X-site Vacant Al-Tourmaline: A New Synthetic End Member

Ulrich Wodara & Werner Schreyer (werner.schreyer@ruhr-uni-bochum.de)

Institut für Geologie, Mineralogie und Geophysik, Ruhr-Universität-Bochum, 44780 Bochum, Germany

For the three Na-bearing tourmaline end members schorl, dravite and elbaite, X-site vacant analogues occurring in nature are the minerals foitite, magnesiofoitite and rossmanite, respectively. No analogue exists for olenite, NaAl₃Al₆[Si₆O₁₈](BO₃)₃O₃ (OH). We were able to synthesize such "Na-free olenite", here called Al tourmaline, in the system Al₂O₃-B₂O₃-SiO₂-H₂O over the PT-range 4-40 kbar, 450-700°C. Starting materials were AlSi gels or oxide mixes with, or without, excess amounts of boron in the form of H₃BO₃ or B₂O₃ for several hypothetical tourmaline formulae. Highest yields up to about 85% were obtained from a composition 3Al₂O₃·2B₂O₃·4SiO₂ plus excess water at 20-30 kbar, 600°C in addition to dumortierite and quartz, rarely also jeremejevite-OH. Powder X-ray diffraction diagrams show all peaks characteristic for tourmaline, but shifted toward higher 2 theta angles. The lattice parameters a = 15.690(6) Å, c = 7.039(5) Å, V = 1500(1) ų are smaller than for natural olenite, but larger than for synthetic olenite with excess boron (Schreyer et al., Eur.J.Mineral, in press). One run product from a more aluminous oxide mix yielded single crystals with lengths up to 50 micrometer, which could be analyzed by electron microprobe including boron. Water had to be calculated by difference. The formula resulting from 20 individual analyses is: (vac_0.96Na_0.04)Al₃(Al_5.83Si_0.20)[Si_4.51B_1.49O₁₈](BO₃)₃(OH)_3.21O_0.79. Na is probably a contamination from the NaCl pressure cell. Thus, like in synthetic olenite, there is excess boron placed on the tetrahedral site. Because (Si+B)>9.0, a small amount of Si was tentatively allocated to octahedral sites, which would otherwise show vacancies. Additional experimentation is required to test as to whether or not the boron excess is a necessary property. There may indeed be a range of Al-tourmaline solid solutions in the system. Nevertheless, an end-member formula for the new X-site vacant tourmaline can be defined as (vac)Al₃Al₆[Si₆O₁₈](BO₃)₃(OH)₂O₂, related to olenite by (vac)HNa_-1.

Fe-Mg Solid Solution of Sursassite (Mg,Fe)₅Al₅Si₆O₂₁(OH)₇)

