Applied Geochemistry (v.26, #9-10)

Geochemical controls on leaching of lignite-fired combustion by-products from Greece by Maria Izquierdo; Nikolaos Koukouzas; Sofia Touliou; Kyriakos D. Panopoulos; Xavier Querol; Grigorios Itskos (1599-1606).
► The metal assemblage in ash fingerprints the geological setting of coal basins. ► Free lime in ash influences the total and leachable contents of volatile elements. ► The mode of occurrence of elements in lignite controls the fly ash leaching. ► Oxyanionic-forming species are the main concern in terms of leaching of ashes.The mobility of inorganic pollutants is of key concern for a range of industrial and engineering applications of fly ash produced during the combustion of lignite in power generation. This paper investigates the role that the geochemical features of lignite, the ash composition and the partitioning of elements during combustion play in determining leaching properties of lignite fired by-products. The work is based on surveys on three lignite-fired power plants in Greece. Calcium-rich ashes show a high abatement potential for SO2 and other gaseous pollutants. For most elements, the concentrations in the parent lignite and the ashes follow the same trend. Relative enrichments in Cd, Co, Cr, Cu, Mo, Ni, Pb, U, V, W, Zn fingerprint the regional and local geological settings of the lignite basins. The total and leachable concentrations of highly volatile elements are strongly influenced by the interaction with ubiquitous free lime. A broad array of elements is highly insoluble in alkaline ash, while a few oxyanionic-forming elements display substantial mobility. Their mode of occurrence in the parent lignite plays a primary role in the leaching of combustion ashes. The outcomes of this study may assist in addressing the impact of co-firing high ash or high Ca alternative fuels on the leaching properties of combustion by-products.

Phosphorus sequestration in Fe-rich sediments from two Brazilian tropical reservoirs by R. Fonseca; T. Canário; M. Morais; F.J.A.S. Barriga (1607-1622).
► Role of Fe-oxide/hydroxide rich sediments of tropical reservoirs on P sequestration. ► P adsorption capacity of sediments is predominantly associated with Fe. ► Clay fraction mineralogy dominated by oxides/hydroxides-Fe and kaolinite. ► Strong retention of phosphate ions through surface adsorption/precipitation phenomena. ► Study of mineralogy/chemistry of sediments could predict P levels in surface waters.This study investigates the important role of Fe-oxide/hydroxide-rich bottom sediments, from two Brazilian tropical reservoirs, on P sequestration from the overlying waters, which decreases the risk of eutrophication. Data on P fractionation indicates that P adsorption capacity of sediments is predominantly associated with Fe, followed by orthosphosphate adsorbed onto Al hydr(oxides) and onto different forms of apatite (Ca–P). The soluble and loosely bound P-forms are low. To have a better understanding of the role of Fe oxides on the adsorption and sequestration of P in the reservoirs, data obtained by sequential extraction of P have been related to concentrations of total Fe and of various Fe-fractions extracted through an optimized fractionation scheme. These analyses indicated high levels of Fe bound to oxides/hydroxides. These mineral oxides, and the preponderance of pH-dependent charged clay minerals (kaolinite) in the clay fraction, induce the sequestration of the P entering the lakes, mainly in particulate form, by strong retention of PO 4 3 - through surface adsorption and precipitation phenomena. Phosphorus-retention provided by the replacement of OH groups on the basal plane of the minerals, for P anions, decreases the negative impact of P inputs in these water systems. There is significant correlation between the concentrations of soluble reactive P in the water (SRP) and major chemical components of sediments (SiO2, Fe2O3, Al2O3, K2O, P2O5), which are related to the dominant bedrock geology. The interaction of SRP with bed-sediments and the chemical conditions of the environment (decrease of redox potential with depth and, in the dry period, a low O2 abundance), can explain fluctuations of SRP and TP (total P) in the water column. Due to the dynamic equilibrium in the sediment–water interface, in the dry season, reducing conditions in hypoliminia enhance the reduction of Fe(III) to the more soluble Fe(II) with subsequent release of the strongly retained P. This mechanism increases the soluble P fraction in sediments and the soluble reactive P in the water column, in this period. The straight linkage between geochemical properties of sediments and P concentration in the water column reported in this article suggests that study of the mineralogical and geochemical composition of bottom sediments could help to improve model predictions of P concentrations in surface waters.

