granulite and the granulite facies

granulite and the granulite facies The term ‘granulite’ was formerly used to refer to a wide range of siliceous metamorphic rocks with a medium- to coarse-grained granoblastic texture (one with equidimensional grains) and lacking a marked schistosity. Since the 1970s, however, it has come to be used exclusively for silicate rocks that have experienced granulite-facies metamorphism. Thus the former use of the term ‘granulite’ for lower-grade psammitic rocks is now obsolete. True granulites are of very varied composition, and their lack of a prominent schistosity arises because micas are scarce or absent in the granulite facies, reflecting the intense dehydration that occurs in this, the metamorphic facies formed at the highest temperatures (see metamorphism, metamorphic facies, and metamorphic rocks).

There has been much controversy over the origin of granulites, and in particular the cause of their intense dehydration. One, probably always predominant, view, has been that granulites are residual rocks that have experienced at least incipient partial melting. Not only has any pore water originally present dissolved in a melt phase, but the temperatures of the granulite facies are high enough for melting to proceed directly by breakdown of hydrous minerals such as biotite, providing a water component for the melt. Granulites can display reaction textures compatible with such a process and appear migmatitic on a larger scale. On the other hand, not all granulites are obviously migmatitic, and some do not show the depletion in incompatible elements that the melting model would indicate. The recognition by Jaques Touret in the 1970s that fluid inclusions in granulite are commonly dominated by carbon dioxide led to an alternative model for granulite formation. This model invoked streaming of carbon dioxide from the mantle through the lower crust, driving dehydration through lowering the partial pressure of water by dilution.

In the course of the ensuing debate, many points of importance have emerged. One of the arguments in favour of the carbon dioxide streaming model was that the geothermometer temperatures for granulite often overlap with those of amphibolite-facies rocks. However, it is now recognized that this is largely the result of cation exchange by diffusion, which means that simple cation-exchange geothermometers often preserve a blocking temperature, rather than the true temperature of formation, of rocks of very high grade. Indeed, further work has identified a number of distinctive assemblages that require exceptionally high temperatures to be stabilized (e.g. sapphirine with quartz) and appear to typify a distinct type of ultra-high-temperature granulite. On the other hand, another compelling argument for the carbon dioxide model is that some granulites (generally charnockites, i.e. rocks with orthopyroxene and potassium-feldspar) can be seen to have been produced very locally from apparently lower-grade wall rocks in certain localities, forming zones of replacement that may be just a few centimetres in width, usually flanking veins. Such charnockites formed in situ may contain much more carbon dioxide in fluid inclusions that the adjacent parent rocks, and this is a rather effective rebuttal to the claim that the carbon dioxide fluids of many granulites arise because water in an original mixed H₂O–CO₂ fluid is selectively dissolved in melt.

At the time of writing, there is increasing consensus about granulites, although many points remain to be resolved. The distinction between granulites produced during slow cooling of deeply emplaced plutons and those formed by progressive metamorphism and dehydration of lower-grade rocks is far from straightforward, and the former mechanism accounts for the apparently anomalous lack of melting features of certain bodies. Likewise, release of carbon dioxide from crystallizing magmas is a much more potent means of introducing it more or less pervasively into lower crustal rocks than long-range migration as a separate fluid, because of its poor wetting properties. This process, and the presence of marbles in some CO₂-rich granulite terranes, probably accounts for the well-documented examples of carbon dioxide involvement in granulites. Most examples of extensive prograde granulite-facies metamorphism appear to be clearly related to melting, and efficient removal of melt accounts for the observation that migmatite features are usually sparse in comparison with those of amphibolite-facies migmatites, from which the melt was unable to migrate without freezing. In some instances, there is clear evidence for the accompanying emplacement of mantle-derived mafic magmas, providing a source for the anomalous heating.

Bruce W. D. Yardley

Bibliography

Harley, S. L. (1998) On the occurrence and characterization of ultrahigh-temperature crustal metamorphism. In Treloar, P. J. and O'Brien, P. J. (eds) What drives metamorphism and metamorphic reactions?, pp. 81–107. Geological Society, Special Publication 138.
Touret, J. L. (ed.) (1985) The deep Proterozoic crust. Reidel, Dordrecht.