis, Holm and Geissman2007). Other studies argued in favour of in situ (U)HP metamorphism (Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Bröcker & Klemd, Reference Bröcker and Klemd1996).
P–T estimates for eclogites suggest peak-pressures > 27 kbar at temperatures of 700–800°C (e.g. Bakun-Czubarow, Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb; Bröcker & Klemd, Reference Bröcker and Klemd1996). Significant differences in metamorphic conditions between discrete eclogite occurrences were not recognized. P–T estimates for granulites vary between ~ 18 kbar and 900°C (Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004) and 21–28 kbar at temperatures of 800–1000°C (e.g. Bakun-Czubarow, Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb, Reference Bakun-Czubarow1992; Kryza, Pin & Vielzeuf, Reference Kryza, Pin and Vielzeuf1996; Klemd & Bröcker, Reference Klemd and Bröcker1999). For the amphibolite-facies overprint, Klemd, Bröcker & Schramm (Reference Klemd, Bröcker and Schramm1995) reported pressure conditions of 4–11 kbar and temperatures of ~ 600–650°C. Microthermometric measurements on fluid inclusions suggest almost isothermal uplift from 7.5 kbar to about 2 kbar at 600°C, followed by isobaric cooling to about 200°C (Klemd, Bröcker & Schramm, Reference Klemd, Bröcker and Schramm1995).
Protolith ages of the (U)HP rocks are only poorly constrained. A sensitive high-resolution ion microprobe (SHRIMP) U–Pb zircon age of c. 525 Ma for an eclogite was repeatedly quoted in the regional literature (e.g. Bakun-Czubarow, Reference Bakun-Czubarow1998, referring to D. Gebauer, unpub. data). Bakun-Czubarow (Reference Bakun-Czubarow1991b) considered the granulites as derivatives of a Cadomian volcanic rock suite. For a felsic granulite, Štípská, Schulmann & Kröner (Reference Štípská, Schulmann and Kröner2004) reported a 206Pb/238U SHRIMP zircon age of 473 ± 8 Ma (1σ) for a single data point, interpreted to date an unspecified event in the history of the protolith. Based on U–Pb SHRIMP zircon dating of a mafic granulite, Lange et al. (Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a) suggested a minimum protolith age of c. 460 Ma.
Eclogites and granulites mostly yielded Variscan metamorphic ages (Sm–Nd, U–Pb, Ar–Ar, Rb–Sr) that range between 350 and 325 Ma (e.g. Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Steltenpohl et al. Reference Steltenpohl, Cymerman, Krogh and Kunk1993; Klemd & Bröcker, Reference Klemd and Bröcker1999; Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000; Marheine et al. Reference Marheine, Kachlik, Maluski, Patocka, Żelaźniewicz, Winchester, Pharaoh and Verniers2002; Lange et al. Reference Lange, Bröcker, Mezger and Don2002, Reference Lange, Bröcker, Armstrong, Trapp and Mezger2005a,b; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009). Significant age differences between eclogites and granulites, and high temperature rocks recording amphibolite-facies overprinting, could not yet convincingly be resolved (e.g. Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a and references therein). A prominent age cluster at c. 340 Ma has mostly been interpreted as a close approximation to peak (U)HP conditions (e.g. Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a). In the case of the granulites, Anczkiewicz et al. (Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007) concluded that the 340 Ma ages record amphibolite-facies overprinting. Based on a Lu–Hf garnet growth age and published zircon ages, these authors argued that the granulites experienced (U)HP metamorphism at some point between c. 387 and 360 Ma, but the exact timing of this metamorphic episode could not be further determined.
3. Sample description
Six samples were selected for zircon geochronology and seven samples for trace element studies from occurrences in and around Międzygórze, Nowa Wieś, Nowa Morawa and Stary Gierałtów (Fig. 1c). Eclogite samples from Międzygórze were collected from outcrops on the hilltop between the Wilczka and Bogoryja rivulets (5301), and on the western slope of Jawor Mountain (5302). Sample NW-1 represents an eclogite lens that is exposed about 1 km NE of Nowa Wieś (Bakun-Czubarow, Reference Bakun-Czubarow1968). Eclogite sample 1111 was taken from a loose block in the forest on the opposite side of the same valley. Eclogite 5306 is derived from the SE slope of Suszyca Mountain near Nowa Morawa. Sample 1106 was collected at Stary Gierałtów from the single outcrop of granulite on the Polish side of the border. Conventional ID-TIMS and ion-probe data for this sample have already been reported by Klemd & Bröcker (Reference Klemd and Bröcker1999) and Lange et al. (Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a). Sample 5312 represents a loose block collected within the granulite belt, about 1.5 km ENE of Stary Gierałtów. GPS coordinates of the sample locations are reported in Tables 1 and 2, and in online Appendix Table A2 available at http://journals.cambridge.org/geo.
Table 1. SHRIMP U–Pb–Th analytical data for zircons from eclogite (5301, 5302) and granulite (5312)
f 206Pb% indicates the percentage of 206Pb that is common Pb; Rad. 206Pb = radiogenic 206Pb. Uncertainties are reported at the 1σ level.
Common Pb corrected using measured 204Pb. Error in standard calibration was 0.41% (samples 5301, 5312) and 0.59% (sample 5302), respectively.
Sample locations: 5301 = Międzygórze, N 50°13.798´, E 016°46.142′ ; 5302 = Międzygórze, N 50°13.548′, E 016°46.097′; 5312 = Stary Gierałtów, N 50°18.545′, E 016°57.472′.
Table 2. LA-ICPMS U-Pb data for (U)HP rocks of the Orlica-Śnieżnik complex
All uncertainties are absolute values (2σ); – indicates 204Pb below detection limit.
