Linking growth episodes of zircon and metamorphic ...

3 downloads 0 Views 2MB Size Report
Sveconorwegian Province of southern Norway. (Fig. 1) consists of ..... 1985; Kars et al. 1980) will not .... norwegian Province of SW Norway. Contributions to.
Linking growth episodes of zircon and metamorphic textures to zircon chemistry: an example from the ultrahigh-temperature granulites of Rogaland (SW Norway) A N D R E A S M O L L E R ~'2, P A T R I C K J. O ' B R I E N 1, A L L E N K E N N E D Y 3 & ALFRED KRONER 2

llnstitut fiir Geowissenschaften, Universitiit Potsdam, Postfach 601553, D-14415 Potsdam, Germany (e-mail: amoeller@ geo. uni-potsdam.de) 2Institut fiir Geowissenschaften, Johannes Gutenberg-Universitgit Mainz, Postfach 3980, D-55099 Mainz, Germany 3Curtin University, Department of Applied Physics, Bentley 6102 WA, Australia In-situ U - T h - P b analyses by ion-microprobe on zircon in intact textural relationships are combined with backscatter and cathodoluminescence imaging and trace element analyses to provide evidence for growth episodes of zircon. This approach helps: (a) to unravel the polymetamorphic history of aluminous migmatitic and granitoid gneisses of the regional contact aureole around the Rogaland anorthosite-norite intrusive complex; and (b) to constrain the age of M2 ultrahigh-temperature (UHT) metamorphism and the subsequent retrograde M3 event. All samples yield magmatic inherited zircon of c. 1035 Ma, some an additional group at c. 1050 Ma. This suggests that loss of Pb by volume diffusion in non-metamict zircon is not an important factor even under extreme crustal conditions. Furthermore, the identical inheritance patterns in aluminous (garnet, cordierite _+ osumilite-bearing) migmatites and orthogneisses indicate a metasomatic igneous instead of a sedimentary protolith for the migmatite. Results for the M1 metamorphic event at c. 1000 Ma BP are consistent in all samples, including those from outside the orthopyroxene-in isograd. The latter do not show evidence for zircon growth during the M2 metamorphic episode. Zircon intergrown with or included within M2 metamorphic minerals (magnetite, spinel, orthopyroxene) give an age of 927 _+ 7 Ma (2o-, n = 20). The youngest observed results are found in zircon outside M2 minerals, some overgrown by M3 mineral assemblages (late garnet coronas, garnet + quartz and orthopyroxene 4-garnet symplectites) and yield a slightly younger pooled age of 908 + 9 Ma (2o; n = 6). These textures are relative time markers for the crystallization of zircon overgrowths during discrete stages of the UHT event. These youngest age groups are consistent with the emplacement age of the Rogaland intrusive complex and the last magmatic activity (Tellnes dyke intrusion), respectively. This is direct and conclusive evidence for UHT metamorphism in the regional aureole being caused by the intrusions, and corrects earlier notions that the events are not linked. Trace element behaviour of zircon (Tb/U and Y content) has been tracked through time in the samples and shows variations both within and between samples. This heterogeneous behaviour at all scales appears to be common in metamorphic rocks and precludes the use of 'rules of thumb' in the interpretation of zircon chemistry, but chemical tracers are useful for recognition of zircon growth or recrystallization during metamorphism.

Abstract:

Geochronology provides the constraints necessary to determine rates and duration of geological processes. In high grade metamorphic rocks it is not always possible to choose samples with a simple history and use them as time markers to unravel complex tectonometamorphic histories. M a n y metamorphic rocks preserve disequili-

brium textures which are helpful in deciphering the P - T path, but preclude the use of isochron dating methods on the minerals that provide the P - T information (e.g. garnet, pyroxene etc.). Accessory phase U - P b geochronology, on the other hand, does not require whole-rock equilibrium and can be carried out using micrometre-

From: VANCE,D., MOLLER,W., & VILLA,I. M. (eds) 2003. Geochronology:Linking the Isotopic Record with Petrologyand Textures. Geological Society, London, Special Publications, 220, 65-81. 0305-8719/03/$15 ~) The Geological Society of London 2003.

66

A. MOLLER ETAL.

scale in-situ measurements on accessory phases, of which zircon is the most important. The challenge then is to link the growth of accessory phases, providing the age information, and the major phases providing the P - T information. This study combines in-situ U - T h - P b analysis by ion-microprobe with backscatter and cathodoluminescence imaging in an area with a complex metamorphic history (e.g. Maijer & Padget 1987; Bingen et al. 1996). Imaging and trace element analyses provide evidence for growth episodes of zircon, and are combined with U - T h - P b isotopic data. Intact textural relationships, preserved by microdrilling the zircon from thin-sections, give constraints on the relative timing and are combined with the U Pb data. Trace element analyses by electron microprobe and ion-microprobe of zones within zircon are used to help identify their growth environment. In-situ analysis thus allows the correlation of mineral inclusion relationships with the petrologic metamorphic history and the trace element (e.g. Y, Hf, rare earth elements (REE)) geochemistry of the zircon. The aim of this study is to determine relative timing of zircon growth with respect to petrologically significant phases (e.g. garnet, orthopyroxene, magnetite, spinel). As stated above, one of the remaining challenges in geochronology is assigning age information to a particular stage within a metamorphic cycle. It has previously been common practice to assume that metamorphic zircon dates the peak of metamorphism, usually without particular reference to the process which would allow this. Central to this interpretation is the concept of closure temperature in geochronology (Dodson 1973, 1986) and the assumption that the closure temperature of zircon was close to the metamorphic peak for most metamorphic terranes. Discordant U - P b ages of zircon and spreads of zircon ages were thus often interpreted to reflect incomplete Pb loss by thermal diffusion. However, experimental studies (Lee et al. 1997; Cherniak & Watson 2000) have shown that Pb diffusion in crystalline zircon is negligibly slow for most crustal temperatures. Our field-based study on zircon mineral separates from the granulites of Rogaland, Norway, also suggests that Pb loss by diffusion does not significantly influence the U - P b ages of nonmetamict zircon, even at peak metamorphic temperatures of 950 ~ (M611er et al. 2002). In U - P b zircon geochronology the interpretation of the obtained dates as metamorphic or magmatic events mostly rests on arguments of morphology and Th/U ratios of zircons from mineral separates. In high grade and polymeta-

morphic rocks, where partial melting and zircon growth or dissolution may have occurred in several phases, these tools are often not sufficient. The term 'metamorphic zircon' may actually be misleading, particularly when melting occurs during metamorphism. In a less strict definition all zircon grown during a metamorphic event, by solid state reactions, crystallizing melt or from fluids, is sometimes called metamorphic. In such a case, however, the different crystallization processes and growth environments are obscured. Taken literally, only zircon grown by solid state (metamorphic) reactions may be called metamorphic, whereas zircon crystallized from a melt is magmatic in origin and zircon crystallized from a fluid is hydrothermal. In this contribution we will thus attempt to be specific about the growth process when describing zircon (see Results section).