Bernd Wunder (wunder@gfz-potsdam.de) & Matthias Gottschalk

GeoForschungsZentrum Potsdam, Germany

In high-pressure experiments in the system MgO-Al₂O₃-SiO₂-H₂O (Schreyer et al. (1986, 1991) synthesized at 50 kbar, 700°C a new phase of the composition Mg₅Al₅Si₆O₂₁(OH)₇. Assuming to have a pumpellyite structure, Schreyer et al. (1986, 1991) named this phase MgMgAl-pumpellyite. Artioli et al. (1999) found for this phase a pumpellyite-type structure having a certain amount of sursassite domains. However, recently Gottschalk et al. (2000) determined the structure of this phase by the Rietveld-method and found it to be isostructural to sursassite. This was confirmed by HRTEM-investigations. This phase is therefore the Mg-analogue of Mn²⁺MgAl₅Si₆O₂₁(OH)₇, firstly synthesized by Reinecke et al. (1988) and therefore was renamed to MgMgAl-sursassite by Gottschalk et al. (2000). Due to its high-pressure, low-temperature stability determined by Fockenberg (1998), this phase is proposed to be an important water reservoir in cold subduction zones. A still open problem is that of a potential solid-solution of MgMgAl-sursassite. In the light of its sursassite-structure a theoretical solid-solutions from Mg₅Al₅Si₆O₂₁(OH)₇ towards a hypothetical end-member Fe₅Al₅Si₆O₂₁(OH)₇ can be proposed, because the ionic-radius of Fe²⁺ lies between those of Mg and Mn²⁺. Based on such considerations we started some synthesis experiments within the system FeO-MgO-Al₂O₃-SiO₂-H₂O, which results are presented here. Starting materials were gels with varying Mg/Fe-ratios on the join Mg₅Al₅Si₆O₂₁(OH)₇-Fe₅Al₅Si₆O₂₁(OH)₇. Buffered by Fe/FeO, sursassite formed in all runs at 40 kbar, 650°C, 3 days, indicating a complete Fe,Mg-solid-solution. Minor amounts (~1 wt.%) of chloritoid and chlorite were detected by the Rietveld-analysis. The cell-dimensions of Fe₅Al₅Si₆O₂₁(OH)₇ were determined to a=8.6355(6)Å, b=5.7417(4)Å, c=9.7137(7)Å, b=108.72(1)°, V=456.16(7)ų. Incorporation of Fe²⁺ into MgMgAl-sursassite can be expected to lower its minimum pressure stability limit (34 kbar according Fockenberg, 1998) and thus make it a potentially still more important receptacle for water transport in regions of more moderate P,T-conditions. Experiments on the lower P-limit of the Fe-endmember are in progress.

Schreyer W, Maresch WV, Medenbach O & Baller T, Nature, 321, 510-511, (1986).

Schreyer W, Maresch WV & Baller T, In: Progress in metamorphic and magmatic petrology, ed. Perchuk LL, Cambridge University Press, U. K, 47-64, (1991).

Artioli G, Fumagalli P & Poli S, American Mineralogist, 84, 1906-1914, (1999).

Gottschalk M, Fockenberg T, Grevel K-D, Wunder B, Wirth R, Schreyer W & Maresch WV, European Journal of Mineralogy (submitted), (2000).

Reinecke T, Koch-Müller M & Langer K, Terra cognita, 8, 74, (1988).

Fockenberg T, American Mineralogist, 83, 220-227, (1998).

Polysomatism of Synthetic Antigorite

Bernd Wunder (wunder@gfz-potsdam.de), Richard Wirth & Matthias Gottschalk

GeoForschungsZentrum Potsdam, Germany

Due to its wave-like modulated structure along the a-axis, antigorite forms a series of discrete compositions on the join chrysotile-talc within the MgO-SiO₂-H₂O (MSH)-system. Individual compositions can be expressed by the formula Mg_3m-3Si_2mO_5m(OH)_4m-6 and typically compositions of natural antigorites are in the range of m=14-23. With increasing metamorphic grade the m-value of antigorites seems to decrease. The aim of this experimental study is to investigate the P,T-dependence of the antigorite polysomatism for samples synthesized over a wide P,T-range in the pure MSH-system. Experiments were performed at temperatures between 450 and 600°C and pressures of 0.2-5.0 GPa. Run-durations varied between 4 and 35 days. Starting materials were synthetic brucite and talc, mixed in the stoichiometric proportions of antigorite with m=17 plus 20 wt.% additional water. Products were characterized by XRD and HRTEM. For the determination of the a lattice modulation (and corresponding m-values), lattice fringe spacings were deduced from diffraction patterns (finite fourier transformation) of selected (010) sections in the lattice fringe image. From XRD it is obvious that in all experiments serpentine had formed. Talc, brucite or forsterite appeared as additional phases. The TEM-study indicates that at pressures * 1.5 GPa antigorite is the dominant serpentine phase. The amount of chrysotile increases with decreasing pressure and temperature; at 0.2 GPa, 450°C only chrysotil is obvious. Antigorites of each experimental run show a heterogeneous distribution of periodicities with m-values in the range 13-21. HRTEM-measurements were therefore repeated over several grains. Maximum m-values of resulting periodicity distributions are a function of pressure and temperature and are more pronounced with increasing run-duration. The highest maximum m-value of 18 was determined for antigorite synthesized at 5.0 GPa, 500°C, the lowest (m=14) at 1.5 GPa, 600°C. The results are treated thermodynamically.