The common mineral assemblage of eclogites comprises garnet, clinopyroxene, zoisite, rutile and quartz. Varietal minerals include phengite, calcic amphibole and kyanite. Apatite, zircon and opaque phases are typical accessories. Retrograde phases include clinopyroxene(II), Ca-amphibole(II), biotite, epidote/clinozoisite, chlorite, albite, titanite and calcium carbonate. Sample 1106 is a mafic granulite mainly consisting of garnet, omphacite, plagioclase and quartz. Biotite and kyanite occur in small quantities. Typical accessories are rutile, apatite, zircon and Fe-oxides. The presence of pseudomorphs after coesite has been inferred from radial fractures around polycrystalline quartz inclusions in garnet (Klemd & Bröcker, Reference Klemd and Bröcker1999). Sample 5312 represents a granulite of intermediate bulk rock composition. The mineral assemblage is similar to 1106, but the modal proportions of garnet and omphacite are much lower, and kyanite has not been recognized. Additional phases, which are present in small quantities, are Ca-amphibole, titanite and K-feldspar. For field and petrological details of the eclogites and the granulites, see Smulikowski (Reference Smulikowski1967), Smulikowski & Smulikowski (Reference Smulikowski and Smulikowski1985), Pouba, Paděra & Fiala (Reference Pouba, Paděra and Fiala1985), Bakun-Czubarow (Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb, Reference Bakun-Czubarow1992), Dumicz (Reference Dumicz1993), Kryza, Pin & Vielzeuf (Reference Kryza, Pin and Vielzeuf1996), Bröcker & Klemd (Reference Bröcker and Klemd1996), Klemd & Bröcker (Reference Klemd and Bröcker1999) and Perchuk et al. (Reference Perchuk, Korchagina, Yapaskurt and Bakun-Czubarow2005).
4. Analytical methods
All eclogites and granulites record indications for amphibolite-facies overprinting. Sample selection was guided by the attempt to collect the most pristine parts of individual occurrences. For U–Pb geochronology and trace element studies, zircon was separated from ~ 1.5–12 kg samples by standard routines (jawbreaker, disc mill, Wilfley table, Frantz magnetic separator, heavy liquids). After polishing to expose the grain interior, CL imaging was applied to reveal the internal zircon structures and to guide spot placement. SIMS U–Pb dating was carried out by use of a sensitive high-resolution ion microprobe (SHRIMP) at the Centre of Isotopic Research (VSEGEI), St Petersburg, Russia. Handpicked zircon grains were mounted in 25 mm epoxy discs together with pieces of the Temora-1 (Black et al. Reference Black, Kamo, Allen, Aleinikoff, Davis, Korsch and Foudoulis2003) and 91500 zircon standards (Wiedenbeck et al. Reference Wiedenbeck, Allé, Corfu, Griffin, Meier, Oberli, von Quadt, Roddick and Spiegel1995; U = 81.2 ppm). Including both spike-calibration and U decay constant uncertainties, the total uncertainty of the Temora-1 standard is 416.8 ± 1.3 Ma, and a value of 417 Ma was used for age calculation in this study. Analytical procedures followed standard operating routines and were similar to those reported by Williams (Reference Williams, McKibben, Shanks III and Ridley1998) and Larionov, Andreichev & Gee (Reference Larionov, Andreichev, Gee, and Pease2004). Cleaned zircon mounts were gold-coated and analysed for U–Th–Pb isotopes using a primary beam diameter of ~ 20 μm. Before analysis, the primary beam was rastered over the target area for ~ 30 s. The data were collected in sets of five scans through 196Zr2O, 204Pb, background (~ 204.5), 206Pb, 207Pb, 208Pb, 238U, 248ThO and 254UO. The Temora standard was analysed after every fifth unknown. The data were reduced using the SQUID v. 1.12 Excel Macro of Ludwig (Reference Ludwig2005a). Uncertainties given for individual SHRIMP data points (ratios and ages) are reported at the 1σ level; error ellipses are shown with 2σ uncertainty. In the text, individual ages and weighted averages are quoted as 206Pb–238U ages with 2σ uncertainty. Common Pb corrections were performed using the relevant Pb composition after Stacey & Kramers (Reference Stacey and Kramers1975) and measured 204Pb (Compston, Williams & Meyer, Reference Compston, Williams and Meyer1984). Most analyses contain very little common Pb and thus are insensitive to the choice of initial isotope composition.