Pb-concentrations and Pb-isotope ratios in soils collected along an east–west transect across the United States by Clemens Reimann; David B. Smith; Laurel G. Woodruff; Belinda Flem (1623-1631).
► Lead concentrations and isotope ratios are reported for soil samples from an east-west transect across the United States. ► The 206Pb/207Pb isotope ratio shows a distinct trend toward lower values when approaching both coasts. ► The 206Pb/207Pb isotope ratios cannot be linked to traffic emissions. ► Weathering changes the 206Pb/207Pb isotope ratio in soil samples.Analytical results for Pb-concentrations and isotopic ratios from ca. 150 samples of soil A horizon and ca. 145 samples of soil C horizon collected along a 4000-km east–west transect across the USA are presented. Lead concentrations along the transect show: (1) generally higher values in the soil A-horizon than the C-horizon (median 21 vs. 16.5 mg/kg), (2) an increase in the median value of the soil A-horizon for central to eastern USA (Missouri to Maryland) when compared to the western USA (California to Kansas) (median 26 vs. 20 mg/kg) and (3) a higher A/C ratio for the central to eastern USA (1.35 vs. 1.14). Lead isotopes show a distinct trend across the USA, with the highest 206Pb/207Pb ratios occurring in the centre (Missouri, median A-horizon: 1.245; C-horizon: 1.251) and the lowest at both coasts (e.g., California, median A-horizon: 1.195; C-horizon: 1.216). The soil C-horizon samples show generally higher 206Pb/207Pb ratios than the A-horizon (median C-horizon: 1.224; A-horizon: 1.219). The 206Pb/207Pb-isotope ratios in the soil A horizon show a correlation with the total feldspar content for the same 2500-km portion of the transect from east-central Colorado to the Atlantic coast that shows steadily increasing precipitation. No such correlation exists in the soil C horizon. The data demonstrate the importance of climate and weathering on both Pb-concentration and 206Pb/207Pb-isotope ratios in soil samples and natural shifts thereof in the soil profile during soil-forming processes.

► Weathering of the black limestone on historical monuments. ► Oxygen and sulfur isotope composition of sulfate salts (outdoors and indoors). ► A data exhibit more scattered δ34S values with respect to δ18O values.The black limestone widely used in Slovenian monuments, particularly in the baroque architecture, is deteriorating extensively due to salt crystallization. Samples of soluble salts from two important historical monuments (in Ljubljana, Slovenia) were investigated in terms of their mineral and isotopic (S and O) compositions. Results revealed the presence of gypsum and soluble salts of the MgSO4·nH2O series, such as starkeyite (MgSO4·4H2O), pentahydrite (MgSO4·5H2O) and hexahydrite (MgSO4·6H2O). Whereas black crusts and subflorescences consisted of gypsum, efflorescences appeared to be an assemblage of gypsum and MgSO4 hydrates. Sample δ18Osulfate values varied from −1.9‰ to +5.5‰ vs. V-SMOW and δ34Ssulfate values from −19.8‰ to +3.2‰ vs. V-CDT. The respective isotopic composition of analysed outdoor and indoor monument samples indicated different sources of contamination.

Direct characterization of airborne particles associated with arsenic-rich mine tailings: Particle size, mineralogy and texture by M.C. Corriveau; H.E. Jamieson; M.B. Parsons; J.L. Campbell; A. Lanzirotti (1639-1648).
► Airborne dust from As-rich gold mine tailings used for recreation was collected. ► Total concentrations of arsenic in the <8 μm fraction varied from 65 to 1040 ng/m3. ► Multiple As minerals in dust are comparable with near-surface tailings samples.Windblown and vehicle-raised dust from unvegetated mine tailings can be a human health risk. Airborne particles from As-rich abandoned Au mine tailings from Nova Scotia, Canada have been characterized in terms of particle size, As concentration, As oxidation state, mineral species and texture. Samples were collected in seven aerodynamically fractionated size ranges (0.5–16 μm) using a cascade impactor deployed at three tailings fields. All three sites are used for recreational activities and off-road vehicles were racing on the tailings at two mines during sample collection. Total concentrations of As in the <8 μm fraction varied from 65 to 1040 ng/m3 of air as measured by proton-induced X-ray emission (PIXE) analysis. The same samples were analysed by synchrotron-based microfocused X-ray absorption near-edge spectroscopy (μXANES) and X-ray diffraction (μXRD) and found to contain multiple As-bearing mineral species, including Fe–As weathering products. The As species present in the dust were similar to those observed in the near-surface tailings. The action of vehicles on the tailings surface may disaggregate material cemented with Fe arsenate and contribute additional fine-grained As-rich particles to airborne dust. Results from this study can be used to help assess the potential human health risks associated with exposure to airborne particles from mine tailings.

