Scott A. Wood

The aqueous geochemistry of the rare-earth elements and yttrium: 1. Review of available low-temperature data for inorganic complexes and the inorganic REE speciation of natural waters☆

Abstract Literature data on the nature and thermodynamics of inorganic complex species of the REE and Y have been critically reviewed. Theoretical considerations suggest that trivalent REE and Y should exhibit strong, predominantly electrostatic complexing with “hard” ligands such as fluoride, sulfate, phosphate, carbonate and hydroxide and this is borne out by the available experimental data. Complexing of these metals with chloride and nitrate is moderately weak and with ammonia and bisulfide is extremely weak to nonexistent. A considerable amount of concordant thermodynamic data are available at 25°C for the sulfate complexes LnSO + 4 and Ln(SO 4 ) − 2 , the fluoride complexes LnF 2+ , LnF + 2 and LnF 0 3 , the nitrate complex LnNO 2+ 3 and the chloride complex LnCl 2+ , where Ln signifies the REE or Y. There are much fewer reliable data available for the hydroxide, carbonate and phosphate complexes and in fact, the exact stoichiometries of REE and Y complexes with these ligands are still somewhat uncertain. However, it appears that the LnOH 2+ , LnCO + 3 , Ln(CO 3 ) − 2 , LnHCO 2+ 3 and LnH 2 PO 2+ 4 complexes have been identified. Calculation of the speciation of Eu ( β Eu = 10 −7 m) in a typical groundwater with β SO 2− 4 = 10 −4 m, β C − = 2·10 −4 m, β F − = 10 −6 m, β CO 2− 3 = 10 −4 m, β NO − 3 = 10 −4 m and β PO 3− 4 = 10 −6 m shows that the simple ion and the sulfate complexes are most important at acidic pH and that the carbonate complexes become predominant at near-neutral to basic pH. Even in relatively saline waters, chloride complexes do not account for a significant fraction of REE in solution and are completely negligible in most surface waters and groundwaters. Nitrate complexes are also negligible even in environments where the concentration of this ligand is artificially high due to pollution. Phosphate and fluoride complexes can attain importance where concentrations of these ligands are somewhat anomalous (i.e. 2–4 orders of magnitude higher than in the model groundwater) due to either natural or anthropogenic factors. Data on the nature and stabilities of REE and Y complexes at low temperature are of critical importance to those concerned with safe nuclear waste disposal, geochemical exploration for REE and Y deposits, and the use of REE and Y as tracers in seawater and fresh water. The information most needed at the present time includes: (1) reliable hydrolysis constants; (2) additional data on phosphate, carbonate and the higher fluoride complexes; (3) data on mixed ligand complexes; and (4) more complete data on the complexation behavior of the +4 and +2 oxidation states of the REE.

The aqueous geochemistry of the rare-earth elements and yttrium: 2. Theoretical predictions of speciation in hydrothermal solutions to 350°C at saturation water vapor pressure

Stability constants for trivalent REE complexes containing fluoride, chloride, sulfate, carbonate or hydroxide ions have been predicted up to 350°C at saturated water vapor pressure using Helgesons electrostatic approach, combined with the isocoulombic approach and available experimental thermodynamic data at low temperatures. In addition, the equilibrium boundary between divalent and trivalent REE ions was calculated at elevated temperatures and 1 kbar for Sm, Yb and Eu as a function of pH and oxygen fugacity. Calculations suggest that the divalent state for all three REE (Sm, Yb and Eu) becomes increasingly important at geologically reasonable pH and oxygen fugacity conditions as temperature increases, but only Eu, and possibly Yb, will be present in the divalent state in most hydrothermal solutions at temperatures of 600°C) the divalent state of many of the REE should prevail over the trivalent state, especially at low pressures (<2 kbar). Extrapolation of the available low-temperature stability constant data indicates that the stabilities of all the trivalent REE complexes considered increase relatively rapidly with temperature, as expected for complexation between hard metal ions and hard ligands. The following table illustrates some of the predicted data for the first cumulative stability constants of sulfate, fluoride and chloride complexes, as well as the first hydrolysis constant for La3+ and Lu3+ at 25° and 300°C: The predicted increase in stability constants with temperature is greatest for fluoride and least for chloride. Even allowing for uncertainty in the estimated stability constants, at geologically reasonable ligand concentrations, fluoride complexes predominate over all others in the mildly acidic to mildly basic pH region at elevated temperatures. Because of the predicted high stability of fluoride and carbonate complexes of the trivalent REE, as well as the empirically observed correlation between REE mobility and the presence of these ligands (in the minerals fluorite and calcite) in many hydrothermally affected geological environments, accurate experimental determination of the stability constants of these complexes should be a priority.