Metamorphic v. magmatic zircon Different processes recognized to produce zircon during a metamorphic event are: from melt (e.g. Roberts & Finger 1997), by solid state reactions (possibly prograde; e.g. Fraser et aL 1997; Pan 1997; Bingen et al. 2001; Degeling et al. 2001), by dissolution-reprecipitation (Vavra et al. 1996; Pidgeon 1992), or by precipitation from fluids (e.g. Williams et al. 1996). The most important processes altering zircon are: (i) annealing of metamict zircon (e.g. Nasdala et al. 2001; Pidgeon 1992); (ii) recrystallization with or without prior metamictization (Black et al. 1986; Friend & Kinny 1995; Hoskin & Black 2000); (iii) dissolution by corrosive melts and fluids (e.g. Watson 1996); and (iv) metamictization (depends on U and Th content and residence time above or below the temperature at which damage accumulates). Identification of metamorphic zircon is partly based on morphology and internal structure, imaged by backscattered electrons or cathodoluminescence. It is widely accepted that zircon which lacks oscillatory zoning is metamorphic (e.g. Kr6ner et al. 1987). Relatively low abundance of U compared to magmatic zircon (in granitoid rocks), and thus comparatively dark backscatter and bright cathodoluminescence intensity, are also typical features. However, the characterization of metamorphic zircon by low Th/U ratios (e.g. Williams et al. 1996; Rubatto et al. 2001) appears to be limited to those rocks in which other phases compete for Th (e.g. monazite, allanite). It has been shown that metamorphic zircon may inherit the Th/U systematics of its inherited magmatic population throughout two metamorphic events up to

IN-SITU ZIRCON DATING OF UHT GRANULITE

ultrahigh-temperature granulite grade (Mrller et al. 2002). For metamorphic studies it is thus interesting to study the change of internal textures and chemical signatures of zircon through metamorphic cycles. Previous studies have suggested that garnet growth leads to heavy rare earth element (HREE) depletion (e.g. Bea & Montero 1999) and zircon in apparent equilibrium with garnet has distinctive flat HREE pattems and overall reduced REE abundances compared to pre-metamorphic magmatic grain interiors (e.g. Schaltegger et al. 1999; Rubatto 2002). Equilibrium between restite and melt in zircon overgrowths from partially melted granulites should be indicated by identical REE patterns of overgrowths, which tend to have steeper patterns than inherited magmatic cores (e.g. Rubatto 2002). Equilibrium with fluid in different metamorphic environmnents also influences trace element patterns: Rubatto & Hermann (2003) describe, for example, significantly different zircon REE patterns from granulites and eclogites due to fluid effects. In this study, the influence of metamorphic processes on zircon is investigated by the relationship between zircon age and zircon chemistry while retaining textural relationships and tracking zircon chemistry

67

(Th/U, and Y content as a proxy for HREE abundance, though not for the REE pattern) through time.

Geological setting The Rogaland-Vest Agder terrane of the Sveconorwegian Province of southern Norway (Fig. 1) consists of deformed units of banded and migmatitic tonalitic and granitic augengneisses, with minor lenses of amphibolite, metapelite, quartzite, calc-silicate rocks and marble (Hermans et al. 1975). A regional contact aureole envelopes the c. 1200km 2 Rogaland intrusive complex, composed of anorthosites and related rocks. Whereas the main stage of anorthosite plutonism is dated at 931 __+2 Ma (zircon and baddeleyite) and jotunite dykes give an age of 931 __+5 Ma, the end of the intrusive activity is dated at 920__+3 Ma on zircon from ilmenite-norite constituting the Tellnes ore-body (c. 2 5 k m south of the study area), and 915___4 Ma on baddeleyite from the foliated margin of the Egersund-Ogna massif, the northwestern part of the intrusive complex (Sch~er et al. 1996). Inherited zircon from the intrusive complex yields upper concordia intercept ages of c. 1240, 1450 and 1690Ma

Fig. 1. Simplifiedgeological map of the Rogaland anorthosite-norite intrusive complexwith sample locations.

68

A. MOLLER E T A L .

(Duchesne et al. 1987; Schfirer et al. 1996), derived either from the sources of the magmas or reflecting crustal contamination during intrusion (Schfirer et al. 1996). These inherited ages are compatible with the Gothian (1.75-1.5 Ga) and Sveconorwegian (1.25-0.9 Ga) orogenic events established for the Baltic Shield (Gafil & Gorbatschev 1987). The gneisses surrounding the Rogaland intrusive complex have undergone a complex metamorphic history which can be divided into four main phases. The first phase (M1) is an upper amphibolite to granulite facies event at 600700 ~ and 6 - 8 kbar (e.g. Jansen et al. 1985) of the Sveconorwegian orogeny, overprinted by a granulite facies stage (M2), preserved in a 10 to 30 km wide aureole around the intrusive complex. Mapped isograds (e.g. Tobi et al. 1985) show temperatures increasing towards the intrusive complex with incoming orthopyroxene, osumilite and pigeonite (Fig. 1) and a peak temperature of more than 1000 ~ reached close to the contact. The temperatures for these isograds have been estimated at 750 ~ for hypersthene-in at the expense of biotite, 880 ~ for osumilite-in and garnet-out within garnet-bearing migmatites and 900-1000 ~ for pigeonite-in (Jansen et al. 1985; Tobi et al. 1985; Westphal et al. 2003). Low pressures of c. 4 kbar for the high temperature M 2 stage were deduced by Jansen et al. (1985) using equilibria between garnet, orthopyroxene, plagioclase and sillimanite assemblages in the osumilite-bearing rocks. Subsequent experiments on the osumilite-bearing metapelite system (e.g. Carrington & Harley 1995) as well as melting experiments on samples from the layered complex (Van der Auwera & Longhi 1994) also support relatively low M2 pressures. Holland et al. (1996) give an estimate of 5.5 _ 0.2 kbar at 800-850 ~ from an osumilite + garnet + orthopyroxene-bearing assemblage. Thermal modelling of the metamorphic aureole suggests a direct link between intrusion and the M2 isograds. By emplacement in two stages separated by less than 5 Ma, the main anorthosite may have pre-heated the area to 600750 ~ whereas the smaller Bjerkreim-Sckndal lopolith produced the ultrahigh temperature (UHT) regional aureoles (Westphal et al. 2003). Isobaric cooling to an amphibolite facies M3 overprint is reflected in corona, symplectite and exsolution textures formed at conditions of 550700 ~ and 3 - 5 kbar and final retrogression (M4) occurred at greenschist or lower grade (Hermans et al. 1976; Maijer et al. 1981; Jansen et al. 1985; Maijer 1987). The age of M 2 metamorphism is disputed despite the field relations between the intrusive

complex and the metamorphic isograds surrounding it. Previously published ages of 1.01.05 Ga for zircon, monazite and titanite from migmatites and orthogneisses of the aureole were interpreted as the age of M2 metamorphism by Nijland et al. (1996), thus much older than the Rogaland intrusive complex. Wielens et al. (1980) also obtained ages of 10201000 Ma for metamorphic zircon. Age limits on the timing of early regional metamorphism come from the dating of intrusive bodies in the area. Zhou et al. (1995) obtained a U - P b zircon age of 1159 __+ 5 Ma on the deformed charnockitic gneiss from Hidderskog (60 km SE of Egersund), which provides a maximum age limit for regional metamorphism. R b - S r wholerock isochrons of 980 + 14Ma for the undeformed Holum granite in the Mandal area (Wilson et al. 1977) and of 998 ___ 14 Ma for the D3 deformed late Homme granite in the Flekkefjord area (Falkum & Pedersen 1979) provide the lower age limit for regional deformation and metamorphism. Evidence for a younger age of further high grade metamorphism linked to the intrusive complex includes a conventional monazite U - P b age of 9 5 7 _ 8Ma from a gneiss directly at the contact to the BjerkreimSCkndal intrusion, a minimum estimate of 931 Ma for zircon from acidic gneisses in the undeformed, leuconoritic Hidra massif (both Pasteels et al. 1979), a 930 • 21 Ma R b - S r age for a Tellnes jotunite dyke (Wilmart et al. 1989), and a K - A r age of 970 __+ 30Ma for osumilite (Maijer et al. 1981), albeit with large error brackets. A first link between reactions involving accessory minerals in the metamorphic aureole and U - P b ages obtained by thermal ionization mass spectrometry (TIMS) dating of zircon, monazite and titanite in orthogneisses was established by Bingen et al. (1996), and Bingen & van Breemen (1998a, b). The Feda suite, a north-trending series of calc-alkaline augengneisses at the northeastern end of the sampled profile at Osen, contains magmatic hornblende, titanite, allanite and zircon as well as metamorphic monazite and thorite and yielded an age of intrusion of around 1050 Ma from magmatic zircon (Bingen & van Breemen 1998a). Metamorphic monazite in these rocks and other charnockitic gneisses, are concordant or nearconcordant at 1024-997 Ma and are interpreted to result from high temperature breakdown of allanite, titanite and biotite during the regional M1 phase (Bingen & van Breemen 1998b), whereas later reactivation, and possibly also cooling, is recorded by a spread of monazite ages down to 970 Ma.