► Kimberlite pathfinder metals likely adsorb to ferrihydrite under minerotrophic peat groundwater conditions. ► Pathfinder metals likely remain soluble as metal–DOM species under ombrotrophic conditions. ► Metal–DOM complexation in peat groundwater is likely controlled by ionic strength. ► Metal–DOM complexation impacts the spatial distribution of pathfinder metals related to buried kimberlite.The speciation and solubility of kimberlite pathfinder metals (Ni, Nd, Ba and K) in shallow peat groundwaters is investigated over the Yankee, Zulu and Golf kimberlites in the Attawapiskat region, James Bay Lowlands, Canada. The purpose of this study is to examine the relationship between dissolved organic matter (DOM) complexation with kimberlite pathfinder metals and determine the spatial distribution of those metals in shallow peat groundwaters along sampling transects over subcropping kimberlites. Nickel, Nd, Ba and K complexation with DOM and the adsorption of these metals onto ferrihydrite were calculated using Visual MINTEQ 3.0 and the NICA-Donnan database. Calculations predict almost 100% of soluble Nd, Ni and Ba form complexes with DOM at sampling sites with little to no contribution from upwelling groundwater (i.e., dissolved organic C (DOC) concentrations = 40–132 mg/L, pH = 3.9–5.5, and log ionic strength ⩽−3). In only the most ombrotrophic peat groundwater conditions does a majority fraction of K bind to DOM. By contrast, under conditions with large contributions from upwelling groundwaters (i.e., DOC concentrations ⩽40 mg/L, pH = 5.5–6.5, and log ionic strength = −3 to −2), as little as 10% of Nd and Ni, and 0% K and Ba are predicted to complex with DOM. The modeling calculations suggest the dominant control on metal–DOM complexation, particularly with respect to Ni and Nd, is competitive effects for DOM binding sites due to elevated ionic strength where there is evidence of strong groundwater upwelling. Visual MINTEQ modeling of metal adsorption on ferrihydrite surfaces predicts that under strong upwelling conditions, Ni and Nd are scavenged from solution due to increased ferrihydrite precipitation and decreased fractions of metals complexed with DOM. Analytical geochemical data are consistent with model predictions of metal adsorption on ferrihydrite. Total dissolved Ni and Nd concentrations at sites of strong upwelling are up to five times lower than waters with little to no upwelling and log ferrihydrite saturation indices (logSIferr) indicate precipitation (values up to 5) at sites of strong groundwater upwelling. Where the majority of Ni and Nd complex with DOM and ferrihydrite is highly under saturated (logSIferr  = −18 to −5), the concentrations of total Ni and Nd are elevated compared to other sites along sampling transects. Metal complexation with DOM effectively inhibits metal scavenging from solution via adsorption and/or from forming secondary mineral precipitates. Also, because alkaline earth metals do not compete strongly with Ni and Nd for adsorption sites on ferrihydrite surfaces, but do compete strongly for insoluble organic sites, Ni and Nd are more likely to adsorb onto ferrihydrite.