For LA-ICPMS geochronological studies, new sample mounts were prepared, because only a relatively small number of grains were available on the SHRIMP epoxy discs. These age determinations and the trace element analyses were performed on a sector field ICP-MS (Element2, ThermoFinnigan) coupled to a 193 nm ArF Excimer laser system (UP193HE, New Wave Research) at the Institut für Mineralogie, Universität Münster. The instrument parameters for both the laser and the ICP-MS are listed in Table A1 (http://journals.cambridge.org/geo). For U–Pb analysis the masses 202, 204, 206, 207 and 238 were measured. 202Hg was also analysed to quantify the interference of 204Hg on 204Pb. Corrections for laser-induced elemental fractionation and instrumental mass bias were done by bracketing groups of five unknowns with two measurements of the GJ-1 standard zircon (Jackson et al. Reference Jackson, Pearson, Griffin and Belousova2004). Age calculations were done offline with an in-house Excel spreadsheet. Corrections were applied for gas blank, time-dependent elemental fractionation (Sylvester & Ghaderi, Reference Sylvester and Ghaderi1997) and common Pb, if present. A common Pb correction (Stacey & Kramers, Reference Stacey and Kramers1975) was only applied if the contribution of the common 206Pb to the total measured 206Pb was 0.7% or higher. Analyses that showed in-run 204Pb anomalies but ‘normal’ average 204Pb values were subjected to a common Pb correction applied to the anomalous ratios only. Uncertainties given for individual LA-ICPMS U–Pb zircon analyses (ratios, ages, error ellipses) are reported with 2σ uncertainty. To monitor reproducibility of the 206Pb–238U ages, the Plešovice standard zircon (337.3 ± 0.4 Ma; Sláma et al. Reference Sláma, Košler, Condon, Crowley, Gerdes, Hanchar, Horstwood, Morris, Nasdala, Norberg, Schaltegger, Schoene, Tubrett and Whitehouse2008) was analysed and processed as an unknown. The results of 30 spot analyses are shown in Figure 2. The plots, regressions and weighted mean age calculations were carried out using Isoplot/Ex 3.22 (Ludwig, Reference Ludwig2005b). Decay constants used for SHRIMP and LA-ICPMS age calculations are those recommended by the Subcommission on Geochronology of IUGS (Steiger & Jäger, Reference Steiger and Jäger1977).
Figure 2. Concordia plot for results of the Plešovice zircon standard with weighted mean average.
For trace element analysis of samples 5301, 5302, 5306 and 5312, available SHRIMP mounts were used, and spot selection was guided by the U–Pb pits induced by ion-probe analysis. For other samples the analyses were performed on the LA-ICPMS zircon mounts within the previously dated CL zone, in most cases directly adjacent to the ablation pits induced by U–Pb dating.
The system has been tuned (torch position, lenses, gas flows) on standard glass NIST 612 measuring 139La, 232Th and 232Th16O to get stable signals and high sensitivity on 139La and 232Th peaks, as well as low oxide rates (232Th16O/232Th ~ 0.1%) during ablation. The NIST 612 glass was used as an external standard (using the preferred values of the GeoReM reference material database, version 11/2006). Groups of five unknowns were bracketed with two calibration standards on both sides to track instrumental drift. The 91500 standard zircon (Wiedenbeck et al. Reference Wiedenbeck, Allé, Corfu, Griffin, Meier, Oberli, von Quadt, Roddick and Spiegel1995) was also measured in all sessions along with the unknowns to monitor accuracy. Results are summarized in the last columns of Table 3. Concentrations of measured elements were calculated using the Glitter software (e.g. Griffin et al. Reference Griffin, Powell, Pearson, O'Reilly and Sylvester2008). Precision is usually better than 10% but depends on the absolute concentration. Elemental abundances were normalized to total Hf concentrations that were determined with a JEOL 8900M Superprobe adjacent to the ablation pits, using natural and synthetic calibration standards. Analytical conditions were 15 kV accelerating potential, a beam current of 15 nA, a spot size of ~ 1 μm, and a counting time of 10 s on peaks and 5 s on background.
Table 3. LA-ICPMS trace element data for zircons from eclogites and granulites
Concentrations in ppm. Eu/Eu* = EuN/√(SmN*GdN), Ce/Ce* = CeN/√(LaN*PrN), normalization to chondrite after Boynton (Reference Boynton and Henderson1984).
Ti-in-zircon temperatures after Ferry & Watson (Reference Ferry and Watson2007), assuming aSiO2 = 1 and aTiO2 = 0.6 for magmatic zircon (CL type 1 and 4) and
aSiO2 = 1 and aTiO2 = 1 for metamorphic zircon (CL type 2 and 3).
SHRIMP and representative LA-ICPMS results are reported in Tables 1, 2 and 3. For additional LA-ICPMS data see online Appendix Tables A2 and A3, available at http://journals.cambridge.org/geo.
5. Results
5.a. CL characteristics
Zircon populations of eclogites (samples 5301 5302, 5306, NW-1 and 1111) consist of two different types. Type 1 zircon is very low luminescent or CL dark and shows diffuse internal features, gradually fading out textures, or faint growth zoning (Fig. 3a). Some grains are almost featureless. The type 1 zircon population includes brown, anhedral grains that often are broken into fragments, and subhedral, mostly turbid grains. Type 2 zircons form euhedral and clear single grains or overgrowths on type 1 zircon. Type 2 zircon is moderately luminescent, lacks remnants of igneous zoning and displays a variety of CL structures that comprise diffuse and blurred zoning patterns, patchy, sector and fir-tree zoning (Fig. 3b).
Figure 3. Cathodoluminescence images of representative zircons from eclogites with SHRIMP and LA-ICPMS ages and spot identification number. Ages are reported with 1σ uncertainty.
Zircons of the granulite sample 5312 show three different types of CL patterns. Type 1 zircons are moderately luminescent, oscillatory zoned grains that often are replaced and rimmed by zircon of type 2 and type 3 (Fig. 4a). Many grains are fractured and strongly corroded. Type 2 zircon is characterized by homogeneous, diffuse or broad patchy zoning patterns, and occurs as single grains or overgrowth around type 1 and type 3 zircon. Type 3 zircons (Fig. 4b) are represented by non-fractured, low luminescent, dark grains that show broad homogeneous domains, patchy CL features or faint growth zoning of variable width.
Figure 4. Cathodoluminescence images of representative zircons from granulites with SHRIMP and LA-ICPMS ages and spot identification number. Ages are reported with 1σ uncertainty.
The zircon populations of granulite sample 1106 consist of anhedral to subhedral grains with variable shapes from rounded to more elongated types. Many grains are fractured. CL imaging revealed internal structures with broad homogeneous domains and sector-zoning (Fig. 4c). Many grains are surrounded by thin overgrowths, which are too small to analyse. Distinct cores were not recognized. Internal features generally have a faint and washed-out appearance. This population is described as type 4 zircon.