IN-SITU ZIRCON DATING OF UHT GRANULITE

The thermal effect of the anorthosite-norite intrusion (M2) and the subsequent cooling (M3) is represented by the growth of clinopyroxene at the expense of hornblende in the Feda suite orthogneiss (this isograd lies to the east of the orthopyroxene-in isograd, Fig. 1). This reaction is accompanied by a distinct phase of monazite growth at 930-925 Ma. Further evidence consists of titanite ages throughout the area clustering tightly at 918 _ 2 Ma, hornblende A r - A r ages in the range 930-904 Ma, and a further, distinct, low-U monazite group yielding an age of 912-904 Ma (Bingen & van Breemen 1998b; Bingen et al. 1998). Rb-Sr, K - A r and A r - A r ages in the range 895-853 Ma are interpreted to reflect subsequent hydrothermal alteration of mica and hornblende (Verschure et al. 1980; Bingen et al. 1998). Conventional (TIMS) U - P b geochronology studies on zircon separates from gneisses and granulites of the Rogaland aureole (Pasteels et al. 1979; Wielens et al. 1980; Bingen & van Breemen 1998a) did not find clear evidence for a link between intrusive activity and metamorphism. In the present study we combine textural relationships of zircon with metamorphic minerals with in-situ ion-microprobe analyses on samples from different zones of the metamorphic aureole around the intrusive complex. Studying the behaviour of zircon from magmatic crystallization through multiple episodes of metamorphism with this approach should help to explain why M2-related new zircon growth may have eluded the previous studies.

Analytical methods The locations of zircons were determined in polished thin-section using petrographic and backscattered electron (BSE) observations. Zircons with wide overgrowths or included in metamorphic minerals were chosen and drilled from the thin-section with a Medenbach microdrill, preserving the original texture. About 60 thin-section discs of different diameter per sample were mounted each on two 1 inch diameter epoxy discs. After further polishing the internal structures of zircon were documented in detail by BSE and cathodoluminescence (CL) imaging. Major elements and selected trace elements (Hf, Y, P in some cases also U, Th, Yb) in zircon were measured with the JEOL 8900 RL electron microprobe at the Institut ftir Geowissenschaften in Mainz, using a 20 kV and 100 nA beam focused to 5 Ixm. Uranium, thorium and lead isotopic measurements were made during four sessions on the Perth Consortium SHRIMP II non-microprobe, employing operating,

69

data-processing and error calculation procedures described by Nelson (1997). Further details of the procedures are given by Mrller et al. (2002) and data tables can be obtained from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, West Yorkshire, LS23 7BQ, UK, as Supplementary Publication No. SUP18191 (5 pages). It is also available online at http://www, geolsoc.org.uk/SUP18191.

Sample description Samples of different peak metamorphic grade were collected along a traverse perpendicular to the margin of the intrusive complex and to the metamorphic isograds (Fig. 1), along state road No. R42 from Osen to Egersund. Two additional UHT granulite locations were sampled further to the north (22 and FFJ). Overall the samples can be divided into two broad groups: aluminous migmatites, previously interpreted to be metasediments (12E, 22H and 22D); and granitic to granodioritic orthogneisses (2B, 10B, 16A, 17C, 19A, FFJ3). The lowest grade sample is 2B, an amphibolitefacies granodioritic augen-gneiss from just outside the orthopyroxene-in isograd. Sample 10B from a roadcut of an old road parallel to and above the current road is from within the orthopyroxenein and the osumilite-in isograd. It is granodioritic and layered, showing thin quartz-rich bands alternating with orthopyroxene + plagioclaserich hands. Sample 12E was collected from an outcrop along the road in Gyadalen, c. 2 km inwards of the osumilite-in isograd. The sample is a migmatitic gneiss with large garnet porphyroblasts, and a garnet + orthopyroxene + biotite-rich selvage from the transition zone between melanosome and leucosome. Sample 16A was taken from a roadcut of the old road, parallel to Gya tunnel, just outside the pigeonite-in isograd. Sample 16A is the transition between a pegmatitic leucosome layer with very large feldspar and a garnet + orthopyroxene + magnetite lens. Sample 17C is a layered charnockite with leucosome and melanosome and was collected along Gyavatnet lake within the pigeonite-in isograd area, from the same outcrop as sample B649 of Bingen & van Breeman (1998b). Sample 19A is charnockitic, is similar to sample 17C, and was collected at the western end of Gyavatnet lake dam. Location 22 with samples 22D and 22H is the osumilite locality investigated by several other authors (e.g. Maijer et al. 1981; stop 8.1 in Maijer & Padget 1987). Leucocratic layers of quartzfeldspar, with occasional large centimetre-sized orthopyroxene, alternate with dark layers and

70

A. MOLLER ETAL.

specs of pinkish hue containing osumilite and garnet relics. Location FFJ has outcrops directly at and to the west of the road from Skjeveland to Nedrab6. This is stop 9.1 in Maijer & Padget (1987). Leucocratic gneisses of the Faurefjell sediments crop out west of the road as part of the marblebearing succession described by Bol et al. (1995). The collected sample is a leucocratic gneiss with macroscopically visible layers of orthopyroxene.

Results M e t a m o r p h i c textures a n d zircon o c c u r r e n c e

Textural relationships between zircon and metamorphic minerals can provide relative brackets on the timing of development of the different assemblages. Thus, the youngest U - P b zircon age from grains included in prograde M2 or retrograde M3 minerals provides a maximum age for the M2 or M3 stages of metamorphism, respectively. U - P b ages on metamorphic rims of zircon intergrown with another metamorphic mineral directly provide the age of this metamorphic stage. Therefore, it is of ultimate importance to observe and describe such textures in detail. Evidence for these textures is presented in photomicrographs in Figure 2 and at larger magnification in secondary electron (SE) images in Figure 3. For the current study the most important textural relationships are: (i) inclusions of zircon found in M2 minerals such as magnetite (sample 12E, see Figs 2A, 3A) or orthopyroxene (samples 10B, 22H in Fig. 2C, D, FFJ3); (ii) intergrowths of zircon with M2 minerals with lobate grain boundaries (with quartz in 12E, with perthitic feldspar in 22H; see Figs 2B, 3B) suggesting crystallization from a melt; or (iii) intergrowths with high angle grain boundaries (with spinel or magnetite in 12E, 16A, 22D, 22H; see Figs 2B, 3B) indicative of textural equilibrium and contemporaneous growth of zircon and metamorphic mineral; and (iv) the zircon attached to other minerals (magnetite, ilmenite, spinel) and overgrown by garnet coronas (samples 12E, 16A; Figs 2E, 3C) and subsequently overgrown by garnet-quartz (12E) or garnet-orthopyroxene (16A; see Figs 2F, 3D) symplectites. These provide the necessary textural constraints on the timing of zircon crystallization, provided that there was no significant change to the U - P b system after inclusion or overgrowth occurred. The youngest obtained ages provide a maximum for the age of the development of the textures and thus

