5.3.4.6 The Migmatitic Zone

The Migmatitic zone has recorded the highest grade metamorphic conditions in the studied area. This zone is characterized by pervasive anatexis of the HHCS rocks and represents the most likely source for the leucogranites. The term migmatite is here used in its broad sense, a more accurate name for these rocks would be «diatexite» as they have undergone extensive partial melting and the leucosomes are volumetrically dominant over the restitic elements. Migmatisation is here so intense that it is not possible to establish whether the protolith of these migmatites are the metasediments of the Phe formation or the Cambro-Ordovician granites. The leucosomes are essentially composed of K-Feldspar and quartz with minor biotite, muscovite, plagioclase, garnet and sillimanite (Fig 5.17). The melanosomes are predominantly formed by biotite, quartz and sillimanite with minor plagioclase, garnet, tourmaline, apatite and monazite. Some muscovite grew late in replacement of sillimanite and biotite during retrograde metamorphism.

The mineral assemblages of the metapelitic (or quartzo-feldspathic) sequence give little information on the peak metamorphic condition that were reached in the migmatitic zone as they have re-equilibrated at low P/lowT during retrograde metamorphism. The virtual absence of muscovite, and the large quantities of leucogranitic melt that escaped from the migmatitic zone to form the overlying intrusion complex clearly indicate, however, that the metamorphic grade was sufficient to allow for fluid-absent melting.

Decimetric boudins of metabasic rocks sometimes occur within the quartzo-feldspathic migmatites. The core of these boudins preserves indications on the peak metamorphic condition of the migmatitic zone. These rocks show a typical granulitic texture and are formed by the assemblage

garnet + clinopyroxene + amphibole + plagioclase + quartz + titanite. (Fig 5.18)

The clinopyroxenes belong to the augite group, are Calcium-rich and have the composition of salites. The amphiboles are hornblendes (s.l.) and more exactly ferropargasites. The garnets are essentially composed of almandine and grossular and the plagioclases are almost pure anorthites. This assemblage is formed through continuous reactions resulting in the modal increase of garnet and clinopyroxene at the expense of hornblende and plagioclase. This is testified by inclusions of the latter minerals in the former ones. A single relict biotite grain was also preserved in a garnet porphyroblast.

The granulitic metabasites also show evidences of retrograde metamorphism, first with the development of hornblende-rich rinds at the contact between the metabasic lenses and the surrounding quartzo-feldspathic migmatites and secondly, with the symplectic growth of hornblende and quartz at the contact between garnets and clinopyroxenes.

The mineral assemblage and the texture of the metabasic rocks typically forms along the kyanite geotherm at temperatures above 700°C (Bucher and Frey, 1994) and can be found both in upper amphibolite or high-pressure granulite facies metabasites. A clear indication that granulite facies conditions were reached would be given by the presence of orthopyroxene which appears in metabasic rocks above ~ 800°C (Spear 1993). Orthopyroxene does however not form in high-pressure granulites, even above 800°C.

As muscovite is virtually absent from the metapelitic migmatites, but biotite is still present and shows no sign of destabilisation, the metamorphic grade must have been sufficient to allow for the breakdown of muscovite but not of biotite. The break-down of muscovite occurs through the reaction:

muscovite + plagioclase + quartz = K-feldspar + Als + melt

As this reaction has a relatively low dP/dT slope, the temperatures at which the break-down of muscovite occurs are highly dependant on the pressure. Three such reaction curves are shown in figure 5.19. The first curve is a theoretical curve, obtained for the end-members muscovite + albite + quartz and predicts temperatures between 700° and 770°C for pressures between 7 and 12 Kbars (Petö, 1976). These temperatures are lower than those shown by the two other curves obtained by Patiño Douce and Harris (1998) through experimental melting of two metapelitic samples from the HHCS (one of them from Zanskar) at different P/T conditions. The natural starting material differs from the model end-member assemblage by the presence of Calcium in the plagioclase and Fe, Mg, Ti and F in the muscovite. The effect of these elements is to extend the thermal stability field of the assemblage Ms + Pl + Qtz of 50° to 80 °C relative to that of the equivalent end-member assemblage. We consider these values obtained through melt experiments on natural samples to be more reliable than the theoretical values obtained through thermobarometric modelling on end-member composition.

We have also represented in figure 5.19 the dehydration-melting curves for biotite as calculated by Vilzeuf and Montiel (1994) and Patino Douce & Beard (1996) for different Mg numbers. The minimal temperature for the break-down of biotite is 800°C and is very little dependent on pressure variations. As these reaction curves are steeper than those for the breakdown of muscovite, these two reaction curves intersect at high-pressure. The minimal pressure for this cross-over to occur is 9 kbar, above which biotite reacts before muscovite to produce melts (above 800°). As the destabilisation of biotites was not observed in our samples, the migmatitic zone should not have exceeded this temperature.

In short, both the metabasic and the metapelitic assemblages constrain the peak temperatures reached by the migmatitic zone to values slightly below 800°C. The peak pressure of the migmatitic zone can be estimated to have reached values above 9 kbar, first because of the Grt ± Cpx + Hbl + Pl + Qtz assemblage of the metabasites and secondly, because the structurally overlying kyanite zone already experienced such pressures (Staurolite-out reaction along the Ky-geotherm occurs ~ 9 Kbar).

At temperatures below 800°C and pressures above 9 kbar, it is quite unlikely that the migmatitic zone produced leucogranitic melts through dehydration melting of muscovite. On the other hand, vapour-present melting is not possible because this would result in melts of trondhjemitic composition and the observed intrusions have a leucogranitic composition.

Thus, to explain the widespread anatexis, the incongruent melting of muscovite and the production of leucogranitic melts that occurred in the migmatitic zone, we infer that these rocks must have undergone isothermal decompression such as to cross the dehydration curve of muscovite. Isothermal decompression paths (such as the one shown in figure 5.19) intersect the relatively flat dehydration melting solidus of muscovite schists, triggering anatexis, which continues until relatively shallow depths. In contrast, the steep solidi for dehydration melting of biotite are almost parallel to isothermal decompression paths. This means that, if biotite schists remain unmolten when buried deeply in an orogen, they are also unlikely to melt during ensuing decompression.

We will see later that other independent arguments sustain the interpretation of peak metamorphic conditions followed by retrograde isothermal decompression.