5.b. SHRIMP and LA-ICPMS U–Pb geochronology
In order to constrain protolith and metamorphic ages, ion-microprobe U–Pb analyses of three samples (38 spot analyses) and LA-ICPMS dating of six samples (366 spot analyses) were carried out. Results for two representative samples are summarized in Tables 1 and 2 and depicted in Figures 5 and 6. Additional LA-ICPMS data are compiled as Table A2 in the online Appendix at http://journals.cambridge.org/geo.
Figure 5. U–Pb concordia diagrams showing SHRIMP analytical data of samples 5301, 5302 and 5312. Data point error ellipses indicate 2σ uncertainties.
Figure 6. U–Pb concordia diagrams showing LA-ICPMS zircon data of eclogites and granulites. Data point error ellipses indicate 2σ uncertainties.
The new data document a large age variability within all studied samples. In eclogites from Międzygórze (5301, 5302), apparent 206Pb–238U ages range from 329 to 599 Ma (n = 66) and from 320 to 408 Ma (n = 21), respectively (Figs 5a, b, 6). The relatively small dataset of sample 5302 is biased because grain selection was focused on presumably metamorphic zircon. In eclogites from Nowa Wieś (1111, NW-1), 206Pb–238U ages vary between 377 and 533 Ma (n = 43) and between 310 and 609 Ma (n = 69), respectively (Fig. 6c, d). The granulites from the Stary Gierałtów area show a similar distribution of apparent ages (sample 1106: 335–531 Ma, n = 79; sample 5312: 292–538 Ma, n = 88) (Fig. 6e, f).
In all samples, most spots are less than 10% discordant, and there is no obvious correlation between apparent age and U concentration. Data points are distributed along concordia between c. 290 and c. 610 Ma without extreme outliers. Probability density distribution diagrams show several peaks (Fig. 7). Neither at the upper nor the lower end of the age spectra do homogeneous age groups occur which would allow a straightforward interpretation as protolith and/or metamorphic ages. Most data points for type 1 and type 4 zircon grains scatter along concordia between c. 400 and 600 Ma. In contrast, type 2 and 3 zircon grains mostly indicate younger apparent U–Pb ages that cluster at c. 330–340 Ma and c. 370 Ma. In general, there seems to be a younging trend from Cambro-Ordovician ages towards 330–340 Ma, which represents the well-established age of Variscan metamorphism in the study area. Maximum apparent ages in three samples (1111, 5312, 1106) are c. 530–535 Ma. For the eclogite samples 5301 and NW-1, the oldest grains indicate an age of c. 600–610 Ma.
Figure 7. Normalized probability distribution diagrams with stacked histograms for LA-ICPMS zircon data of eclogites and granulites. Bin width = 10 Ma.
5.c. Trace element characteristics
Based on chondrite-normalized REE distribution patterns, the zircon populations can be subdivided into two well-defined groups which closely correlate to distinct CL features. Type 1 and type 4 zircon grains of both rock types are characterized by positive Ce and negative Eu anomalies, and steep HREE slopes (LuN/GdN = 8.1–32.3) (Fig. 8a–d, f, g; Tables 3, A3). In two samples (5312, 1111), some type 1 grains show flat LREE patterns with a negative slope (Fig. 8d, h), possibly indicating accidental analysis of apatite micro-inclusions. Compared to grains with other CL features, concentrations of total REEs (eclogites: 555–6945 ppm; granulites: 413–1928 ppm), Y (eclogites: 910–11425 ppm; granulites: 524–3039 ppm) and Th (eclogites: 127–6496 ppm; granulites: 28–530) are high (Tables 3, A3). U contents vary between 135–2896 ppm (eclogites) and 74–932 ppm (granulites). Th/U ratios range from 0.56–2.24 (eclogites) and from 0.24–0.68 (granulites).
Figure 8. Chondrite-normalized REE patterns of eclogites and granulites from the Orlica-Śnieżnik complex. Normalization to chondrite after Boynton (Reference Boynton and Henderson1984). Line represents magmatic zircons (type 1 or 4); line with diamond symbols represents metamorphic zircon (type 2 and 3). For better visualization, results for sample 5312 are shown in two different diagrams (g, h).
Type 2 and type 3 zircon grains are characterized by flat HREE patterns (LuN/GdN = 0.6–4.4) and a positive Ce anomaly (Fig. 8a–c, e, g; Table 3, A3). Eu anomalies are practically absent (Eu/Eu* = 0.4–1.7). Zircon grains of these groups have low concentrations of total REE (eclogites: 11–157 ppm; granulite: 57–110 ppm), Y (eclogites: 17–197 ppm; granulite: 25–176 ppm) and Th (eclogites: < 1–39 ppm; granulite: 9–326 ppm). U contents vary between 33–3131 ppm (eclogites) and 74–932 ppm (granulite). In eclogites, Th/U ratios generally are very low (< 0.01–0.18), whereas Th/U ratios of the granulite sample 5312 are more variable (< 0.01–1.83) (Table 3).
For type 1 and 4 zircon grains of both rock types, Ti-in-zircon thermometry (Ferry & Watson, Reference Ferry and Watson2007) yielded apparent temperatures of about 790 to 870°C. Ti-in-zircon temperatures for type 2 and type 3 zircon of eclogites, uncorrected for pressure, fall within the range of 690–820°C; such zircons from granulite sample 5312 indicate a temperature of about 885°C.
