Strontium isotope heterogeneities in amphibolite facies, banded metasediments

Abstract

Abstract provided by author:

Within the framework of these data, Rb-Sr small-domain isotope analyses were carried out from banded metasediments of the Kuiseb Formation (1) to investigate the applicability of the method to the dating of metamorphic processes and (2) to study the isotopic homogeneity of these metasediments. Six samples (KB167, KB117, KB307, KB421, KB115, KB103) were selected which are characterized by a compositional layering of either sedimentary or metamorphic origin. The samples differ in their structural overprint as well as in their metamorphic grade. Rockprofiles between 4 and 16 cm in length were cut perpendicular to the banding of the samples and subsequently divided into mm thick slabs. Apart from whole-rock analyses of each slab, mineral fractions (biotites, muscovites, plagioclase-quartz fractions, apatites) were analysed from individual slabs

The obtained element distribution patterns of Rb and Sr within the metamorphically banded samples show a characteristic positive correlation of Rb and Sr both within the phyllosilicate-rich (P-) and quartz-rich (Q-) domains of the banding which suggests that quartz has been removed from the P-domains during prograde metamorphism

Put into isochron diagrams, the isotope data of all six rock profiles align along straight lines (MSWD 0. 5-3. 2). The corresponding whole-rock age data range from 449 ± 20 Ma (KB117) to 509 ± 17 Ma (KB307). In contrast, the investigated biotite fractions show a more limited range of mineral-whole-rock ages (485 ± 10 to 498 ± 10 Ma), despite the composite character of the fractions and the structurally and metamorphically different history of the samples. Plagioclase-whole-rock pairs yielded between 476 ± 15 and 497 ± 14 Ma. K-Ar and 40Ar- 39Ar data obtained from selected biotite and muscovite fractions are consistent with the respective Rb-Sr data. Compared to the U-Pb monazite ages, all of the Rb-Sr and the K-Ar/40 Ar-39 Ar mineral data clearly fall below the supposed peak of metamorphism. The pronounced uniformity of the mineral data (from the metasediments as well as Rb-Sr data for mineral fractions from the Donkerhuk Granite) points towards a regional event or regionally significant processes responsible for the isotopic equilibration of the minerals after the peak of regional metamorphism. The resetting of the isotopic systems is attributed either to the temperature-induced selfdiffusion which continues during regional cooling of the rocks or to a late deformational and/or high temperature retrograde event which led to a partial recrystallization of the minerals

With regard to the relation between the mineral and whole-rock isotope data, the rock-profiles KB117 and KB115 appear to be most important for an evaluation of the significance of the apparent whole-rock ages. In these samples, the regression of the whole-rock data from the individual slabs of the rock profiles revealed younger ages (449 ± 20 Ma, MSWD 0. 5; 479 ± 11 Ma, MSWD 0. 7) than the mineral-whole-rock pairs from individual slabs. The significance of the calculated whole-rock ages is discussed by means of profile diagrams which illustrate the spatial and the computed temporal development of the 87Sr/86Sr for the individual slabs. This form of diagram thus allows a geometrical control over the isotope distribution within the rock profiles

The most striking feature of the spatial distribution and the computed temporal development of the 87Sr/86Sr isotope ratios is that for most rock profiles no homogeneous distributions of the 87Sr/86Sr isotope ratios can be calculated for the time of the apparent whole-rock ages. Instead, the isotope distribution patterns are characterized by isotopic heterogeneities. It is important that the isotopic heterogeneities within the distribution patterns cannot have evolved from an undisturbed, previously equilibrated isotope distribution. However, partial isotope equilibria apparently exist within the rock profiles. In accordance with previous studies, the isotope distribution patterns which show the least variation of the isotope ratios, may be interpreted as representing a metamorphic event leading to a partial equilibration of the isotope ratios although the regression lines do not represent true isochrons. This dating by means of disequilibrium distributions might be applied to the samples KB167, KB307, KB421 and KB103

The two rock profiles KB117 and KB115 show comparatively homogeneous distribution patterns of their Sr/ Sr ratios for the time of the calculated whole-rock ages. In both samples, plagioclase-quartz concentrates were analysed which demonstrate that there is no homogeneity of isotopes between the plagioclases and the respective whole rocks at the time of the apparent whole-rock equilibrium. In three slabs of the rock profile KB115, the analysed plagioclase-quartz fractions rather have a homogeneous isotope distribution with the respective whole rock at the calculated biotite-whole-rock ages. Furthermore, it could be shown for the rock profile KB115 that the homogenization of the plagioclases with their respective whole rocks occurred discontemporaneously within different slabs. These apparent isotope heterogeneities strongly suggest that the apparent whole-rock ages obtained from the two samples do not denote the timing of whole-rock isotope homogenization, since the major constituents of the rocks were not in isotope equilibrium with their respective whole rocks at that time

Two theories may be discussed to explain the quite uncommon relationship between the mineral and whole-rock ages within the two rock profiles. The first model is based on the assumption that no isotope exchange occurred between the whole rocks of the individual slabs, but between the minerals within the individual slabs. It seems unlikely, however, to postulate that no isotope exchange at all occurred between the mm- to cm-thick slabs of the investigated amphibolite facies metasediments. The second model is based on the assumption that an equilibration only between the plagioclases continued, whereas the homogenization of the plagioclases with the respective biotites and whole rocks within and between the contiguous slabs ceased at an earlier time. This model would imply that plagioclase closes at a later stage than biotite which, however, has not yet been observed in other studies. Thus both of these models do not seem to be compelling for the interpretation of the isotope systematics of the two rock profiles. A third possible mode of explanation would encounter an open system behaviour of the whole rocks. It might be assumed that the interaction of the rock with fluid phases, for example by an addition of Sr in an interstitial phase, can alter either the whole-rock isotopic composition or the concentrations of the elements. The minerals would remain closed systems. In practice, there are arguments pro and contra an open system behaviour of the rocks. The occurrence of certain retrograde mineral phases confirms the presence of a fluid phase in the late stages of metamorphism. Although the genesis of the apparent "isochrons" is not yet clear, it can undoubtedly be shown that they do not bear geologic relevance. The present study showed that apparent "isochron" ages derived from Rb-Sr whole-rock small-domain analyses of metasedimentary rocks can lead to geologically meaningless ages which casts doubts on the applicabilty of the Rb-Sr small-domain method as a tool in dating metamorphic events and on the significance of previous Rb-Sr whole-rock data from metamorphic rocks

The evolution of the Rb-Sr isotopic systems most likely is governed by a combination of processes which partly will operate in a contrary way. The different Rb-Sr ratios will permanently enforce the development of gradients within the 87Sr/86Sr isotopic compositions between compositionally different rock volumes. The homogenization of strontium isotopes, on the other hand, may be inferred to result from a continuous temperature induced selfdiffusion and more episodically increases in the rate and extent of isotope exchange due to deformational processes and metamorphic reactions with a net transport of Rb and Sr. It is also assumed that the change in the physical conditions of metamorphism and the change in the mineralogy of the rocks will result in changing rates and mechanisms of isotope homogenization. Selfdiffusion may become the most important process of isotope exchange when the main deformation and also metamorphic reactions have ceased. The existence of the strontium isotope heterogeneities within the isotope distribution patterns of the investigated microprofiles suggests that the extent of strontium diffusion during the final stages of the Pan-African metamorphism did not exceed a few millimetres or centimetres