Applied Geochemistry (v.16, #9-10)

Stable isotope ratios were used as a tracer for S flow and transformations in an irrigation experiment with 5 different German forest soils. Seventy-five lysimeters constructed from soil cores, 15 from each site, were irrigated over 20 months with SO4-rich artificial canopy throughfall, simulating 3 different S input levels: 35 kg S ha−1 in treatment I, 63 kg S ha−1 in treatment II, and 131 kg S ha−1 in treatment III. The δ34S value of the irrigation SO4 was more than 22‰ higher than those of total S in the untreated soils. Mass and isotope balances for different soil S compounds were used to assess the patterns and mechanisms of S retention in individual soil horizons and their dependence on S deposition levels. Independent of the S deposition level, on average 12±5 kg ha−1 of the applied S were bound organically by the microbial biomass in all soils. Immobilization of irrigation SO4 occurred predominantly in the topsoil horizons with the formation of C-bonded S being more prevalent than the synthesis of organic sulfates. Tracer retention via formation of organic soil S compounds accounted for up to 50% of the irrigation SO4 in treatment I, from 16 to 25% in treatment II, and less than 20% in treatment III. The dominant process of inorganic S retention in the soils appeared to be adsorption of SO4, but precipitation of aluminum hydroxy sulfate minerals constituted a second potential inorganic retention process in some soils. Sulfate adsorption increased with increasing sesquioxide content of the soils and with increasing S deposition rates. In soils with high sesquioxide contents, typically more than 70% of the irrigated S was retained inorganically, whereas in the soil with the lowest sesquioxide content, generally less than 50% of the labeled irrigation S was detected in inorganic form. In the latter soil, the sesquioxide content was not high enough to fully adsorb the elevated SO4 inputs in treatments II and III. Consequently, increased tracer S export with the seepage water SO4 was observed in the experimental variants with elevated SO4 deposition rates. In soils with high sesquioxide contents, the elevated SO4 inputs in treatments II and III were fully retained in the soil horizons in inorganic form during the 20 months of the experiment and thus increased seepage water export of labeled SO4 was not observed. The ability to inorganically retain tracer S in the mineral soil horizons was identified as the major factor regulating the extent of tracer S export with the seepage water at 60 cm depth. The high retention of labeled S in all soils combined with the comparatively low recovery of irrigation SO4 with the seepage water implies that the mean transit time of S in the uppermost 60 cm of the acid forest soils varies between several years and many decades, much longer than previously thought.

A set of chlorinated hydrocarbons (TCE, PCE, DCM, 1,1,1-TCA, chloroform) provided by four manufacturers has been isotopically characterised for both C and Cl, using a new sensitive method. A very large range of δ 13C (from −51.66 to −24.07‰/PDB) associated with a very large range of δ 37Cl (from −2.7 to +3.4‰/SMOC) was obtained. This range of δ 37Cl is much larger than that of inorganic Cl (±1‰ SMOC) and most individual solvents show a very distinct δ 37Cl compared to inorganic Cl isotopic signatures. Moreover, δ 37Cl/δ 13C pairs are distinct from one solvent/manufacturer to another. In a δ 13C versus δ 37Cl diagram, δ 37Cl / δ 13C pairs show different trends for the products of a single manufacturer compared to another. This suggests that Cl isotopic compositions are probably highly fractionated during organic synthesis. The δ 37Cl values can be interpreted in terms of the probable manufacturing processes. Unlike the data published previously, with one exception, all the new results for samples reported here have positive δ 37Cl values which might differentiate natural Cl from that derived from degradation. This method has significant potential as a tool for investigating environmental pollution problems; in particular, it offers the possibility for validating models of transport and fate of pollutants.

