Alkali metal and rare earth element evolution of rock-forming minerals from the Gatumba area pegmatites (Rwanda): Quantitative assessment of crystal-melt fractionation in the regional zonation of pegmatite groups

This study presents a general model for the evaluation of Rayleigh fractional crystallisation as the principal differentiation mechanism in the formation of regionally zoned common and rare-element pegmatites. The magmatic evolution of these systems from a granitic source is reconstructed by means of alkali element and rare earth element (REE) analyses of rock-forming minerals (feldspars, micas and tourmaline), which represent a whole sequence of regional pegmatite zonation. The Gatumba pegmatite field (Rwanda, Central Africa) is chosen as case study area because of its well-developed regional zonation sequence. The pegmatites are spatially and temporally related to peraluminous G4-granites (986 +/- 10 Ma). The regional zonation is developed around a G4-granite and the proximal pegmatites grade outwardly into biotite, two-mica and muscovite pegmatites. Rare-element (Nb-Ta-Sn) pegmatites occur most distal from the granite. Alkali metal fractionation trends in pegmatitic K-feldspar (Rb 350-6000 ppm, Cs 2-160 ppm) and muscovite (Rb 670-7000 ppm, Cs 10-150 ppm) define a single and continuous trend, which is modelled by Rayleigh fractional crystallisation, starting from a parental granite composition (G4-composition: K 3.1 wt%, Rb 222 ppm, Cs 11 ppm). The fractionation model shows, moreover, that the pegmatites adjacent to the parental pluton are the least fractionated, and the distal pegmatites are the most fractionated. Biotite pegmatites form from 0% to 69% crystallisation, two-mica pegmatites from 69% to 92%, and muscovite pegmatites from 92% crystallisation onwards. The extreme Rb- and Cs-enrichment in rare-element pegmatites requires at least 98% fractionation of the initial G4-granite composition. Mathematical derivation of the K/Rb versus Cs relationship in K-feldspars confirms Rayleigh fractional crystallisation as the main differentiation process in the development of regional pegmatite zonation. Moreover (1) it demonstrates the continuity of the fractionation process from biotite pegmatites to rare-element pegmatite and indicates a genetic link among them; and (2) it allows a general evaluation of pegmatite fields in terms of system parameters, i.e. the initial element concentration in the granitic melt and partition coefficients. REE patterns of the rock-forming minerals (feldspars, muscovite, biotite and schorl) show a distinct transition along the regional zonation. They evolve from sloping fractionated to more flat patterns starting from the biotite up to the muscovite pegmatites. Minerals from the rare-element pegmatites (feldspars, muscovite and elbaite) show again more fractionated, heavy REE-depleted patterns. The observed evolution in the REE patterns corresponds to early crystallisation of light REE-enriched monazite and a late crystallisation of mainly cogenetic heavy REE-enriched phases such as apatite, columbite-group minerals and beryl. Modelling of Rayleigh fractionation, starting from an initial parental granite (G4-composition: La-N/Yb-N 10 and Sigma REE 83), shows that this evolution in mineral REE pattern is the result of fractional crystallisation of the pegmatitic melt and precipitation of REE-incorporating minerals such as monazite and apatite. Consequently, trace element modelling indicates that Rayleigh fractional crystallisation governs the mineralogical and geo-chemical evolution from a granite source to common and eventually rare-element pegmatites. This mechanism shows that granitic pegmatites are extremely fractionated, residual melts which are genetically and directly connected to a main granite body. (C) 2014 Elsevier Ltd. All rights reserved.