Gold speciation in natural waters: I. Solubility and hydrolysis reactions of gold in aqueous solution

Solubility measurements of Au in dilute to concentrated aqueous NaOH solutions at 25°C have been carried out. The data were fitted to the general half reaction Au(c) + 2 H2O → AuO2H4−n1−n + nH+ + e− in order to identify the stoichiometry and stability of the hydrolyzed species formed. The monohydroxide, AuOH(H2O)0, is found to be the most stable species up to pH ~ 12. The equilibrium constant for the formation of this species (i.e., n = 1) is logK1 = −22.57 ± 0.44. Consideration of competitive complexation of Au by a number of inorganic ligands which are encountered in natural waters indicates that AuOH(H2O)0 is the most stable inorganic Au species over a wide range of Eh, pH, and ligand activities. The only inorganic ligands which may occur in natural waters at concentrations high enough to stabilize Au(I) include HS− under reducing conditions, S2O32− under alkaline oxidizing conditions, Cl− in very acidic, oxidizing brines, and possibly CN− locally, in environments where there is biogenic and/or anthropogenic production of cyanide. Calculated equilibrium pe-pH diagrams for Cl− and ΣS activities typical of both fresh and sea water show that AuOH(H2O)0 is probably the dominant dissolved Au species in these environments, with the exception of anoxic ocean and lake waters.

Thermodynamic calculations of the volatility of the platinum group elements (PGE): The PGE content of fluids at magmatic temperatures☆

Abstract The volatilities of the platinum-group elements as metals, oxides and chlorides were calculated at temperatures of 800–1600 K. Only Pd is significantly volatile as the metal. At logfH2O = 1 Kbar and 1200 K., the concentration (weight) of Pd in the vapor reaches 1 ppt and at 1600 K attains several ppb. The PGE oxides are extremely volatile at atmospheric oxygen fugacities. However, only Os and Ru have significant volatilities (≥ ppt) as oxides (OsO4, RuO3) at oxygen fugacities typical of magmatic PGE deposits (near QFM) and only at temperatures greater than 1400 K. Data on the volatility of PGE chlorides exist only for Pd and Ru, both of which are somewhat more volatile as chlorides than as oxides. At 1400 logfH2O = 1 bars, fHCl = 100 bars and at QFM, the calculated vapor concentrations of PdCl2 and RuCl3 are 500 ppt and 20 ppt, respectively (and less in the presence of sulfur). However, higher concentrations of PGE may be attained at higher temperatures, higher fO2, higher fHCl or lower fH2. Also, any interactions between water vapor and PGE vapor species (e.g. ionization, solvation) would tend to increase the vapor concentration of PGE. Volatility of Ir as IrF6 is insignificant at all conditions. Vapor transport of the more volatile PGE as chlorides may play some role in the transport of these metals in mafic igneous complexes such as the Stillwater or the Bushveld. However, under the conditions where the PGE are most volatile, the metals Fe, Ni and Cu are several factors often more volatile, so that enrichment of the PGE and Cu over Ni and Fe cannot be explained by chloride transport alone.