6. Discussion
6.a. Magmatic v. metamorphic zircon origin
CL imaging revealed four types of zircon: (1) oscillatory-zoned or relatively homogeneous grains with very low to moderate luminescence and without visible inherited cores, (2) moderately luminescent overgrowths and single grains showing a range in internal structures, including relatively homogeneous CL domains, faint and blurred zoning patterns, sector and fir-tree zoning, (3) low luminescent or dark grains that show broad homogeneous domains, patchy CL features or faint growth zoning of variable width, and (4) grains with broad homogeneous CL domains and sector-zoning with a general faint and washed-out appearance. Type 1 and 4 zircon grains are interpreted to be of magmatic origin, whereas type 2 and 3 zircon grains are related to metamorphic processes. In the case of all eclogites, presumed metamorphic and magmatic origins correspond well to extremely low Th/U ratios for overgrowths and related single grains, and moderate Th/U values for protolith zircon. This is not entirely correct for metamorphic zircon of granulite 5312, which has highly variable Th/U ratios up to 1.83 (Table 3). However, the Th/U ratio in zircon is often strongly influenced by protolith characteristics and/or the local chemical environment of formation (e.g. Harley, Kelly & Möller, Reference Harley, Kelly and Möller2007 and references therein), and thus, not too much emphasis should be placed on this parameter to distinguish between magmatic and metamorphic modes of formation.
6.b. What is the reason for the scatter of data points along concordia?
The 206Pb–238U dates for zircon grains of all studied samples show a large variability, and no single age can be calculated. Protolith ages and the time of metamorphic overprints cannot directly be deduced from the new data. Although spot selection was guided by CL images, it cannot be ruled out that some of the scatter is an artifact of beam overlap on neighbouring growth zones producing mixed ages. However, both the LA-ICPMS and SHRIMP datasets show a similar age variability, despite the fact that these techniques exploit considerably different sample volumes. Judging from the CL images, it is considered unlikely that mixing is the major cause for the diffuse distribution of data points. Instead, it is suggested that Pb-loss mainly controls this spread along concordia. A large spread in apparent ages along concordia has already been described in a previous SHRIMP study by Lange et al. (Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a) for the granulite 1106 from Stary Gierałtów. These authors observed no correspondence between degree of apparent resetting and U concentration, and they therefore concluded that non-metamict magmatic protolith zircon experienced variable Pb-loss during Variscan high-grade metamorphism at c. 340 Ma and/or c. 360 Ma. The postulated relationship between resetting and Variscan metamorphism is a plausible interpretation, due to the presence of a distinct younging trends towards an age that is typically associated with the Variscan metamorphic overprint. However, the resetting of presumably non-metamict zircon demands further explanation. Due to the high closure temperature for diffusion of Pb in zircon (> 900°C; Lee, Williams & Ellis, Reference Lee, Williams and Ellis1997; Cherniak & Watson, Reference Cherniak and Watson2001), there are not many possibilities to reset non-metamict zircon under normal crustal conditions. Ashwal, Tucker & Zinner (Reference Ashwal, Tucker and Zinner1999) argued that volume and/or fracture-assisted diffusional resetting is possible, if the blocking temperature for zircon is exceeded for a long time span, or if zircon has been repeatedly affected by granulite- to amphibolite-facies metamorphism. Lange et al. (Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a) considered protracted or multiple high-grade metamorphism as a realistic scenario for the Orlica-Śnieżnik complex, but noted that unambiguous geochronological evidence for such a background had not been established at the time of their study. This situation has changed. Previous Sm–Nd and U–Pb geochronology of granulites yielded ages of c. 350–330 Ma (Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Klemd & Bröcker, Reference Klemd and Bröcker1999; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a), and this age group has mostly been related to short-duration (U)HP metamorphism. However, a recent Lu–Hf and Sm–Nd chronometric study of granulites from the Stary Gierałtów area presented convincing arguments that (U)HP conditions were already attained at some point between c. 387–360 Ma (Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007). These authors related the c. 340 Ma ages to a high temperature metamorphic episode under lower pressures on the retrograde P–T path. Although the geological significance of the 340 Ma ages (eclogite stage or subsequent overprinting) remains controversial (Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009), it is obvious from the Lu–Hf data that the time span during which the granulites experienced high-grade metamorphic conditions is much longer than previously assumed. Thus, the geological context for the granulites seems to agree with conditions that have been shown elsewhere to permit Pb-loss in non-metamict zircon. There are indications that the eclogites have also been affected by prolonged and/or multiple HT events, but for this rock type a different interpretation is more plausible. In the eclogites, many type 1 zircon grains have a very low luminescence, and grains with brown colour are common. These observations are consistent with a high degree of radiation damage, and suggest that in these rocks Pb-loss is mainly a consequence of partial metamictization. It is interesting to note that chondrite-normalized REE patterns of magmatic type 1 and type 4 zircon grains seem to be completely unaffected by the processes that have led to disturbance of the U–Pb system. Despite variable degrees of Pb-loss and partial age resetting, such zircons preserved an igneous REE signature, whereas all metamorphic zircons (type 2 and 3; mostly 330–380 Ma) show flat HREE patterns (Fig. 8). There are no transitional types between both groups.
6.c. What is the protolith age of eclogites and granulites?
We expected that for each sample, a relatively small number of ion-probe analyses (~ 10–15) would be sufficient to determine homogeneous age groups for the central grain parts of igneous zircon. However, due to the considerable spread in apparent U–Pb ages, no single age group could be defined for the first three samples studied, indicating the need for a different analytical strategy. The oldest date for each sample may be used as approximation to the protolith age, but it is not really convincing to place geological significance on a single spot analysis that may be compromised by inheritance or Pb-loss. A cluster of dates would allow a more robust interpretation (e.g. Gehrels, Valencia & Pullen, Reference Gehrels, Valencia, Pullen and Olszewski2006). In order to achieve this objective, a large number of spots were analysed by LA-ICPMS.