► Evidence of Fe(II) in <200 nm–>10 kDa fraction in anoxic groundwaters and soil waters. ► Evidence for independent speciation of Fe and As under natural anoxic conditions. ► Control of the organic matter on Fe(II) speciation in natural anoxic conditions. ► Predictive modelling of mobility and bioavailability of arsenic in As-affected areas.Arsenic speciation in groundwater and interstitial waters from soil irrigated with As-rich groundwater in Bangladesh was investigated through a size fractionation approach performed in the field and under conditions that allowed the preservation of natural anoxic conditions. Based on a simple chemical characterization, the presence of Fe(II) in the colloidal fraction (<200 nm–>10 kDa) in anoxic groundwaters and soil waters of flooded paddy fields was assessed. Arsenic was found to be present mainly as a ”solute” oxyanion as indicated by the small amount of As associated with the 200 nm–10 kDa colloidal fraction (most of the time, less than 10 ± 5% of total As). When As-colloids were observed, they were linked to high Fe concentration and the As-colloid percentage was relatively well correlated with the total Fe concentration in solution. This study underlined that under natural anoxic conditions, there was no impact of PO 4 3 - , SiO 4 4 - and Mn on the As-bearing colloids, which is not in agreement with results obtained under laboratory controlled conditions. Therefore, it is suspected that the impact of other naturally occurring ligands, such as organic matter, could modify Fe(II) speciation, and, in turn, As speciation. To assess the complexation of Fe(II) by organic matter, the WinHumicV model was used to calculate Fe(II) speciation. Using the standard parameters to describe organic matter reactivity, results showed that organically-bound Fe(II) ranges from 5% to 100% and 30% to 100%, for groundwater and soil water, respectively, depending on the pH and the Fe(II) and DOC contents. Evidence for independent speciation of Fe and As under natural anoxic conditions is of great importance for predictive modelling of the mobility, the availability and then the bioavailability of As in As-affected areas.

► Biogenic zinc sulfides exhibit short-range structural order. ► This structural order increases resistance to re-oxidation. ► Dissolved Fe and Zn formed distinct sulfide phases.The biologically mediated formation of metal sulfide precipitates in anoxic sediments represents a potentially important mechanism for the sequestration of toxic metals. Current knowledge of the structure and reactivity of these biogenic metal sulfides is scarce, limiting the ability to effectively assess contaminant sequestration in, and remobilization from, these solids. In this study, SO4-reducing bacteria (Desulfovibrio sp.) were grown for 5 days in a high-SO4, minimal metal media amended with Zn at either 30 or 300 micromolar. Zinc speciation in the reactor solids was determined using X-ray absorption spectroscopy, and the results compared to spectra of known metal sulfide mineral phases and freshly formed metal sulfides synthesized through purely chemical processes. Biogenically mediated Zn sulfides showed significantly more short range crystallographic order than the abiotically prepared amorphous precipitates. The presence of dissolved Fe2+ at similar concentrations did not affect the nature of the Zn precipitates formed. The biogenic ZnS solids were also more resistant to re-oxidation than the chemical precipitates but more soluble than sphalerite mineral samples. These results suggest that Zn sulfides formed in anaerobic sediments are likely to be more resistant to re-oxidation than would be expected based on dissolution of Fe sulfides and/or sediment acid volatile sulfides.