CO2 gas pools in Jiyang sag, China by Leping Zheng; Shijie Wang; Yongsheng Liao; Zujun Feng (1033-1039).
The CO2 gas pools of Jiyang sag are located along the Gaoqing–Pingnan fault within a region of alkaline basalts. The concentration of CO2 in the gas pools is in the range of 68.85–96.99%. All of the geochemical tracers for the CO2 gas pools support the suggestion that CO2 was mainly derived from mantle degassing. The δ13C values of CO2 in the gas pools are in the range of −5.67–−3.41‰, which are higher than those of organogenic CO2, and near to those of abiogenic CO2. Their 3He/4He ratios are 2.80–4.47×106, i.e. the R/Ra ratios are 2.00–3.19, showing that the Jiyang sag had undergone strong mantle degassing. CO2/3He ratios are 0.59–0.89×109, which are identical to those for N-MORB, indicating that CO2 in these CO2 gas pools was mainly derived from the mantle. Accompanying the intrusion of mantle-derived magma, the mantle-derived CO2 migrated upwards along deep faults and was trapped in advantageous structures forming gas pools.

The dilution factors (D i ) and removal fractions (R i ) of pollutants from acid mine drainage (AMD) were quantitatively estimated using two different methods, the conservative component and mass balance method, along Imgok Creek in Korea. The conservative component method assumes that SO4 is a perfectly conservative component and calculates D i and R i from the concentration ratios of SO4. The mass balance method solves the simultaneous equations relating the concentrations of dissolved components to their precipitation stoichiometries to obtain D i and R i . The results from both methods are little different, indicating that SO4 concentration is a good indicator of dilution for Imgok creek. The calculated D i 's of pollutants quickly decrease from the site of AMD input to the site a few km downstream, but then remain more or less constant over the reaches farther downstream. This is because D i loses its sensitivity in the reaches where difference in SO4 concentration between the main stream and combining tributaries significantly diminishes. The calculated R i 's show that approximately 90, 95, and 75% of the original Fe input were removed from the streamwater in October 1996, April 1997, and October 1997, respectively. Aluminum was almost completely removed in April 1997, but only 50% of the original Al was removed in October 1997. The removal of Fe was due to the precipitation of schwertmannite or ferrihydrite and Al due to amorphous Al4(OH)10SO4. The maximum removal fraction of dissolved SO4 was only 5%. The other metals from AMD were significantly removed from the stream water only in April 1997. These metals were removed not by precipitation but by adsorption on and/or coprecipitation with Fe/Al-compounds. The relatively abundant freshwater supply in April 1997 might raise stream pH higher than the adsorption edge and consequently, contribute to rapid metal attenuation by forcing not only more precipitation but also more adsorption of the dissolved metals.

Laboratory hydrothermal experiments provide unique information regarding the fate of volatile and/or incompatibles (e.g. B, Li, and As) during oceanic crust subduction. Examination of chemical redistribution between the subducted slab, mantle wedge, arc volcanics and overlying ocean water during subduction is critical to gain further insight into arc volcanism and chemical oceanic budgets. For instance, efficient mobilization of B at shallow depths may be a key aspect of its oceanic budget, and can help to explain the systematics of B-δ11B and B-10Be in arc lavas. Fluid–rock interactions at elevated temperatures and pressures in accretionary prisms were studied using a rocking autoclave hydrothermal apparatus to monitor sediment–porewater interaction over the range of 25 to 350°C, at 800 bars. Clay-rich hemi-pelagic sediments from the décollement zone of Ocean Drilling Program Site 808, Nankai Trough, were reacted with NaCl–CaCl2 solutions at water/rock ratios of ∼3.5 to 1.5 (w/w) to mimic alteration processes in the shallow subduction zone. Fluids were extracted at 25–50°C intervals and were analyzed for major and trace chemical constituents. The fluid chemistry changed significantly during the course of these experiments, but there was generally only minor modification of the solid phase; only B, Li, As, Br, and Pb are significantly depleted. During the heating cycle, dissolved Na, Mg and SO4 decreased sharply and NH4, SiO2, K, B, and Li increased at T⩾300°C. Calcium drops gradually at low temperatures, but concentrations rise sharply at T⩾300°C. Decomposition of organic matter, SO4 2− depletion, and Mg-fixation dominate at low temperatures; however, albitization of calcic plagioclase leads to marked Na depletions and Ca enrichments at T⩾300°C. Dissolved SiO2 remained below saturation with respect to quartz and amorphous silica throughout the entire experiment. B and Li mobilization with large isotopic fractionations occurred at low temperature. Exchangeable B (δ11B=∼15‰) is completely leached before reaching 150°C. Substantial O2 exchange between fluids and the solid phase occurred at T⩾200°C in the spiked experiment, where δ18O varies more than 100‰ in the fluids. During retrograde cooling, dissolved Mg, SO4, Ca, Si, K and Sr are released as a result of carbonate or anhydrite dissolution, and marked B re-adsorption occurred at temperatures below 60°C.

