Experimental Investigation of Andalusite-Topaz Relations In Synthetic Peraluminous Haplogranites

Many evolved peraluminous granite plutons contain either andalusite (Al₂SiO₅) or topaz (Al₂SiO₄(OH,F)₂), however some plutons contain both these A/NK = ∞ phases. Textural and chemical evidence shows that most andalusite in peraluminous granitic rocks is of primary magmatic origin, and that at least some topaz can also crystallize directly from the silicate melt. The appearance of such primary magmatic topaz is an important mineralogical expression of fluorine activity in the melt, and may be an indicator of spatial proximity to fluorine-related, high field-strength, Sn-W-U-Mo polymineralic mineralization. We present the results of experiments using water-saturated synthetic peraluminous haplogranites to locate the andalusite-topaz boundary in T-P-X space, where T = 700-650^oC, P = 200 MPa, and X = A/NK (1.1, 1.2, 1.3, 1.4) and F (0.0, 1.0, 2.0, 4.0 wt.%). We also try to determine if the melt-andalusite-topaz relationship is cotectic or peritectic. The starting materials are synthetic gels in the system KNASH (with fluorine added as AgF), run times are 5-7 days, and analysis of run products is by electron microprobe. Runs also include one part-per-thousand each of seeds of natural andalusite and synthetic topaz to assist in overcoming nucleation problems. As expected, for all values of A/NK, low fluorine concentrations favour andalusite stability, whereas high fluorine concentrations favour topaz stability. For intermediate fluorine concentrations and A/NK ≤ 1.3, either mullite is stable or no A/NK = ∞ phase is present. For intermediate fluorine concentrations and A/NK = 1.4, andalusite alone, topaz alone, or andalusite + topaz are stable, depending on temperature. Experimental runs and natural samples with well-developed topaz reaction rims on andalusite suggest that topaz is the product of a peritectic reaction between an early primary magmatic andalusite and a late fluorine-enriched melt.