Hanford tank residual waste – Contaminant source terms and release models by William J. Deutsch; Kirk J. Cantrell; Kenneth M. Krupka; Michael L. Lindberg; R. Jeffery Serne (1681-1693).
► Residual waste from five Hanford spent fuel process storage tanks was evaluated. ► Gibbsite is a common mineral in tanks with high Al concentrations. ► Non-crystalline U–Na–C–O–P ± H phases are common in the U-rich residual. ► Iron oxides/hydroxides have been identified in all residual waste samples. ► Uranium release is highly dependent on waste and leachant compositions.Residual waste is expected to be left in 177 underground storage tanks after closure at the US Department of Energy’s Hanford Site in Washington State, USA. In the long term, the residual wastes may represent a potential source of contamination to the subsurface environment. Residual materials that cannot be completely removed during the tank closure process are being studied to identify and characterize the solid phases and estimate the release of contaminants from these solids to water that might enter the closed tanks in the future. As of the end of 2009, residual waste from five tanks has been evaluated. Residual wastes from adjacent tanks C-202 and C-203 have high U concentrations of 24 and 59 wt.%, respectively, while residual wastes from nearby tanks C-103 and C-106 have low U concentrations of 0.4 and 0.03 wt.%, respectively. Aluminum concentrations are high (8.2–29.1 wt.%) in some tanks (C-103, C-106, and S-112) and relatively low (<1.5 wt.%) in other tanks (C-202 and C-203). Gibbsite is a common mineral in tanks with high Al concentrations, while non-crystalline U–Na–C–O–P ± H phases are common in the U-rich residual wastes from tanks C-202 and C-203. Iron oxides/hydroxides have been identified in all residual waste samples studied to date. Contaminant release from the residual wastes was studied by conducting batch leach tests using distilled deionized water, a Ca(OH)2-saturated solution, or a CaCO3-saturated water. Uranium release concentrations are highly dependent on waste and leachant compositions with dissolved U concentrations one or two orders of magnitude higher in the tests with high U residual wastes, and also higher when leached with the CaCO3-saturated solution than with the Ca(OH)2-saturated solution. Technetium leachability is not as strongly dependent on the concentration of Tc in the waste, and it appears to be slightly more leachable by the Ca(OH)2-saturated solution than by the CaCO3-saturated solution. In general, Tc is much less leachable (<10 wt.% of the available mass in the waste) than previously predicted. This may be due to the coprecipitation of trace concentrations of Tc in relatively insoluble phases such as Fe oxide/hydroxide solids.

► Analyses of biomarkers from diatoms and dinoflagellates. Oils and rocks come from the East Sirte Basin.The present paper involves a detailed evaluation of specific steroid biomarkers by gas chromatography–mass spectrometry (GC–MS) and GC-metastable reaction monitoring (MRM) analyses of several crude oils and source rocks from the East Sirte Basin. 24-Norcholestanes, dinosteranes, 4α-methyl-24-ethylcholestanes and triaromatic steroids have been identified in both source-rocks and crude oils of the East Sirte Basin. Diatoms, dinoflagellates (including those potentially associated with corals) and/or their direct ancestors are amongst the proposed sources of these biomarkers. These biomarker parameters have been used to establish a Mesozoic oil–source correlation of the East Sirte Basin. Hydropyrolysis of an extant coral extract revealed a similar distribution (although immature) of dinosteranes and 4α-methyl-24-ethylcholestanes also observed in the Sirte oils and source-rocks. This is consistent with the presence of dinoflagellates present during the deposition of the Mesozoic aged East Sirte Basin Formations.A good data correlation for the rock extracts revealed a similar distribution of 3,24-dimethyl triaromatic steroids, 3-methyl-24-ethylcholestanes, 4-methyl-24-ethylcholestanes and 2-methyl-24-ethylcholestanes observed in one of the oil families and associated source-rocks for the East Sirte Basin.

Weathering rind formation in buried terrace cobbles during periods of up to 300ka by H. Yoshida; R. Metcalfe; S. Nishimoto; H. Yamamoto; N. Katsuta (1706-1721).
► Weathering rinds in sandstone and basalt cobbles buried for up to 300 ka have been investigated. ► The aim was to determine the formation process and elemental mass balances during rind development. ► Elemental mass balances across the rinds were determined by using open system mass balance (τi , j ) calculations. ► The formation rates are slower than the tropical areas due to the lower rainfall in the studied area.Weathering rinds formed in Mesozoic sandstone and basalt cobbles buried in terrace deposits for up to 300 ka have been investigated. The aim was to determine the formation process and elemental mass balances during rind development. The ages of terraces distributed in the western part of Fukui prefecture, central Japan have been determined as 50 ka, 120 ka and 300 ka based on a tephro-stratigraphic method. Detailed investigations across the weathering rinds, consisting of microscopic observations, porosity measurements, and mineralogical and geochemical analyses using X-ray diffractometry (XRD), X-ray fluorescence (XRF), secondary X-ray analytical microscopy (SXAM), scanning electron microanalyser (SEM) and electron probe microanalysis (EPMA) have been carried out. The results revealed that the Fe concentrations in the weathering rind of a basalt cobble slightly decreased from the cobble’s surface (rim) towards the unweathered core. In contrast, in a sandstone cobble formed under the same environmental conditions over the same period of time there is an Fe-rich layer at some distance below the cobble’s surface. Elemental mass balances across the rinds were determined by using open system mass balance (τi,j ) calculations and show that the Fe was precipitated as Fe-oxyhydroxides in the basalt cobbles, although Fe was slightly removed from the rims. In sandstone cobbles, on the other hand, Fe migrated along a Fe concentration gradient by diffusion and precipitated as Fe-oxyhydroxide minerals to form the weathering rinds. Presumably, precipitation was due to the relatively higher pH conditions caused by mineral dissolution within the pores, principally involving calcite, but probably also silicates including feldspar. The detailed characterization of the weathering rinds revealed the influence of lithology on the accumulation and dissolution of Fe-oxyhydroxides, causing weathering rinds with different characteristics to develop in different kinds of buried cobbles under the same conditions. Relatively large climatic changes in the study area did not cause discernable variations in the mean formation rates of the studied rinds, which were in the order of 10−8 m/a for both basalt and sandstone cobbles. These rates are 1–2 orders of magnitude slower than those reported for tropical areas elsewhere, most probably due to the lower rainfall in the studied area.

