The hydrothermal geochemistry of tungsten in granitoid environments: I. Relative solubilities of ferberite and scheelite as a function of T, P, pH, and m(NaCl)

The characteristics of granitoid-related tungsten deposits hosted in silice ous (carbonate-free) rocks (e.g., Panasqueira, Cligga Head, Paste Bueno) ar e reviewed and the ranges of physicochemical parameters of the ore-forming fluids are summarized. The two important tungsten minerals in these deposit s are wolframite and scheelite, which were deposited mostly between 200 deg rees and 500 degrees C and 200 and 1,500 bars. The salinities of the minera lizing fluids were typically less than 15 wt percent but commonly were sign ificantly higher (up to 55 wt %). The two predominant dissolved components are Na+ and Cl- with subordinate Ca2+, K+, and carbonate species (CO32-/HCO 3-). The contents of CO2 are highly variable, but X-CO2 values typically ra nge from 0 to 0.1. Limited pH and f(O2) estimates indicate a moderately aci dic fluid with oxygen fugacities between those of the QFM and HM buffers. T hese parameters were used to guide solubility and speciation modeling of W in hydrothermal fluids in granitoid environments. Experimentally derived thermodynamic data for scheelite, ferberite, aqueous Ca, Fe, and W species, and other required aqueous species were critically evaluated and the most reliable data were adopted. where necessary, missing data were estimated. The resultant thermodynamic database provides a basis for solubility and speciation calculations in the system Ca-Fe-W-Cl-O-H. T he simultaneous solubilities of scheelite and ferberite in NaCl-HCl-K2O sol utions were calculated at temperatures from 200 degrees to 600 degrees C, p ressures from 500 to 1,000 bars, pH from 3 to 6, and m(NaCl) from to 0.1 to 5.0 moles/kg H2O. The solubility model takes account of the species H+, OH -, Na+, Cl-, NaCl0, HCl0, NaOH0, H2WO40, HWO4-, WO42-, Fe2+, FeCl+, FeCl20, FeOH+, FeO0, HFeO2-, Ca2+, CaCl+, CaCl20, CaOH+, NaHWO40, and NaWO4-. The calculations indicate the following: (1) solubilities of scheelite and/or f erberite can attain values as high as hundreds to thousands of parts per mi llion as the tungstate species H2WO40, HWO4-, WO42-, NaHWO40, and NaWO4-; t hus, tungsten-chloride, -fluoride, or -carbonate complexes, or more exotic species are not required to transport sufficient W to form an ore deposit; (2) the tungsten concentration in equilibrium with scheelite and ferberite increases strongly with increasing temperature, increasing NaCl concentrati on and decreasing pH, but is only weakly dependent on pressure; (3) the Ca/ Fe ratio of a solution in equilibrium with both scheelite and ferberite dec reases strongly with increasing temperature, i.e., the field of stability o f scheelite expands with increasing temperature; the implication, therefore , is that simple cooling of a solution with a constant Ca/Fe ratio cannot r esult in the replacement of ferberite by scheelite, and that held observati ons of the late-stage replacement of ferberite by scheelite require an incr ease in the Ca/Fe ratio concomitant with cooling; (4) the Ca/Fe ratio is re latively independent of pH; and (5) the effect of NaCl concentration on thi s ratio changes as a function of temperature and pressure. At less than 400 degrees C the ratio is independent of, or decreases with, increasing NaCl concentration; at higher temperatures the ratio first decreases and then in creases with increasing NaCl concentration. Experimental data on the solubi lity of scheelite and the Ca/Fe ratio of fluids in equilibrium with scheeli te + ferberite, and which are not used in parameterizing our model, general ly agree with the results of calculations performed using our thermodynamic database within an order of magnitude. However, our critical examination o f available thermodynamic data reveals that significant uncertainty remains in several parameters (e.g., the solubility products of scheelite and ferb erite and the association constants for alkali tungstate ion pairs). This u ncertainty can only be reduced via carefully conceived, executed, controlle d, and interpreted experiments, taking into account the various experimenta l pitfalls identified in this paper.