Daniel Harlov (dharlov@gfz-potsdam.de) & Hans-Jürgen Förster (forhj@gfz-potsdam.de)
GeoForshungsZentrum, Telegrafenberg, D-14473 Potsdam, F.R. Germany
The REE geochemistry of metamorphic rocks is directly influenced by the chemistry of apatite, monazite and xenotime. However, the geochemical and crystallochemical relationship between these three phosphate minerals is not well understood (Pan et al. 1993). For example, in chlorapatites in a pegmatoid from the Ødegårdens Verk, Bamble Sector, SE Norway (Leiftink et al. 1994), numerous monazite and xenotime grains are observed in areas where the chlorapatite was metasomatically altered to a F-poor hydroxyapatite under amphibolite facies conditions. In contrast, regions of unaltered chlorapatite in the same crystal do not contain monazite or xenotime inclusions. This suggests that introduction of OH into the chlorapatite structure induced the formation of these REE phosphates. Two experiments were performed in the piston cylinder. A non-metasomatised sample of the chlorapatite was crushed into 50-200 mm size grains. In the first experiment, chlorapatite (20 mg) and H2O (5 mg) were placed in a Pt capsule and the capsule welded shut. In the second, a 50/50 molar H2O/F fluid (6 mg) was used with CaF2 as the F source. The two capsules were placed in a CaF2 cell with the NiCr thermocouple located between them and taken up to 900°C and 1 GPa for one week. In the H2O-chlorapatite experiment, numerous co-existing monazite and xenotime grains volunteered and the chlorapatite was altered to a hydroxyapatite with a minor Cl component. In the H2O-F-chlorapatite experiment, no monazite or xenotime grains volunteered and the chlorapatite was altered to a fluor-apatite with a minor OH and Cl component. These results suggest that, in the H2O-chlorapatite experiment, the REE in the apatite went into nucleating monazite and xenotime within the body of the apatite crystal structure whereas in the H2O-F-chlorapatite experiment the REE complexed with the F and were subsequently leached out of the apatite into the fluid.
Pan Y, Fleet ME, and Macrae ND, Mineral Mag, 57, 697-707, (1993).
Leiftink DJ, Nijland TG, and Maijer C, Can Mineral, 32, 149-158, (1994).
Daniel Harlov (dharlov@gfz-potsdam.de) & Stefan Melzer (smelzer@gfz-potsdam.de)
GeoForschungsZentrum, Telegrafenberg, D-14473 Potsdam, F.R. Germany
Orthopyroxene-bearing granulite facies terranes (charnockites) are often characterized by extreme depletion in Rb (e.g. Hansen et al. 1995) despite abundant high Ti-biotite which should act as a natural sink for Rb. Some workers have speculated that low H2O activity supercritical (K,Na)Cl-rich brines, originating from mantle derived basaltic underplating, could play a significant role in charnockite formation (Newton et al. 1998) as both a source of heat transport and fluid. If so they could also play a role in the depletion of Rb in these terranes. However, recently some workers (Melzer and Wunder 1999) have demonstrated experimentally that for high H2O activity (K,Rb)Cl brines, (> 0.9), Rb is strongly partitioned into the phlogopite. Phlogopite synthesis experiments were performed at 800°C and 2 GPa for concentrations of Rb less than 10% relative to K in the presence of a supercritical (K,Rb)Cl brine with an approximate Cl molality of about 100 (Cl/(H2O+Cl) = 0.7) and an H2O activity of 0.09 (cf. Aranovich and Newton 1997). The solid run products consisted of phlogopite along with smaller amounts of K-feldspar and very minor quartz. Rb fractionated strongly into the phlogopite, whereas, for the K-feldspar, Rb was distributed almost equally distributed between the feldspar and fluid. The derived exchange coefficients, KD(Rb-K) phl-fluid and KD(Rb-K) fsp-fluid (Henry's law), are 1.88±0.12 and 0.86±0.14, respectively. Because these charnockites are highly depleted in Rb, our experimental results suggest that the original fluid responsible for the dehydration of these rocks must have already had an extremely low Rb concentration relative to K. They also suggest that the fluid could not have been depleted in Rb during the dehydration process but rather must have been enriched during the breakdown of the amphiboles to pyroxenes with this excess Rb partitioned into the biotite-phlogopites further up the rock column away from the source basaltic underplating.
Hansen EC, Newton RC, Janardhan AS, and Lindenberg S, J Geol, 103, 629-651, (1995).
Newton RC, Aranovich LYa, Hansen EC, and Vandenheuvel BA, Precamb Res, 91, 41-63, (1998).
Melzer Sand Wunder B, EOS, 80, 361, (1999).
Frédéric Hatert (fhatert@ulg.ac.be)1, André-Mathieu Fransolet (amfransolet@ulg.ac.be)1, Fernande Grandjean (fgrandjean@ulg.ac.be)2 & Gary J. Long3
1 Université de Liège, Laboratoire de Minéralogie, B.18, Sart Tilman, B-4000 Liège, Belgium
2 Université de Liège, Institut de Physique, B.5, Sart Tilman, B-4000 Liège, Belgium
3 Department of Chemistry, University of Missouri-Rolla, Rolla, MO 65409-0010, USA
The chemical compositions of alluaudites occurring in Li-rich granitic pegmatites generally range between two ideal endmembers, Na2Mn(Fe2+Fe3+)(PO4)3 and NaMnFe3+2(PO4)3. Moreover, the calculation of the structural formulae of these minerals shows that Mn also shares the A(1) site with Na, or the M(2) positions with Fe (Fransolet et al., 1994). It is the reason why we first envisage the synthesis of phosphates in the Na2O-MnO-Fe2O3-P2O5 system.
The phosphates have been synthesized by solid state reaction in air, between 800 and 950°C. The central part of the Na-Mn-Fe3+ ternary diagram is occupied by compounds exhibiting the alluaudite structure. These compounds form extensive solid solutions covering 25% of the diagram surface. Other phases have also been identified: NaMn4(PO4)3 (fillowite structure), (Mn,Fe)3(PO4)2 (graftonite structure), FePO4 (berlinite structure), and Na3Fe3+2(PO4)3 (nasicon structure). Interestingly, it must already be pointed out that associations alluaudite + fillowite, and alluaudite + graftonite are known to occur in granitic pegmatites. The unit-cell parameters of 14 alluaudite-like samples have been calculated and compared with the unit-cell parameters of natural alluaudites particularly enriched in Fe3+. Rietveld refinements of selected samples, Na2Mn2Fe3+(PO4)3, Na1.5Mn1.5Fe3+1.5(PO4)3, and Mn2.25Fe3+1.5(PO4)3, have been performed in order to tackle the cation distribution between the A(2)', A(1), M(1) and M(2) crystallographic sites. The 57Fe-Mössbauer spectra indicate the presence of small amounts of Fe2+ in the Fe-rich alluaudites-like compounds, confirming the wet chemical analyses realised by Hatert et al. (in press). This method also sheds some light on the distribution of Fe atoms between the M(1) and M(2) sites.
Fransolet AM, Antenucci D, Fontan F & Keller P, I.M.A. 16th General Meeting, Abstract volume, 125-126, (1994).
Hatert F, Keller P, Lissner F, Antenucci D & Fransolet AM, Eur. J. Mineral., in press
Index of EMPG VIII Volume
Further EMPG VIII Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page