Although a comprehensive U–Pb dataset is now available, the protolith ages still remain obscure due to the lack of a well-defined age cluster at the upper end of the age spectra. Normalized probability diagrams and histograms (Fig. 7) document the complexity of this dataset for individual samples by indicating apparent peaks with unclear geological significance between the well-established age of Variscan metamorphism (c. 330–340 Ma) and the upper end of the age spectra, which most likely includes the time of protolith formation. Such complex age patterns are typical for detrital zircon populations, however, at least for the eclogites and the mafic granulite (1106) a sedimentary origin is unrealistic. These rocks are considered to be derivatives of igneous mafic rocks, but the specific original rock type is unknown. For sample 5312 a tuffaceous origin cannot completely be ruled out. Judging from the similarity to the age patterns of samples derived from magmatic precursors and the lack of typical characteristics of sedimentary grains, the zircon population of sample 5312 is not considered to be significantly contaminated by detrital components.
It can be argued that U–Pb ID-TIMS dating, combined with a micro-drill and/or a chemical abrasion technique, has the potential to provide a higher precision of individual data points. However, this analytical approach would not significantly change the general picture. The main problem is not the analytical precision of the SHRIMP and LA-ICPMS dating methods, but variable and possibly multi-stage Pb-loss. Further complexities are added by metamorphic zircon growth and re-equilibration processes, and, to a minor extent, possible mixing of different aged domains during analysis. An additional problem is added by the possibility that inheritance may contribute to the spread in ages. Zircon populations of igneous mafic rocks commonly record no or only minimal inheritance, but examples documenting the presence of inherited components were repeatedly described (e.g. Pilot et al. Reference Pilot, Werner, Haubrich and Baumann1998; Root et al. Reference Root, Hacker, Mattinson and Wooden2004; Peytcheva et al. Reference Peytcheva, von Quadt, Georgiev, Ivanov, Heinrich and Frank2008; Koglin, Kostopoulos & Reischmann, Reference Koglin, Kostopoulos and Reischmann2008; Presnyakov et al. Reference Presnyakov, Lepekhina, Belyatsky, Shuliatin, Antonov and Sergeev2008). All of these factors create a situation where accurate protolith ages cannot be determined with any precision. No solution to the problem is free from ambiguity.
At this point we attempt to outline two possible scenarios that assume no or only minor contamination by inherited components. We consider such circumstances to be very likely, but emphasize that in case of significant contamination by inherited components, other scenarios are possible. If the zircon populations are free of inheritance, the oldest age in each sample can be interpreted as the closest approximation to the true protolith age. This approach would indicate apparent magmatic crystallization ages of c. 531 and c. 538 Ma for the granulites, and of c. 533 Ma for eclogite sample 1111. The oldest grains of eclogites 5301 and NW-1 yield ages of c. 599 and c. 609 Ma, respectively. Accordingly, all samples would provide apparent protolith ages that are older than the precursors of the orthogneisses (c. 490–510 Ma; Oliver, Corfu & Krogh, Reference Oliver, Corfu and Krogh1993; Kröner et al. Reference Kröner, Jaeckel, Hegner and Opletal2001; Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004), supporting interpretations that suggest tectonic juxtaposition of (U)HP rocks and their country rocks (e.g. Don et al. Reference Don, Dumicz, Wojciechowska and Żelaźniewicz1990; Don, Reference Don2001; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004). In each sample, the one to two oldest ages represent single data points that are separated from the remaining population. Assuming that these ages are related to inheritance, it is possible to calculate homogeneous age groups based on the four to six oldest grains of the remaining population. For the eclogite samples 5301 and NW-1, this approach indicates protolith ages > 550 Ma (Fig. A1 of online Appendix; http://www.cambridge.org/journals/geo), whereas eclogite sample 1111 yields an apparent protolith age of 506 ± 6 Ma (Fig. A1). Granulite samples 1106 and 5312 provide apparent ages of 509 ± 9 Ma and 477 ± 7 Ma, respectively (Fig. A1). This would imply that the magmatic precursors of some (U)HP rocks formed largely coevally to the protoliths of the enclosing orthogneisses.
For now, all attempts to unravel protolith ages from this dataset lack definiteness. The time of igneous crystallization of the parent rocks cannot be resolved with confidence. Due to the general scarcity of eclogites and granulites in the Orlica-Śnieżnik complex and the good coverage of the main outcrops by our sample selection, we consider the prospects for determining accurate and precise protolith ages in future studies to be extremely small.