for the metamorphic stage preserved in this texture. Internal zoning o f zircon

Description of the investigated zircon and interpretation of the analytical results relies mainly on features observed in CL and BSE images. It is therefore necessary to summarize how these features are interpreted in this study. Brightness of CL and BSE is usually anticorrelated in the studied zircon, with CL-dark areas usually appearing BSE-bright. Some features may be seen in only one of the image types rather than both. Figure 3 thus shows both CL and BSE images for representative grains from some samples. Parts of grains with discrete fine-scale oscillatory zoning (some cores in Fig. 3; see also fig. 2 of Mrller et al. 2002) are interpreted to reflect magmatic growth (e.g. Pidgeon 1992), whereas sector zoning is interpreted to reflect slow growth from a fluid-rich boundary layer (Watson & Liang 1995; Schaltegger et al. 1999). Metamorphic zircon is represented by recrystallized parts of grains as well as overgrowths on existing grains. We interpret those areas showing little or no internal zoning and with shapes unrelated to existing features (i.e. magmatic cores) as overgrowths (see Fig. 3B, C and fig. 2 of Mrller et al. 2002). Dissolution may have occurred before and be unrelated to the precipitation process, usually producing sharp boundaries between inherited magmatic portions and overgrowths (see Fig. 3B and fig. 2 of Mrller et al. 2002). Areas with lobate or transgressive boundaries into zoned parts of grains (showing, in some cases, internal zoning and relics of earlier features) are interpreted to result from complete or partial solid-state recrystallization (see Fig. 3C, D; e.g. Pidgeon 1992; Vavra et al. 1996; Hoskin & Black 2000), a process faster than solid-state diffusion (Mrller et al. 2002). Note that post- or latemagmatic solid-state recrystallization, possibly aided by late-stage fluids, can produce similar features to recrystallization during metamorphic events (probably because the fluid-related mechanism is essentially the same). Zircon with wide metamorphic overgrowths suitable for ion-microprobe occur mainly in two key textures in these granulites: melanosomes or melanocratic selvages on leucosomes in migmatites (e.g. sample NRI2), and accessory phaseand orthopyroxene-rich layers in granitic to granodioritic rocks. Zircon with preserved oscillatory zonation, with or without metamorphic overgrowths or alteration due to recrystallization has been found throughout the

IN-SITU ZIRCON DATING OF UHT GRANULITE

71

Fig. 2. Representative photomicrographs of key textural relationships between zircon and metamorphic minerals analysed in this study. (A) Sample 12E, position 5.i. Large zircon is included in titaniferous magnetite (with corundum exsolution along rim and cracks) and biotite. Outer rim with M2 metamorphic age has inclusion of xenotime and has elevated Y and P contents. (B) Sample 22H, position 6.8. Zircon attached to magnetite and spinel, with lobate grain boundary towards feldspar. (C) Sample 22H, position 6.19. Zircon attached to magnetite and included in M2 orthopyroxene porphyroblast. (D) Sample 22H, position 6.2. Zircon attached to magnetite and included in M2 orthopyroxene porphyroblast, with high angle grain boundaries at lower end. (E) Sample 16A, position 2.7. Zircon is attached to magnetite, an M3 retrograde garnet corona has developed on magnetite and encloses the zircon. (F) Sample 16A, position 2.8. Zircon is attached to magnetite, an M3 retrograde garnet corona has developed between orthopyroxene and magnetite and overgrows the zircon. This is overgrown by M3 symplectite of orthoyproxene and garnet.

area, i n d e p e n d e n t of m e t a m o r p h i c grade. Several phases of overgrowth can be identified in zircon from m a n y samples, and overgrowths or recrystallized zones are found to be widest, and zircon most abundant, in melanocratic layers of

the investigated gneisses. The overgrowths m a y be epitaxial and continuous on the inherited cores, in some cases preserving the m a g m a t i c prismatic shape (e.g. Fig. 3A). In other cases, zircon was found to be anhedral, with featureless

72

A. MOLLER ETAL.

Fig. 3. Representative secondary electron (SE), cathodoluminescence (CL) and backscattered-electron (BSE) images. Images taken after SHRIMP analysis (note ion-microprobe pits, best visible on BSE images). U-Pb results are dates with 1~r errors. Larger pits without age information are from ion-microprobe analyses indicated a s 2~ for trace element analyses (unpublished data). (A) Sample 12E, position 5.i. Large zircon is included in titaniferous magnetite. Outer rim with M 2 metamorphic age has inclusion of xenotime and has elevated Y and P contents (Table 2). (B) Sample 22H, position 6.8. Zircon with M2 metamorphic overgrowth on inherited core. Note lobate boundary with feldspar. (C) Sample 16A, position 2.7. Zircon enclosed in M3 (retrograde garnet corona shows distinct inherited core and recrystallized area towards grain boundary with magnetite. (D) Sample 16A, position 2.8. Zircon attached to magnetite, and enclosed in M3 retrograde assemblages. The recrystallized rim gives the youngest observed age in all samples and thus a maximum age for development of M3 assemblages. rims discontinuously overgrowing older parts (e.g. Fig. 3C). Anhedral, internally featureless zircon, possibly completely newly grown, has only rarely been observed: most zircon grains contain an inherited core.

U - P b geochronological results The same age groups of zircon have been found in all samples and can be summarized into five main groups. Two groups of ages (Fig. 4C) can

be distinguished within the timeframe of the emplacement of the Rogaland intrusive complex by in-situ analyses. The younger group of analyses at 908 + 9 Ma consists of some results with lower T h / U from sample 12E and two more analyses from orthogneisses 10B and 16A. These are rims of zircon grains overgrown by M3 assemblages (garnet coronas on opaques and garnet + orthopyroxene symplectites), thus marking a m a x i m u m for the age of M 3. The older of the two groups at 927 __ 7 M a is more c o m m o n

1N-SITU ZIRCON DATING OF UHT GRANULITE

73 o

~6 ~

~

~D

~

~

~:

N

~

o

~:

N

N

~ ~ .~ = :

~

~

~

~

0~.~

9

~

.0

0

~,~

~

. . . .

~ 0 ~ . ~ "~ ~ ' ~ . ~. ~ ~ . ~ ' o~~

aom~,~

~

9

o ~

oo,JD

~ = ~:.~. I

.~

~.~ I

~ ,..~

~ .~ ~ ~ .~=

~

s~e

.~-~

74

A. MOLLER ETAL.

(n = 20) and consists of rims on grains included in or intergrown with Me minerals (e.g. magnetite, spinel, orthopyroxene). Another important result for the UHT M2 event is the lack of M2 zircon in samples from outside the orthopyroxene-in isograd (M611er et al. 2002). The MI metamorphic phase can be dated at about 1000 Ma based on results from mineral separates (M611er et al. 2002) and the present in-situ study (Fig. 4a, b). Most of the zircons from this age group can be identified as metamorphic zircons because they are featureless rims with varying chemical character (Fig. 5), overgrowths as well as recrystallization products. Inherited magmatic cores are preserved in all analysed samples and yield essentially the same age results in all samples (Fig. 4a, b, d) with a larger group of results at 1035 Ma and a smaller one at c. 1050 Ma, consistent with conventional TIMS dating of zircon by Bingen & van Breemen (1998a).

