Trace-element zoning in a metamorphic garnet D. D. Hickmott, N. Shimizu Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology Cambridge, Massachusetts 02139 F. S. Spear Department of Geology, Rensselaer Polytechnic Institute Troy, New York 12181 J. Selverstone Department of Earth and Planetary Sciences, Harvard University Cambridge, Massachusetts 02138 ABSTRACT Trace-element zoning has been measured in an amphibolite garnet from the Tauern window, Austria, by using an ion microprobe. Humps in the zoning profiles of Na, Sc, V, Y, and the heavy rare-earth elements mark a period of open-system behavior. These humps corre- spond to a part of the major-element zoning profile that is interpreted as a P-T reversal. The source of the mass excesses of these elements remains ambiguous: they were derived either on a thin-section scale by the breakdown of trace-element-enriched refractory minerals or externally from unusual trace-element-enriched fluids. P-T paths de- termined from garnet zoning may require modification if open-system behavior is important during garnet growth. INTRODUCTION The major-element zoning of metamorphic garnets has been used to investigate variations in the P-T conditions in metamorphic terrains prior to the final equilibration of matrix minerals and garnet rims (Tracy et al., 1976; Spear and Selverstone, 1983). The chemical signatures of early metamorphism are assumed to be preserved in garnet cores up to the amphibolite facies because chemical diffusion in garnet is slow at low to moderate metamorphic temperatures (Woodsworth, 1977; Cygan and Lasaga, 1985; Freer, 1981). Ubiquity of garnet in a wide range of bulk compositions contributes to its importance as a probe of metamorphic processes. Accurate interpretation of chemical zoning in garnet leads to significantly improved understanding of P-T paths of metamorphism and hence of the tectonics of mountain belts; therefore, it is essential to develop techniques that elucidate metamorphic processes during garnet growth. In this article we examine whether trace-element variation during garnet growth (trace-element zoning) measured with an ion microprobe can provide additional constraints on the processes affecting major- element zoning profiles. We use a well-studied garnet from the Tauern window, Austria (Selverstone et al., 1984; Selverstone, 1985), as an exam- ple. The ion-microprobe techniques used to investigate zoning in the more abundant trace elements (i.e., Sc, Ti, V, Cr, Y) are similar to those de- scribed elsewhere (Shimizu and Hart, 1982; Shimizu et al., 1978). Rare-earth elements (REE) were analyzed in situ by using a slightly modi- fied technique, outlined below. This example demonstrates that trace ele- ments are sensitive indicators of the occurrence of open-system behavior during garnet growth. Elements such as Sc, Y, and the heavy REE, all normally considered to be immobile during prograde metamorphism, ex- hibit open-system behavior on thin-section scale. ION PROBE ANALYSIS (REE) For REE analysis a primary beam of negatively charged oxygen ions is focused to a spot 50-60 ixm in diameter (compared to the 5-/jm spot used for the more abundant trace elements), and secondary ions are col- lected in the electron multiplier of a Cameca IMS-3F ion microprobe. Measurements of isotope ratios of REE as a function of secondary high- voltage offset reveals that a modest energy filtering (i.e., 30-40 electron volts [eV] as opposed to the 80-100 eV used for abundant trace elements) is sufficient to suppress the majority of molecular ion interferences for heavy REE-enriched garnets. Unzoned pyrope-rich garnets from kimber- lites and peridotite xenoliths in kimberlites are used as standards to verify that linear relations between relative intensity (REE/Si) and concentra- tions are obtainable at 30-40 V offsets. Replicate analyses of a pyrope-rich working standard from Monastery kimberlite suggest that REE/Si ratios are reproducible within less than 10% (one sigma) for most elements. GEOLOGIC SETTING The sample investigated, FH-1M, is an amphibolite from the Lower Schieferhulle of the Tauern window, Austria. Details of geologic setting, phase relations, petrography, and P-T conditions of metamorphism can be found elsewhere (Selverstone et al., 1984; Selverstone, 1985). The sample contains the assemblage hornblende + kyanite + staurolite + garnet + biotite + chlorite + epidote + plagioclase + ankerite + quartz + rutile + ilmenite + late-stage margarite. The garnets are large (2-4 mm) and euhe- dral, and they contain inclusions of plagioclase, epidote, ankerite, ilmenite, and rutile. Garnet growth occurred by continuous reaction in a constant, divariant assemblage that included chlorite, epidote, and ankerite as reac- tants. On the basis of geothermometry, geobarometry, and thermodynamic modeling of garnet zoning, Selverstone et al. (1984) suggested that garnets grew during decompression to rim equilibration conditions of 550 °C and 7 kbar. They also stated, on the basis of major-element zoning, that a P-T reversal of 300-400 bar occurred during garnet growth as a result of a brief interval of burial during the uplift-dominated P-T path. GARNET ZONING The garnet studied is concentrically zoned in both major and trace elements. It is imperative to understand the interrelation between major- and trace-element variations because accurate interpretation of metamor- phic processes is facilitated by a sound knowledge of the development of major-element zoning. The traverse portrayed in Figure 1 extends 1.5 mm from garnet core to rim and is identical in its major features to zoning profiles in two other garnets from the same sample. (A detailed discussion of intrasample trace-element variations and trace-element zoning in other minerals from FH-1M will be presented in a subsequent publication.) Gaps occur in the traverse where inclusions are intersected by the primary beam of the ion microprobe. Major-element zoning (Fe, Ca, Mn, Mg) mirrors that presented by Selverstone et al. (1984; see their Fig. 11). As growth occurs from the core to point 4 (Fig. 1), almandine and pyrope components increase, whereas grossular and spessartine decrease. Selver- stone et al. (1984) interpreted this as an initial period of garnet growth GEOLOGY, v. 15, p. 573-576, June 1987 573 on December 11, 2014 geology.gsapubs.org Downloaded from