Solubility of Pt and Pd sulfides and Au metal in aqueous bisulfide solutions

An experimental study of the solubility of Pt and Pd sulfides and Au metal in aqueous bisulfide solutions was conducted at temperatures from 200° to 350 °C and at saturated vapor pressure. A 500-mL Bridgemantype pressure vessel constructed of titanium, and equipped with a motor-driven magnetic stirrer was employed. The pH and the oxidation state were buffered by the coexistence of H2S/HS−/SOinf4sup2−. The pH at temperature was calculated to be in the range 5.91–9.43, and ∑S was 0.3–2.2 m. Under the experimental conditions, the measured solubility of gold is about two to three orders of magnitude greater than that of either platinum and palladium, and the measured solubility of platinum is, in general, approximately equal to that of palladium, in molal units. The solubilities are found to be in the range: platinum 4–800 ppb, palladium 1–400 ppb, and gold 2–300 ppm. The solubility data can be modeled adequately using the following reactions: Au+H2S+HH−=Au(HS)2−+1/2H2 (K14); PtS+HS−+H+=Pt (HS)20(K15); PdS+HS−+H+=Pd (HS)20(K16); PtS2+H2=Pt (HS)20(K21).With equilibrium constants determined as follows (errors represent two standard deviations): Preliminary measurements of the solubilities of metallic Pt, Pd and Au as hydroxide complexes were also conducted using a second titanium pressure vessel, at temperatures of 200° to 350 °C and vapor saturation pressure, with pH and the oxidation state controlled or buffered by adding known amounts of NaOH and H2 gas. The concentration of NaOH was in the range 0.01–1.3 m, and the partial pressure of H2 at 200 °C was 62–275 bars, initially. Under the temperature and pressure conditions of these experiments, the solubility of platinum in 1 m NaOH solution is less than 100 ppb, that of palladium is less than 10 ppb and that of gold is less than 0.2 ppm; and in 0.01 m NaOH solutions, both Pt and Pd solubilities are less than 1 ppb. These data indicate that the contributions of hydroxide complexes to the total solubilities in the bisulfide runs, where the pH was in the range of 5.9–9.4, are negligible. The concentrations of both Pt and Pd as bisulfide complexes in the Salton Sea geothermal system predicted using the stability constants determined in this work agree very well with those values measured by McKibben et al. (1990). This calculation strongly suggests that the PGE are transported in moderately reducing, near neutral hydrothermal fluids as bisulfide complexes, as is gold. However, the much lower maximum solubility of the PGE relative to gold severely constrains models of re genesis, and may explain the relative rarity of hydrothermal PGE deposits compared to the relative abundance of hydrothermal Au deposits.

Gold speciation in natural waters: II. The importance of organic complexing—Experiments with some simple model ligands

The solubility of Au has been measured at 25°C in aqueous solutions in the presence of various organic ligands (acetate, benzoate, oxalate, phthalate, salicylate, and thiosalicylate). These ligands were chosen as simple analogs of humic acid moieties in order to model the complexation of Au by humic and fulvic acids in natural waters. With the first five ligands (ΣL = 0.1 M), solubilities were generally below 25 μg/l, whereas in the thiosalicylate solutions (ΣL = 0.45 M), a maximum Au concentration of 680 mg/1 was measured. Acetate and benzoate complexes are too weak to detect by the solubility method. Oxalate appears to have a reducing effect on Au in solution, and both oxalate and phthalate complexes of Au(I) may be coordinatively unfavorable. It was only possible to identify one salicylato complex (logβ2 = 17.5 ± 0.5) and two thiosalicylato species (logβ1 = 29.9 ± 0.3 and logβ2 = 31.7 ± 0.3). In addition, stability constants for a number of O-, N-, and S-donor complexes of Au(I) were estimated from linear free energy relationships with Cu(I), Ag(I), and Hg(II). General trends in stability constants of Au-organic complexes with various donor atoms are S ⪢ N ≥ O. Calculations based on a simple model of a fulvic acid suggest that Au is almost exclusively bound to S-donor sites under reducing conditions, but AuOH(H2O)0 and complexing by organic O-donors are more important in oxidizing environments.

Experimental determination of the hydrolysis constants of Pt2+ and Pd2+ at 25°C from the solubility of Pt and Pd in aqueous hydroxide solutions☆

The solubilities of Pt and Pd metal were measured at 25°C in 10−4 to 10.0 molal NaOH solutions under a reduced oxygen atmosphere in order to determine the stoichiometry and stability constants for Pt and Pd hydroxide complexes. Equilibration times of over one year were employed. The Pd data are consistent with the existence of Pd(OH)20(aq) from pH 9 to 12 and Pd(OH)3− from pH 12 to 15.5. No conclusive evidence for a Pd(OH)42− complex was obtained, but the data do not preclude its existence at high pH. For Pt, the data are consistent with a single complex for pH = 9 to 15.5, I.E., Pt(OH)20(aq). A graphical treatment of the data yields the following cumulative stability constants: log β2 = 18.9 ± 1.0 and log β3 = 20.9 ± 1.0 for Pd and log β2 = 29.9 ± 1.0 for Pt. The stepwise stability constant for Pd(OH)3−log K3 = 2.0 is in relatively good agreement with that derived from data in the literature (log K3 = 1.8). However, the cumulative stability constants for Pd measured in this work are considerably smaller than those reported in the literature. The log β2 = 29.9 ± 1.0 value measured for Pt compares relatively well with a theoretically estimated value of 28.3. The data suggest that the predominant inorganic form of Pt and Pd in freshwaters may be the neutral hydroxide species. In seawater, the hydroxide complex of Pt is also predicted to predominate over the chloride complex, but, in the case of Pd, the hydroxide complex appears to be less stable and it is presently not clear whether the chloride or the hydroxide complex will predominate. In fluids responsible for serpentinization, Pt and Pd may also be mobilized as hydroxide complexes.