Tracking trace element chemistry of zircon through time

Figure 5 shows the relationship of Th/U and of Y content to the 2~ age for zircon from the samples analysed in-situ. Note the different scale in the age axis for sample 12E and that Y contents are not available for all zircon analysed by SHRIMP. Zircon in Figure 5a, from the melanosome of migmatite sample 12E, clearly shows that the Th/U ratio cannot be a guide for the distinction between metamorphic and magmatic zircon. It is clear that although most analysed zircons have ages corresponding to the M 1 and M2 event, the bulk of the analyses do not fall below the line at a Th/U ratio of 0.1, thought to indicate metamorphic zircon (e.g. Williams et al. 1996). It has been argued that concurrent monazite growth is necessary to produce the low Th/U signature in zircon because of competition for Th (e.g. Williams 2001), but M1 monazite (and possibly M2 is present in the migmatite samples and neither of them shows consistently depleted Th/U in the zircon of metamorphic age. This may be in part the result of inheritance of trace element characteristics during solid-state recrystallization of zircon, as advocated by Pidgeon (1992), a closed system behaviour that can account for unchanged Th/U and Y through time (e.g. sample 10B). However, the dramatic changes observed in Th/U over time in sample 22H (below 0.1 to 0.3 in magmatic zircon and 0.2 to 0.8 in M2 metamorphic rims) or the heterogeneous distribution observed in sample 12E (zircon with Th/U below 0.1 in the M2 and M1 age groups, but zircon of all ages with Th/U

as high as 0.6) may be explained by the dependence of zircon chemistry on local equilibria or different growth mechanisms (crystallization from partial melt versus recrystallization). Some analyses from the granitic gneiss 16A show low Th/U for MI zircon (Fig. 5b), but higher values again for M2 zircon. This may be interpreted to reflect the dissolution of monazite (or other high Th/U phases) during M2 or between M1 and M2. Figure 5C shows two distinct groups of analyses: a high-Y group for the timespan of the metamorphic events, and a low-Y group covering all ages. The high-Y analyses are from rims of zircon grains which generally have xenotime inclusions (note that not all inclusions may be visible in the plane cut by the thin section). These xenotime grains observed within M2 metamorphic rims (Fig. 3A) do not appear to be due to exsolution of xenotime component previously present as solid solution in high temperature magmatic or metamorphic zircon cores, because no xenotime has been found in magmatic cores of grains, and cores do not have elevated Y or P contents. Instead, the xenotime-bearing rims have high Y and P, indicating that xenotime coprecipitated while zircon was saturated in xenotime solid solution. Whole-rock analyses for sample 12E also show elevated Y, more than five times enriched compared to the granitoid gneisses. The enrichment of Y on the whole-rock scale is interpreted to be the reason for the observed feature, the cause of which has not been determined. Figure 5C shows that this enrichment in Y is an early process possibly originating during M~, affecting only few zircons. The simplest scenario to account for the observed results and textures is one in which prograde partial melting during M1 produced melt + garnet at the expense of biotitebearing assemblages. Melt escape and the restitic garnet would produce the magmatic texture and a melanosome enriched in Y. Rare high-Y M1 zircon (Fig. 5C) may then have grown during crystallization of the melt in equilibrium with garnet within this melanosome. The subsequent prograde breakdown of garnet during the UHT M2 event released Y and Zr to produce the highY zircon and coprecipitated xenotime. It is suspected that the low-Y M2 zircon results from recrystallization processes rather than new growth, and thus did not tap the high-Y reservoir. These examples show that these types of diagram and the tracking of zircon chemistry through time can be instructive for the interpretation of processes responsible for zircon growth and recrystallization during metamorphism. The results can then be cross-checked with qualitative

IN-SITU ZIRCON DATING OF UHT GRANULITE

75

o

~N

O o

g.= ,,.-.,

E

.=

,.t:Z ' , ~

9

'Saz

76

A. MOLLER ETAL.

image interpretation of internal zoning patterns (Fig. 3).

Discussion Metamorphic reactions in the Rogaland regional aureole

For the sake of brevity, we refer only to the mineral reactions important for the present study here or to reactions recently studied for their potential for growth of new zircon. Detailed accounts of the metamorphic reactions during the different metamorphic episodes affecting the various rock types found in Rogaland have been presented by Maijer (1987). The metamorphic reactions observed in our samples and described above and below are part of these reaction sequences and can therefore be tied to specific stages in the evolution of the metamorphic complex. Maijer et al. (1981) describe orthopyroxene, osumilite, cordierite, hercynitic spinel-magnetite intergrowths, plus quartz and high temperature feldspar as the characteristic M 2 metamorphic minerals in the orthogneisses and migmatites, as described above to include zircon with rims belonging to the 927 _+ 7 Ma age group. Mineral assemblages of the M3 event of lower granulite to amphibolite facies, on the other hand, are characterized by plagioclase, orthoclase, quartz, and second generations of biotite, orthopyroxene, garnet and sillimanite. The effect of M3 retrograde cooling is seen in garnet coronas on magnetite + spinel intergrowths. Westphal & Schumacher (1997) also describe secondary garnet § quartz rims on orthopyroxene and primary garnet (outer 150200 pore) as post-peak-M2 assemblages. Garnet § orthopyroxene symplectites have been described and explained by Maijer (1987) to be the result of the breakdown of high-A1 UHT orthopyroxene and also belong to the M3 assemblage. The typical succession of M3 reaction with garnet corona between opaque phase and orthopyroxene and garnet § orthopyroxene symplectite is the texture shown in Figure 2F to overgrow zircon with a rim belonging to the 908 _+ 9 Ma age group. Maijer (1987) attributes these retrograde assemblages to slow to very slow cooling from M2 to M3, a view consistent with our geochronological results. The potential of metamorphic reactions f o r zircon growth

The potential of metamorphic solid-state reactions as an important source of Zr and thus their

potential for new metamorphic zircon growth has been evaluated by in-situ analyses of the Zr content of the reactant minerals and balancing the reactions (e.g. Fraser et al. 1997). Degeling et al. (2001) have studied the Rogaland granulites with this method. They propose that the reaction garnet + biotite + quartz + sillimanite osumilite § orthopyroxene § spinel + magnetite which defines the osumilite-in and garnet-out isograd in rocks of appropriate composition (e.g. Tobi et al. 1985; Kars et al. 1980) will not produce zircon because Zr is taken up by osumilite and orthopyroxene. A different reaction in aluminous migmatites, occurring at lower grade close to the hypersthene-in isograd, is: garnet § sillimanite quartz --~ cordierite. This is interpreted by the authors to occur during decompression after M1 in Rogaland at c. 5.6 kbar and 710 ~ and can be balanced to produce zircon because Zr will not enter cordierite (Degeling et al. 2001). Textural evidence supports this interpretation with micro-zircons as abundant inclusions in cordierite. This is one of the reactions studied earlier by Fraser et al. (1997), who also concluded that there was potential for new zircon. Combined with the new SHRIMP U - P b data, this reaction or similar retrograde reactions after M1 may account for some of the 980 to 965 Ma results on zircon, which cannot be explained by overlapping analyses or incomplete recrystallization. Detailed studies on balancing metamorphic reactions for zircon growth potential in the samples used in this study are in progress and will be the subject of a separate publication. As discussed above, the textural evidence alone presented above (Fig. 2) can bracket the age of the M 2 and M3 metamorphic events. A comment on previous geochronological results

It can be speculated that one of the reasons for sparse evidence for M 2 UHT zircon ages from previous conventional TIMS studies is the textural relationship of M2 metamorphic zircon with magnetite. As described above, zircon is often found included in, intergrown with, or attached to magnetite (Fig. 2). These types of zircon may have escaped attention in standard procedure zircon separates, because zircon is usually obtained from the least magnetic

IN-SITU ZIRCON DATING OF UHT GRANULITE

fractions during magnetic separation. Zircon not separated from magnetite would therefore have been overlooked, unless a specific effort had been made. It has been observed that zircons in magnetite-free leucosomes are usually prismatic grains with thinner metamorphic rims and would therefore likely be more prone to inheritance and yield results close to earlier, magmatic events. Thin metamorphic rims may also be removed during air abrasion of grains in preparation for TIMS U - P b zircon dating.