6.d. U–Pb ages and trace element characteristics of metamorphic zircon
The metamorphic evolution of the Sudetes, and in particular the Orlica-Śnieżnik complex, has been a matter of considerable debate. Several studies suggested a polymetamorphic history with discrete HT events that affected Cambrian protoliths during pre-Variscan and Variscan time, in addition to the well-established Carboniferous event (for the most recent overviews, see Żelaźniewicz et al. Reference Żelaźniewicz, Nowak, Larionov and Presnyakov2006; Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009, and references therein). The cumulative probability diagrams of all studied samples show several peaks between the presumed time of protolith formation and the 340 Ma event. Although it cannot be ruled out that these age patterns reflect to some extent distinct tectonometamorphic processes, it is impossible to verify such speculations. At this stage the most reasonable interpretation is to consider most of these peaks as geologically meaningless. The only exception is the accumulation of data points at c. 390–360 Ma (Fig. 7a–f; Fig. A1b), because such dates were also recognized in other geochronological studies that used different chronometers. For example, a recent Lu–Hf and Sm–Nd study documented c. 386–370 Ma ages for (U)HP granulites and closely associated middle crustal metapelites from the Stary Gierałtów area (Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007). In addition, 370–360 Ma ages have also been reported for different occurrences of gneisses, leucosomes and granulites (e.g. Bröcker, Cosca & Klemd, Reference Bröcker, Cosca and Klemd1997; Klemd & Bröcker, Reference Klemd and Bröcker1999; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Gordon et al. Reference Gordon, Schneider, Manecki and Holm2005; Lange et. al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a), but the geological significance remained unclear, due to methodological limitations of the dating techniques used and/or difficulties in linking age information with a specific P–T stage. In contrast to the granulites, previous geochronology did not establish convincing evidence for pre-350 Ma (U)HP processes affecting eclogites and associated gneisses. The new U–Pb data allow the reasoned inference that zircons of eclogites may record variable degrees of metamorphic crystal-chemical changes and/or new mineral growth between 390 and 360 Ma. It is tempting to speculate that these processes reflect (U)HP conditions, as suggested for the granulites of the Orlica-Śnieżnik complex, but there is yet no conclusive evidence to support this hypothesis.
Numerous studies have documented the importance of a c. 340 Ma event for the metamorphic evolution of the Orlica-Śnieżnik complex (e.g. Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Steltenpohl et al. Reference Steltenpohl, Cymerman, Krogh and Kunk1993; Klemd & Bröcker, Reference Klemd and Bröcker1999; Turniak, Mazur & Wysoczanski, Reference Turniak, Mazur and Wysoczanski2000; Marheine et al. Reference Marheine, Kachlik, Maluski, Patocka, Żelaźniewicz, Winchester, Pharaoh and Verniers2002; Lange et al. Reference Lange, Bröcker, Mezger and Don2002, Reference Lange, Bröcker, Armstrong, Trapp and Mezger2005a,b; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009). The new U–Pb ages for type 2 and 3 zircons mostly cluster at 330–350 Ma and further accentuate the importance of Carboniferous metamorphism. Similar zircon ages (SHRIMP, ID-TIMS, Pb–Pb evaporation) were also reported for an eclogite from Nowa Morawa (Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009) and granulites from the Červený Důl and Stary Gierałtów areas (Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a; Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007). The new U–Pb ages closely correspond to results obtained by other chronometric methods for the same rocks or different samples from the same outcrops. For sample NW-1, Bröcker et al. (Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009) reported Rb–Sr and Sm–Nd mineral ages of 346.3 ± 4.2 Ma and 352.2 ± 3.4 Ma, respectively. 40Ar–39Ar and Rb–Sr phengite dating of sample 1111 yielded ages of c. 348 Ma and 331.3 ± 6.5 Ma, respectively. A different eclogite sample which was taken from the same outcrop as sample 5301 provided 40Ar–39Ar and Rb–Sr phengite ages of c. 349 Ma and of 330.9 ± 4.9 Ma. An eclogite sample collected close to the outcrop from which sample 5302 was taken yielded 40Ar–39Ar and Rb–Sr phengite ages of c. 348 Ma and 327.3 ± 4.1 Ma.
The geological significance of Carboniferous ages recorded in eclogites and granulites is controversial and has either been related to peak (U)HP conditions (e.g. Brueckner, Medaris & Bakun-Czubarow, Reference Brueckner, Medaris and Bakun-Czubarow1991; Štípská, Schulmann & Kröner, Reference Štípská, Schulmann and Kröner2004; Lange et al. Reference Lange, Bröcker, Armstrong, Żelaźniewicz, Trapp and Mezger2005a), the waning stages of eclogite-facies metamorphism (e.g. Bröcker et al. Reference Bröcker, Klemd, Cosca, Brock, Larionov and Rodionov2009) or the amphibolite-facies overprint (Anczkiewicz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007). This controversy arises mainly because Sm–Nd ages do not allow unambiguous distinction between late HP metamorphism and amphibolite-facies overprinting, due to overlapping temperatures for both stages that are close to the closure temperature of garnet. Our study provides new arguments for this discussion, because the REE compositions of metamorphic zircon from both rock types strongly suggest contemporaneous crystallization of zircon and garnet. This interpretation is based on the observation of very low total REE abundances, flat HREE patterns, and the absence of an Eu anomaly. All these features strongly point to zircon formation under eclogite-facies conditions (e.g. Rubatto, Reference Rubatto2002; Rubatto & Hermann, Reference Rubatto and Hermann2003; Whitehouse & Platt, Reference Whitehouse and Platt2003). The slight negative Eu-anomaly of type 2 and 3 zircon grains from the granulite sample 5312 (Fig. 8g) is interpreted as an inherited feature. Contemporaneous zircon and garnet growth during amphibolite-facies overprinting is unlikely because only one generation of garnet has been recognized in eclogites and granulites (e.g. Bröcker & Klemd, Reference Bröcker and Klemd1996; Perchuk et al. Reference Perchuk, Korchagina, Yapaskurt and Bakun-Czubarow2005). For granulites, Anczkiewicz et al. (Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007) reported variable preservation of original prograde garnet growth zoning that has been modified by diffusional homogenization creating flat compositional profiles, and minor compositional changes at garnet rims due to partial re-equilibration with matrix phases. Such intra- and intergrain diffusional processes would not affect the REE charcteristics of newly grown zircon. Non-cogenetic garnet generations with distinct chemical compositions were not recognized. Trace element characteristics suggest that zircon growth in both eclogites and granulites can be linked to a HP episode. U–Pb dating of such zircon grains indicates a Carboniferous age for this P–T stage. This does not rule out the possibility that similar Variscan ages of the orthogneisses may record a different metamorphic regime that was attained at the same time in other parts of the orogen.
