Alkali-free tourmaline in the system MgO-Al 2O 3-B 2O 3-SiO 2-H 2O

Abstract

An end member of the tourmaline series with a structural formula □(Mg 2Al)Al 6(BO 3) 3[Si 6O 18](OH) 4 has been synthesized in the system MgO-Al 2O 3-B 2O 3-SiO 2-H 2O where it represents the only phase with a tourmaline structure. Our experiments provide no evidence for the substitutions Al → Mg + H, Mg → 2H, B + H → Si, and AlAl → MgSi and we were not able to synthesize a phase “Mg-aluminobuergerite” characterized by Mg in the (3a)-site and a strong (OH)-deficiency reported by Rosenberg and Foit (1975). The alkali-free tourmaline has a vacant (3a)-site and is related to dravite by the □ + Al for Na + Mg substitution. It is stable from at least 300°C to about 800°C at low fluid pressures and 100% excess B 2O 3, and can be synthesized up to a pressure of 20 kbars. At higher temperatures the tourmaline decomposes into grandidierite or a boron-bearing phase possibly related to mullite (“B-mullite”), quartz, and unidentified solid phases, or the tourmaline melts incongruently into corundum + liquid, depending on pressure. In the absence of excess B 2O 3 tourmaline stability is lowered by about 60°C. Tourmaline may coexist with the other MgO-Al 2O 3-B 2O 3-SiO 2-H 2O phases forsterite, enstatite, chlorite, talc, quartz, grandidierite, corundum, spinel, “B-mullite,” cordierite, and sinhalite depending on the prevailing PTX-conditions. The (3a)-vacant tourmaline has the space group R3m with a =15.90 A ̊ , c = 7.115 A ̊ , and V = 1557.0 A ̊ 3 . However, these values vary at room temperature with the pressure-temperature conditions of synthesis by ±0.015 A ̊ in a , ±0.010 A ̊ in c , and ±4.0 A ̊ 3 in V , probably as a result of MgAl order/disorder relations in the octahedral positions. Despite these variations intensity calculations support the assumed structural formula. Refractive indices are n o = 1.631(2), n E = 1.610(2), Δn = 0.021. The infrared spectrum is intermediate between those of dravite and elbaite. The common alkali and calcium deficiencies of natural tourmalines may at least partly be explained by miscibilities towards (3a)-vacant end members. The apparent absence of (3a)-vacant tourmaline in nature is probably due to the lack of fluids that carry boron but no Na or Ca.