Investigation of rock-to-water release and fate of major, minor, and trace elements in the metabasalt–serpentinite shallow aquifer of Mt. Reventino (CZ, Italy) by reaction path modelling by Carmine Apollaro; Luigi Marini; Teresa Critelli; Donatella Barca; Andrea Bloise; Rosanna De Rosa; Francesca Liberi; Domenico Miriello (1722-1740).
► Rock-to-water release and fate of elements from metabasalts and serpentinites, with emphasis on Cr(VI). ► Role of amphibole as Cr source, a fact which was rarely recognized in previous papers. ► Reconstruction of Cr(VI) speciation in the aqueous solution, based on the stability constants of . ► Significant amounts of the neutral complexes CaCrO 4 o and MgCrO 4 o , more mobile and more bio-available than charged solutes. ► Correspondence between the results provided by the Double Solid Reactant Method and the Ideal Solid Solution Approach.The progressive dissolution of metabasalts and serpentinites hosting the shallow aquifer of Mt. Reventino was simulated by means of the EQ3/6 software package, version 8.0, adopting both the Ideal Solid Solution Approach (ISSA) and the Double Solid Reactant Method (DSRM), which provided comparable results. A detailed field and laboratory study was performed on rock samples and local groundwaters to constrain and validate reaction path modelling. The prevalence of Ca–HCO3 over Mg–HCO3 compositions suggests that groundwaters interact chiefly with metabasalts and secondarily with the less abundant and less permeable serpentinites. The most important and active Cr source is a tremolite-rich amphibole, whose role as supplier of dissolved Cr has rarely been recognized in previous studies carried out in areas where ophiolitic rocks crop out. Speciation calculations indicate that hexavalent dissolved Cr is mainly present as chromate ion, followed by the neutral complexes CaCrO 4 o (14–32 mol%) and MgCrO 4 o (2–12 mol%), which are more mobile and more bio-available than charged solutes.All dissolved trace elements are supplied to shallow groundwaters by gradual dissolution of local rocks and, therefore, contributions linked to anthropogenic sources can be ruled out. In particular: (i) Ni is chiefly contributed to the aqueous phase by the tremolite-rich amphibole; (ii) different amounts of Sr, Ba, and Pb are provided by calcite dissolution (with Sr ≫ Ba > Pb), whereas the solid solution of orthorhombic carbonates acts as sink for these trace elements (with X strontianite  ≫  X witherite  >  X cerussite); (iii) the principal source of Cu and Zn is again calcite, whereas the solid solution of trigonal carbonates represents their major sink.