Regional implications

Thermal modelling (Westphal et al. 2003) studies and our geochronological results demonstrate that the M2 UHT metamorphic evolution in the region is directly linked to the intrusion of the Rogaland intrusive complex. Textural relationships linked with in-situ ion-microprobe zircon dating unequivocally constrain the age of the prograde mineral assemblage to be younger than 927 __+ 7 M a and subsequent M3 retrograde assemblages to be younger than 908 + 9 Ma. Because these results are very closely spaced in time, it can be concluded that M 2 and M3 metamorphism stages are part of the same, relatively short-lived regional-scale contact metamorphic event as part of a long period with higher than normal geothermal gradient. This is supported by the thermal modelling of Westphal et al. (2003), who calculated that the country rocks to the Rogaland intrusive complex must have been at c. 600 ~ already at c. 15 km depth (geothermal gradient of approximately 40 ~ when intrusion occurred in two phases and heated the rocks further to UHT conditions, which raised the geothermal gradient at the pigeonite-in isograd to almost 60 ~ (880 ~ at 5 kbar). Hornblende Ar/Ar ages of 870880Ma in the region (Bingen et al. 1998) indicate that despite the relatively shallow crustal level, the region did not cool below the closure temperature of hornblende (c. 570 ~ until about 50 Ma after intrusion and peak M 2 metamorphism. An integrated cooling rate of c. 6 ~ is very slow for this mid-crustal level, and an additional, slowly decaying heat source must be considered to explain this. We propose that the most likely scenario that can account for the magmatism as well as raised geothermal gradients over a longer period of time is a very shallow asthenospheric mantle immediately underlying the crust, i.e. a scenario of delaminated lithospheric mantle. This is not unlike the extensional scenario proposed by Bingen et al. (1998) for the M 2 and M3 phases of

77

Sveconorwegian metamorphism in the Rogaland region. The calculations of Westphal et al. (2003) also indicate that the intrusions would not have lifted the regional temperatures significantly (by less than 100 ~ at the distance of the orthopyroxene-in isograd, a result consistent with the observation that M2 zircon growth has not been observed in rocks further from the intrusion than our sample NR10, beyond the osumilite-in isograd (M611er et al. 2002; this study). It may thus be inferred that the regional orthopyroxenein isograd (and the clinopyroxene-in isograd even further away) may at least partly be a feature of the M1 metamorphic event, an assumption which has been made previously by some authors (e.g. Bingen et al. 1998). Our U - P b zircon results also support the conclusions that syndeformational magmatic activity in the area occurs largely within a short time period between c. 1050 Ma and 1030Ma, during the later compressional phases of the Sveconorwegian orogeny (Bingen et al. 1998). It should be noted that the aluminous migmatitic gneisses, which have mostly been regarded as metasediments, contain the same age groups of inherited zircon as the surrounding granitoid orthogneisses. A metasedimentary origin of these migmatites is therefore concluded to be unlikely. They may be explained as metasomatically altered granitic gneisses. Conclusions

A compilation of the available SHRIMP U - P b data on concordant zircon shows three main age groups (Fig. 4d): (i) a magmatic group at c. 1035 Ma; (ii) an M1 metamorphic group at c. 1000 Ma; and (iii) an M2-M3 metamorphic group at the time of anorthosite-norite intrusive activity at c. 930 Ma to 910 Ma. It is important to note that the age of the inherited age components is not shifted due to Pb loss in even the highest grade granulites and that identical age groups, within error, have been found throughout the area. We interpret this as evidence that Pb loss by volume diffusion is not an important process for non-metamict zircon under crustal conditions and that, instead, recrystallization in the solid state and interaction with fluids and melts are far more effective in affecting zircon in high grade metamorphic rocks. Either recrystallization or dissolution-reprecipitation within a closed system may be responsible for the fact that in some of the samples zircon shows no change in Th/U or Y characteristics between magmatic and both M1 and M2 metamorphic zones (e.g. sample 10B, see Fig. 5). It is also important to note that a large

78

A. MOLLER ETAL.

enough number of analyses can reveal that zircon of the same age population can have widely different trace element characteristics (e.g. sample 12E and 22H, Fig. 5). Yttrium content and Th/U, on the other hand, do not show simple correlation trends, which suggests that trace element chemistry of tetravalent and trivalent cations, and thus age and trace element chemistry, are decoupled in these samples. These are observations that we interpret to reflect either heterogeneity of processes or disequilibrium within the sample and between zircon and the rest of the sample.

The use of trace element characteristics of zircon as fingerprints for geological processes This study shows that Th and U alone as geochemical tracers have limited use for the interpretation of zircon ages and, more specifically, the distinction between magmatic zircon and zircon yielding metamorphic ages. It is important to point out that metamorphic zircon is in many cases not recognizable by low Th/U, contrary to what has been observed by a number of other studies (e.g. Williams et al. 1996; Rubatto et al. 2001). Relatively high Th/U, or no change in comparison to magmatic zircon, has instead been observed (Fig. 5) and we suggest that each step in the reaction and mass balance history of the rock affecting the environment of zircon has the potential to change its trace element characteristics. Geochemical characterization of zircon by trace elements (Hf, Y, P and REE) has the potential to become a useful interpretative tool in the future only when the mechanisms controlling trace element behaviour are better understood and lead to a less empirical interpretation of zircon geochemistry.

approach has been successfully applied to insitu analysis of monazite (e.g. Foster et al. 2000). In the Rogaland metamorphic aureole, zircon included in and intergrown with M2 minerals give an age of 927 + 7 Ma (2o; n = 20). Zircon outside M2 minerals, some overgrown by M3 mineral assemblages (distinguished by low T h / U in migmatite sample 12E), yield a distinctly younger pooled age of 908 _+ 9 Ma (2o-, n = 6). These age groups are consistent with the emplacement age of the Rogaland intrusive complex at 931 • 2 Ma and the last magmatic activity at 920 + 3 Ma (Tellnes dyke intrusion), respectively (Sch/irer et al. 1996). We interpret these results as direct and conclusive evidence for UHT metamorphism in the regional aureole being caused by the intrusions, which corrects earlier models in which the events were not linked. The microgeochronological approach thus marks an important step towards a more direct correlation of U - P b zircon geochronology with petrologically determined P - T paths. This is true especially in polymetamorphic terranes with a complex and multiphase history, which prevents isochron isotopic methods requiring equilibrium to yield correct results. We thank N. Groschopf,M. Mtiller and B. Schulz-Dobrick for assistance with the electron microprobe analyses. We are grateful for the reviews of D. Vance, G. Foster and M. Whitehouse, which helped to improve the manuscript. A.M. thanks B. Mocek for assistance in sample preparation, data presentation and for discussions. U-Th-Pb zircon analyses were carried out on the Sensitive High Resolution Ion Microprobe mass spectrometer (SHRIMP II) operated by a consortium consisting of Curtin University of Technology, the Geological Survey of Western Australia and the University of Western Australia with the support of the Australian Research Council. This research was financially supported by the Deutsche Forschungsgemeinschaft (DFG) through grants Kr 590/62-1 and Kr 590/62-2 to A.K.