6.e. Ti-in-zircon thermometry
Ti-in-zircon thermometry (Watson, Wark & Thomas, Reference Watson, Wark and Thomas2006) is based on experiments performed at 10 kbar, and is dependent on activities of SiO2 and TiO2, and pressure (Ferry & Watson, Reference Ferry and Watson2007). In the studied samples, crystallization of magmatic zircon most likely occurred in the presence of quartz and ilmenite (a SiO2 = 1 and a TiO2 = 0.6) at pressures broadly approximated by the experimental conditions. For zircons of CL types 1 and 4, Ti-in-zircon thermometry (Ferry & Watson, Reference Ferry and Watson2007; Table 3) indicates apparent temperatures of c. 790–870°C that are lower than crystallization temperatures of basic melts. Lower than expected Ti-in-zircon temperatures may indicate late stage zircon growth in evolved magmatic systems (e.g. Kaczmarek, Müntener & Rubatto, Reference Kaczmarek, Müntener and Rubatto2008), but other processes influencing the incorporation of Ti into zircon have also been shown to induce a similar effect (Fu et al. Reference Fu, Page, Cavosie, Fournelle, Kita, Lackey, Wilde and Valley2008). Thus, caution is warranted in applying Ti-in-zircon thermometry, because many aspects of this method are not yet fully understood (e.g. Fu et al. Reference Fu, Page, Cavosie, Fournelle, Kita, Lackey, Wilde and Valley2008). For example, Ferriss, Essene & Becker (Reference Ferriss, Essene and Becker2008) questioned the reliability of the Ti-in-zircon thermometer for UHP rocks, because under such conditions, Ti substitution into the Zr instead of the Si site becomes more important. In the studied eclogites and granulites, metamorphic zircon formed in the presence of quartz and rutile (a SiO2 = 1 and a TiO2 = 1) at high to ultrahigh pressure conditions. Ferry & Watson (Reference Ferry and Watson2007) suggested for the Ti thermometer a pressure correction of 5°C/kbar at 750°C, but quantum-mechanical calculations (Ferriss, Essene & Becker, Reference Ferriss, Essene and Becker2008) predict that the necessary pressure correction should be twice as large. Such corrections were not applied, because zircon growth can only broadly be linked to eclogite-facies conditions, but not to a specific pressure. As a result, metamorphic temperatures may be underestimated by c. 50–130°C at pressures between 15 and 28 kbar. Ti-in-zircon temperatures for metamorphic zircon of eclogites, uncorrected for pressure, fall within the range of 690–820°C that has been estimated for these rocks using conventional geothermometry (e.g. Bakun-Czubarow, Reference Bakun-Czubarow1991a,Reference Bakun-Czubarowb; Bröcker & Klemd, Reference Bröcker and Klemd1996). Similarily, for the granulite sample 5312, Ti-in-zircon thermometry indicates metamorphic temperatures of about 885°C, conforming to the previously estimated 800–1000°C range.
7. Conclusions
Attempts to determine accurate and precise protolith ages for eclogites and granulites from the Orlica-Śnieżnik complex have been severely hampered by variable degrees of isotopic resetting that affected the U–Pb isotope system of individual zircon grains. For both rock types, SHRIMP and LA-ICPMS data show a large spread in apparent U–Pb ages, and a younging trend towards 330–340 Ma, which is the well-established age of Variscan metamorphism in the study area. Pb-loss is considered as the most important cause for this age variability. The disturbance of the U–Pb zircon system is observed in samples representing different P–T paths, and can be related to two processes: partial metamictization and/or protracted metamorphism. The individual contributions of these processes to the overall result and their timing are difficult to assess. In the case of the eclogites, radiation damage in combination with metamorphic overprinting is considered to be the dominant factor for post-crystallization disturbance of the isotope system. In the case of the granulites, exposure of these rocks to protracted or multiple HT metamorphism may have produced favourable circumstances to induce partial resetting of non-metamict zircon. The exact age relationships between (U)HP rocks and their host rocks (coeval, younger or older country rocks) remains unresolved.
The last metamorphic overprint at c. 350–330 Ma represents a major episode of new zircon growth and other crystal-chemical changes affecting this phase. The contributions of earlier metamorphic events have yet to be resolved. In eclogites a conspicuous data cluster at c. 360–380 Ma was recognized. It is not understood whether this feature indicates metamorphic processes similar to those indicated by Lu–Hf dating of the granulites (Ancziekiwcz et al. Reference Anczkiewicz, Szczepański, Mazur, Storey, Crowley, Villa, Thirlwall and Jeffries2007) or a fortuitous accumulation of dates representing incompletely reset grains. U–Pb zircon ages of c. 330–350 Ma complement existing datasets that, at least for the eclogites, are largely based on other chronometers. Trace element characteristics of metamorphic zircon from both eclogites and granulites suggest contemporaneous crystallization with garnet, supporting interpretations that relate Carboniferous ages in (U)HP rocks to late-stage eclogite-facies metamorphism and not to middle crustal overprinting.
Acknowledgements
This study was funded by the Deutsche Forschungsgemeinschaft (grant BR 1068/11-1). Reviews by S. Mazur and an anonymous referee are greatly appreciated. Supplementary online Appendix is available at http://journals.cambridge.org/geo.