► To constrain the recharge areas and flow paths of SGD in Rishiri Island. ► To determine the local meteoric water lines of Rishiri Island. ► An altitude effect of −0.12‰ per 100 m for δ18O was found. ► In temperate zone, the importance of winter precipitation in groundwater recharge. ► The submarine groundwater discharge (SGD) had a deeper flow system and longer flow paths than did spring water.An isotopic and chemical study was conducted on precipitation, spring water, streams, groundwater wells and submarine groundwater discharge (SGD) to constrain the recharge areas and flow paths of SGD. The isotopic values of precipitation were used to determine the local meteoric water lines (LMWLs) of Rishiri Island. The d-excess values of precipitation showed seasonal variation, with lows of 2.5‰ in the summer and highs of 24.2‰ in the winter. The d-excess values of spring water, streams, groundwater wells and SGD ranged from 12.5‰ to 23.0‰, indicating that the resulting waters were a mix of two seasons of precipitation. The isotopic composition of the groundwater wells sampled along the coast and SGD showed more negative values than that of the spring water sampled along the coast. This indicated that SGD recharged at high altitudes and flowed into the sea. The isotopic and chemical composition of SGD indicated unidirectional flow from land to sea.

► The solubility of schwertmannite and hydrobasaluminite has been re-evaluated. ► These two minerals are major controls on FeIII and Al concentration in acid pit lakes. ► The obtained log  K Sch (18.8 ± 1.7) matches with the value proposed by . ► The obtained log  K Hyb (23.9 ± 0.7) is close to values reported for aged basaluminite. ► These log  K values accurately predict the behaviour of FeIII and Al in acid pit lakes.The solubility of schwertmannite and hydrobasaluminite and their control on the concentration of Fe(III) and Al in acid-sulfate aqueous systems has been re-examined through a series of titration experiments with waters from two acidic pit lakes (Cueva de la Mora and San Telmo) from the Iberian Pyrite Belt in SW Spain. The work also includes a microscopic study (SEM–EDS) of natural Fe(III) precipitates found in the water column of both lakes. The microscopic study of natural precipitates confirms that schwertmannite constitutes a major control of Fe3+ concentration in the lakes at pH ∼ 3.0, whereas jarosite may also be abundant at near-surface conditions and is typical of low pH (<2.5). Goethite is eventually found, although it basically results from ageing of less stable precursor phases. The titration study comprised chemical analysis of waters sampled at increments of 0.2 pH units in the pH range 2.6–10.0, along with chemical (ICP-AES, XRF) and mineralogical (XRD, SEM, EDS, TEM) analysis of the resulting precipitates at different pH values (3.5, 4.5, 4.8, 5.2). The experimental results indicate that the precipitation of a highly hydrated schwertmannite (with empirical formula Fe8O8(SO4) x (OH) y ·nH2O, where x  = 1.4–1.5, y  = 5.0–5.2, and n  = 13–17) strongly buffers the solutions in the pH range of 3.0–3.5 and appears to control the aqueous concentration of Fe3+ up to pH ∼ 5. Precipitation of Al3+ also constitutes an important buffering system at pH 4.2–4.7 and is controlled by the formation of poorly crystalline hydrobasaluminite (with empirical formula Al4(SO4)(OH)10·15H2O), which is highly unstable in low relative humidity conditions and tends to dehydrate to basaluminite (Al4(SO4)1.2(OH)9.6·9–10H2O). Removal rates for Fe3+ and Al3+ by precipitation of these two minerals approached 96–98% at pH 3.5 and 5.2, respectively. The obtained compositional stoichiometries, along with computation of the activities of Fe3+, Al3+, SO 4 2 - and H+ in solution, allowed the calculation of ionic activity products (log IAP) and solubility product constants (log  K sp) for the precipitating phases. The solubility product constants have been deduced by two independent approaches. The first one averages ionic activity products obtained for a given pH range, whereas the second one deduces log  K sp values from linear regression lines in ion activity-pH plots. Both methods tend to converge and give log  K sp  = 18.8 ± 1.7 for schwertmannite, and log  K sp  = 23.9 ± 0.7 for hydrobasaluminite. In the pH range 5–9, the results are compatible with additional solubility controls from ferrihydrite and an amorphous to nano-crystalline Al(OH)3 phase on the aqueous concentrations of Fe and Al, respectively, although formation of these latter compounds could not be demonstrated. The present work confirms the results originally obtained for schwertmannite solubility and reports a reliable solubility product constant for freshly precipitated hydrobasaluminite. Integration of the resulting log  K sp values in the MINTEQ.V4 database and the PHREEQCI geochemical modeling program provides a precise description of the geochemical behaviour of Fe and Al in these acidic pit lakes.