Microgeochronology: linking growth episodes of zircon to metamorphic textures

References

Metamorphic zircon overgrowths are best observed in-situ, because some zircon and its growth textures will be destroyed by standard mineral separation techniques. The combination of in-situ observations and imaging techniques with in-situ electron microprobe trace element analysis allows distinction of growth phases of zircon. These can then be targeted by ionmicroprobe isotope analysis to obtain ages to be directly linked to distinct episodes in the metamorphic history and are thus more meaningful for geological interpretations. This

BEA, F. & MONTERO, P. 1999. Behavior of accessory phases and redistribution of Zr, REE, Y, Th, and U during metamorphism and partial melting of metapelites in the lower crust: an example from the Kinzigite formation of Ivrea-Verbano, NW Italy. Geochimica et Cosmochimica Acta, 63, 1133-1153. BINGEN, B. & VAN BREEMEN, O. 1998a. Tectonic regimes and terrane boundaries in the high-grade Sveconorwegian belt of SW Norway, inferred from U-Pb zircon geochronology and geochemical signature of augen gneiss suites. Journal of the Geological Society London, 155, 143-154.

IN-SITU ZIRCON DATING OF UHT GRANULITE BINGEN, B. & VANBREEMEN,O. 1998b. U - P b monazite ages in amphibolite- to granulite-facies orthogneiss reflect hydrous mineral breakdown reactions: Sveconorwegian Province of SW Norway. Contributions to Mineralogy and Petrology, 132, 336-353. BINGEN, B., DEMAIFFE, D. & HERTOGEN, J. 1996. Redistribution of rare earth elements, thorium, and uranium over accessory minerals in the course of amphibolite to granulite facies metamorphism: The role of apatite and monazite in orthogneisses from southwest Norway. Geochimica et Cosmochimica Acta, 60, 1341-1354. BINGEN, B., BOVEN, A., PUNZALAN, L., WIJBRANS, J. R. & DEMAIFFE, D. 1998. Hornblende 4~ geochronology across terrane boundaries in the Sveconorwegian Province of S. Norway. Precambrian Research, 90, 159-185. BINGEN, B., AUSTRHEIM,H. & WHITEHOUSE,M. 2001. Ilmenite as a potential source for zirconium during high-grade metamorphism? Textural evidence from the Caledonides of Western Norway and implications for zircon geochronology. Journal of Petrology, 42, 355-375. BLACK, L. P., WILLIAMS, I. S. & COMPSTON, W. 1986. Four zircon ages from one rock: the history of a 3930-Ma-old granulite from Mount Sones, Enderby Land, Antarctica. Contributions to Mineralogy and Petrology, 94, 427-437. BOL, L. C. G. M., NIJLAND, I. G., SAUTER, P. C. C., JANSEN,J. B. H. & VALLEY,J. W. 1995. Preservation of premetamorphic oxygen and carbon isotopic trends in granulite facies marbles from Rogaland, southwest Norway. American Journal of Science, 295, 1179-1219. CARRINGTON, D. P. & HARLEY, S. L. 1995. The stability of osumilite in metapelitic granulites. Journal of Metamorphic Geology, 13, 613-625. CHERNIAK, D. J. & WATSON, E. B. 2000. Pb diffusion in zircon. Chemical Geology, 172, 5-24. DEGELING, H., EGGINS, S. & ELLIS, D. J. 2001. Zr budgets for metamorphic reactions, and the formation of zircon from garnet breakdown. Mineralogical Magazine, 65, 749-758. DODSON, M. H. 1973. Closure temperature in cooling geochronological and petrological systems. Contributions to Mineralogy and Petrology, 40, 259-274. DODSON, M. H. 1986. Closure profiles in cooling systems. Mathematical Science Forum, 7, 145-154. DUCHESNE, J. C., CARUBA, R. & IACCONI, P. 1987. Zircon in charnockitic rocks from Rogaland (Southwest Norway); petrogenetic implications. Lithos, 20, 357-368. FALKUM, T. & PEDERSEN, S. 1979. R b - S r age determination on the intrusive Precambrian Homme Granite and consequence for dating the last regional folding and metamorphism in the Flekkefjord region, South Norway. Norsk Geologisk Tidsskr!ft, 59, 59-65. FOSTER, G. L., KINNY, P., VANCE, D., PRINCE, C. P. & HARRIS, N. 2000. The significance of monazite U T h - P h age data in metamorphic assemblages; a combined study of monazite and garnet chronometry. Earth and Planetary Science Letters, 181, 327-340.

79

FRASER, G., ELLIS, D. & EGGINS, S. 1997. Zirconium abundance in granulite-facies minerals, with implications for zircon geochronology in highgrade rocks. Geology, 25, 607-610. FRIEND, C. R. L. & KINNY, P. D. 1995. New evidence for protolith ages of Lewisian granulites, northwest Scotland. Geology, 23, 1027-1030. GA,~L, G. & GORBATSCHEV,R. 1987. An outline of the precambrian evolution of the Baltic shield. Precambrian Research, 35, 15-52. HERMANS, G. A. E. M., TOBI, A. C., POORTER, R. P. E. ~; MAIJER, C. 1975. The high-grade metamorphic Precambrian of the Sirdal-Orsdal area, Rogaland/ Vest-Agder, S. W. Norway. Norges Geologiske UndersCkeise, Bulletin, 318, 51-74. HERMANS, G. A. E. M., HAKSTEGE, A. L., JANSEN, J. B. H. • POORTER, R. P. E. 1976. Sapphirine occurrence near Vikes5 in Rogaland, southwestern Norway. Norsk Geologisk Tidsskrifi, Supplement, 56, 397-412. HOLLAND, T. J. B., BABU, E. V. S. S. K. & WATERS, D. J. 1996. Phase relations of osumilite and dehydration melting in pelitic rocks: a simple thermodynamic model for the KFMASH system. Contributions to Mineralogy and Petrology, 124, 383-394. HOSKIN, P. W. & BLACK, L. P. 2000. Metamorphic zircon formation by solid-state recrystallization of protolith igneous zircon. Journal of Metamorphic Geology, 18, 423-439. JANSEN, J. B. H., BLOK, R. J. P., Dos, A. & SCHEELINGS, M. 1985. Geothermometry and geobarometry in Rogaland and preliminary results from the Datable area, S Norway. In: TOBI, A. C. & TOURET, J. L. R. (eds) The Deep Proterozoic Crust in the North Atlantic Provinces. Reidel, Dordrecht, 499-516. KARS, H., JANSEN, J. B. H., TOBI, A. C. & POORTER, R. P. E. 1980. The metapelitic rocks of the polymetamorphic Precambrian of Rogaland, SW Norway; Part II, Mineral relations between cordierite, hercynite and magnetite within the osumilite-in isograd. Contributions to Mineralogy and Petrology, 74, 235-244. KRONER, A., WILLIAMS, I. S., COMPSTON, W., BAUR, N., VITANAGE, P. W. & PERERA, L. R. K. 1987. Zircon ion microprobe dating of highgrade rocks in Sri Lanka. Journal of Geology, 95, 775-791. LEE, J. K. W., WILLIAMS, I. S. & ELLIS, D. J. 1997. Pb, U and Th diffusion in natural zircon. Nature, 390, 159-162. MAIJER, C. 1987. The metamorphic envelope of the Rogaland intrusive complex. In: MAIJER, C. & PADGET, P. (eds) The Geology of Southernmost Norway: an Excursion Guide. Norges Geologiske UndersCkelse, Special Publication, 1, 68-73. MAIJER, C. & PADGET, P. 1987. The Geology of Southernmost Norway: an Excursion Guide. A geological excursion guide with thematic articles prepared for the NATO Advanced Study Institute Meeting held at Moi, Norway, 1984. Norges Geologiske UndersOkelse, Special Publication, 1.