Solubility and Transport of Platinum-Group Elements in Hydrothermal Solutions: Thermodynamic and Physical Chemical Constraints

Thermodynamic calculations and physical chemical considerations suggest that at least some of the platinum-group elements (PGE) may be mobile in geological fluids such as chloride, hydroxide, bisulphide, polysulphide, thiosulphate or ammonia complexes, depending on pH, fO2, temperature and lignad concentration.

A preliminary petrogenetic grid for REE fluorocarbonates and associated minerals

The bulk of the worlds economic LREE reserves occur as fluorocarbonate minerals, notably bastnaesite. However, despite the importance of these minerals, very little is known about the physicochemical conditions controlling their formation. In this paper we attempt to partly redress this deficiency by qualitatively determining P-T and compositional relationships for part of the system Ln(CO3)F-CaCO 3-F2(CO3) −1-H2O, including the minerals fluorite, calcite, bastnaesite, parisite, synchysite and fluocerite. This degenerate (n + 3)-phase multisystem has 23 possible base P- T topologies, plus their mirror images and trivial conjugates, from which we have been able to select a single probable stable topology using a combination of published experimental phase equilibrium data, molar volume and entropy estimates and natural assemblage data. Compositional relationships in the system have been established by constructing log (aca2+ · aF−2) vs. log (aF−2aCO32−) diagrams for each of the stable divariant regions shown on the P-T net. Important conclusions of the study with respect to P-T relationships are 1. (1) that all of the above REE-fluorocarbonate minerals can form at comparatively low pressure and temperature; 2. (2) that bastnaesite + fluorite is a low-temperature assemblage and, in the presence of synchysite or calcite, is also restricted to low or high pressure, respectively; 3. (3) that parisite + fluorite is stable to higher temperatures; 4. (4) that bastnaesite + synchysite + calcite is restricted to high P-T conditions; 5. (5) that parisite reacts to form bastnaesite and calcite at high temperatures (<620°C at 1 kb); and 6. (6) that bastnaesite-(La) decomposes by a decarbonation reaction at temperatures <750°C at 1 kb and at lower temperatures with decreasing ionic radius of the lanthanide. The principal conclusions with respect to compositional relationships are 1. (1) that transformations among the REE fluorocarbonates cannot occur through changes in F− activity alone, but can result from variations in either Ca2+ or CO32− activities, and 2. (2) that fluocerite can coexist with any of the fluorocarbonate minerals at high F− or low CO32− activity depending on the activity of Ca2+, pressure and temperature. Fuller understanding of the genesis of REE fluorocarbonate deposits will require extensive fluid inclusion and related studies and systematic determinations of phase relations through well-constrained, reversed experiments.

Experimental determination of the hydrothermal solubility and speciation of tungsten at 500°C and 1 kbar1,2☆☆☆

The solubility of crystalline WO3 was measured in pure water and solutions of HCl (0.5–5 molal), NaOH (0.01–1 molal), NaCl (1–6 molal) and KCl (1 molal) at 500°C and approximately 1 kbar in order to determine the speciation of tungsten in hydrothermal solutions in the presence of Cl−, Na+ and K+. The average solubility of WO3 in pure water was measured to be 515 ppm W under the above conditions. No significant increase in solubility was observed in the presence of up to 5 molal HCl implying that Cl− complexing of W is not important in natural solutions. A log K value for the reaction WO3(s) + H2O(l) = H2WO4(aq) of −2.55 ± 0.2 was derived. WO3 solubility was considerably increased by the addition of NaCl and even more so by the addition of NaOH. The dependence of W concentration on NaCl and NaOH concentrations, together with quench pH measurements, suggest the importance of a cation-tungstate ion pair such as NaHWO40, but the possibility of a polymeric tungsten species cannot be excluded. A value of log K = 1.0 ± 0.1 was derived for the reaction WO3(s) + NaOH0(aq) = NaHWO40(aq) and a value of log K = −4.22 ± 0.09 was derived for the reaction WO3(s) + NaCl0(aq) + H2O(1) = NaHWO40(aq) + HCl0(aq). The solubility of assemblages containing WO2 was also investigated. These solubilities were higher than those measured for WO3 alone. This combined with the intense blue color of the solutions indicates the presence of a significant amount of a reduced (W(V) species.