80

A. MOLLER ETAL.

MAIJER, C., ANDRIESSEN, P. A. M., HEBEDA, E. H., JANSEN, J. B. H. & VERSCnURE, R. H. 1981. Osumilite, an approximately 970 Ma old hightemperature index mineral of the granulite-facies metamorphism in Rogaland, SW Norway. Geologie en Mijnbouw, 611, 267-272. MOLLER, A., O'BRIEN, P. J., KENNEDY, A. & KRONER, A. 2002. Polyphase zircon in ultrahightemperature granulites (Rogaland, SW Norway): constraints for Pb diffusion in zircon. Journal of Metamorphic Geology, 20, 727-740. NASDALA, L., WENZEL, M., VAVRA, G., IRMER, G., WENZEL, T. & KOBER, B. 2001. Metamictisation of natural zircon: accumulation versus thermal annealing of radioactivity-induced damage. Contributions to Mineralogy and Petrology, 141, 125-144. NELSON, D. R. 1997. Compilation of SHRIMP U-Pb Zircon Geochronology Data, 1996. Geological Survey of Western Australia, Perth, Report, 1997/2. NIJLAND, T. G., MAIJER, C. • DE HAAS, G.-J. L. M. 1996. The Stokkafjell Troctolite, SW Norway: its bearing on the early P - T evolution of the Rogaland terrane. Neues Jahrbuch fiir Mineralogie, Abhandlungen, 1"71, 91-117. PAN, Y. 1997. Zircon- and monazite-forming metamorphic reactions at Manitouwadge, Ontario. The Canadian Mineralogist, 35, 105-118. PASTEELS, P., DEMAIFFE, D. & MICHOT, J. 1979. U Pb and R b - S r geochronology of the eastern part of the South Rogaland igneous complex, southern Norway. Lithos, 12, 199-208. PIDGEON, R. T. 1992. Recrystallisation of oscillatory zoned zircon: some geochronological implications. Contributions to Mineralogy and Petrology, 110, 463 -472. ROBERTS, M. P. & F~NGER, F. 1997. Do U - P b zircon ages from granulites reflect peak metamorphic conditions? Geology, 25, 319-322. RImATTO, D. 2002. Zircon trace element geochemistry: partitioning with garnet and the link between U - P b ages and metamorphism. Chemical Geology, 184, 123-138. RUBATTO, D. & HERMANN, J. 2003. Zircon formation during fluid circulation in eclogites (Monviso, Western Alps): implications for Zr and Hfbudget in subduction zones. Geochimica et Cosmochimica Acta, 67, 2173-2187. RUBATTO, D., WILLIAMS, I. S. & BUICK, |. S. 2001. Zircon and monazite response to prograde metamorphism in the Reynolds Range, central Australia. Contributions to Mineralogy and Petrology, 140, 458-468. SCHALTEGGER, U., FANNING, C. M., G(JNTHER, D., MAURIN, J. C., SCHULMANN, K. & GEBAUER, D. 1999. Growth, annealing and recrystallization of zircon and preservation of monazite in high-grade metamorphism: conventional and in-situ U-Pb isotope, cathodoluminescence and microchemical evidence. Contributions to Mineralogy and Petrology, 132, 186-201. SCH,~RER, U., WILMART, E. & DUCHESNE, J.-C. 1996. The short duration and anorogenic character of anorthosite magmatism: U - P b dating of the

Rogaland complex, Norway. Earth and Planetary Science Letters, 139, 335-350. TOBI, A. C., HERMANS, G. A. E. M., MAIJER, C. & JANSEN, J. B. H. 1985. Metamorphic zoning in the high-grade Proterozoic of Rogaland-Vest Agder, SW Norway. In: TOBI, A. C, & TOURET, J. L. R. (eds) The Deep Proterozoic Crust in the North Atlantic Provinces. Reidel, Dordrecht, 477-497. VANDEN AUWERA,J. & LONGHI,J. 1994. Experimental study of a jotunite (hypersthene monzodiorite); constraints on the parent magma composition and crystallization conditions (P, T, fO2) of the Bjerkreim-Sokndal layered intrusion (Norway). Contributions to Mineralogy and Petrology, 118, 60-78. VAVRA,G., GEBAUER,D., SCHMID,R. & COMPSTON,W. 1996. 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. Contributions to Mineralogy and Petrology, 122, 337-358. VERSCHURE, R. H., ANDRIESSEN,P. A. M., BOELRIJK, N. A. I. M., HEBEDA, E. H., MAIJER, C., PRIEM, H. N. A. & VERDURMEN, E. A. T. 1980. On the thermal stability of R b - S r and K - A r biotite systems; evidence from coexisting Sveconorwegian (c. 870 Ma) and Caledonian (c. 400 Ma) biotites in SW Norway. Contributions to Mineralogy and Petrology, 74, 245-252. WATSON, E. B. 1996. Dissolution, growth and survival of zircons during crustal fusion: kinetic principles, geological models and implications for isotopic inheritance. Transactions of the Royal Society of Edinburgh: Earth Sciences, 87, 43-56. WATSON, E. B. & LIANG, Y. 1995. A simple model for sector zoning in slowly grown crystals: Implications for growth rate and lattice diffusion, with emphasis on accessory minerals in crustal rocks. American Mineralogist, 80, 1179-1187. WESTPHAL, M. & SCHUMACHER, J. C. 1997. Links between the thermal model, reaction textures and mineral zoning in garnet-bearing granulites at Rogaland, southwestern Norway (abstract, DMG Meeting). European Journal of Mineralogy, 9, 384. WESTPHAL, M., BOSCHERT, S. ~: SCHUMACHER,J. C. 2003. UHT-metarnorphisrn and the role of magmatic heat sources at the Rogaland Anorthosite complex in southwestern Norway. Journal of Petrology, 44, 1145-1162. WIELENS, J. B. W., ANDRIESSEN,P. A. M., BOELR1JK, N. A. 1. M., HEBEDA, E. H., PRIEM, H. N. A., VERDURMEN, E. A. T. & VERSCHURE, R. H. 1980. Isotope geochronology in the high-grade metamorphic Precambrian of southwestern Norway: new data and reinterpretations. Norges geologiske UndersOkelse, 359, 1-30. WILLIAMS, I. S. 2001. Response of detrital zircon and monazite, and their U - P b isotopic systems, to regional metamorphism and host-rock partial melting, Cooma Complex, southeastern Australia. Australian Journal of Earth Sciences, 48, 557-580.

IN-SITU ZIRCON DATING OF UHT GRANULITE WILLIAMS, I. S., BUICK, I. S. & CARTWRIGHT,I. 1996. An extended episode of early Mesoproterozoic metamorphic fluid low in the Reynolds Range, central Australia. Journal of Metamorphic Geology, 14, 29-47. WILMART, E., DEMAIFFE, D. ~r DUCHESNE,J. C. 1989. Geochemical constraints on the genesis of the Tellnes ilmenite deposit, Southwest Norway. Economic Geology, 84, 1047-1056.

81

WILSON, J. R., P~DERSEN, S., BERTHELSEN, C. R. & JAKOBSEN, B. M. 1977. New light on the Precambrian Holum granite, South Norway. Norsk Geologisk Tidsskrift, 57, 347-360. ZHOU, X. Q., BINGEN,B., DEMAIFFE,D., LIt~GEO1S,J. P., HERTOGEN, J., WEIS, D. & MICHOT, J. 1995. The 1160 Ma Hidderskog meta-charnockite; implications of this A-type pluton for the Sveconorwegian Belt in Vest Agder (SW Norway). Lithos, 36, 51-66.