Thermodynamic data for all fate-determining processes are needed in order to predict the fate and transport of metals in natural systems. The surface complexation properties of a synthetic MnO2, δ-MnO2, have accordingly been investigated using glass electrode potentiometry. Experimental data were interpreted according to the surface complexation model in conjunction with the diffuse double layer model of the solid/solution interface. Adsorption constants were determined using the non-linear optimisation program FITEQL. Surface complexation parameters determined in this way were validated against results obtained from the literature. Best fits of alkalimetric titration data were obtained with a 2-site, 3 surface-species model of the δ-MnO2 surface. Site concentrations of 2.23×10−3 mol g−1 and 7.66×10−4 mol g−1 were obtained. Corresponding logarithms of formation constants for the postulated surface species are −1.27 (≡XO), −5.99 (≡YO) and 3.52 (≡YOH2 +) at I=0.1 M. The surface speciation of δ-MnO2 is dominated by ≡XO over the pH range investigated. Metal adsorption was modelled with surface species of the type ≡XOM+, ≡XOMOH, ≡YOM+, ≡YOMOH (M=Cu, Ni, Zn, Cd and Pb) and ≡XOM2OH2+ (M=Pb). For Cu, Ni and Zn, titration data could be modelled with ≡XOM+, ≡XOMOH, ≡YOM+ and ≡YOMOH, whereas for Cd, ≡XOM+ and ≡YOM+ were sufficient. Lead data were best modelled by assuming the dinuclear species ≡XOM2OH2+ to be the only surface species to form. Adsorption constants determined for Ni, Cu and Zn follow the Irving-Williams sequence. The model suggests an adsorption order of (Pb, Cu) > (Ni, Zn) > Cd. The discrepancy between model predictions and published adsorption results is similar to the variability observed in experimental results from different laboratories.

Contamination of deep formation waters by drilling fluids is a problem that concerns most types of drilling operations (petroleum wells, geothermal boreholes) and is crucial in the course of feasibility and safety studies of potential radioactive-waste repositories. Residual contamination of formation-water samples has an important impact on the accuracy of the characterisation of the natural hydrogeochemical background of the study sites. Based on a literature review and on experience acquired on the sites of the French Agency for Nuclear Waste Disposal (ANDRA), this article proposes a general method for the correction of residual contamination, including estimates of the associated errors. The quantification of the contamination is based on tracing techniques and on a geochemical survey during the pumping test preceding sampling. The correction and estimation of errors require repeated measurements of the tracer(s) and the chemical and isotopic species during pumping. The method is applied to a pumping test in a research well in deep granite of the Vienne district (France) where multi-tracing of drilling fluids has been used.

A study of groundwater chemistry and isotopic composition was conducted in Cochabamba Valley, Bolivia of an alluvial fan aquifer system located along the northern edge of the valley. In total, 75 wells were sampled for major and minor (F, Br, Si4+, NO3 , Fetotal, H2S, PO3 4−) ion chemistry, 18O and 2H, and 3H. In the alluvial fan aquifer, groundwater flow patterns appear to be controlled by the unique hydrostratigraphy of the fan sediments. The radial morphology of alluvial fans and the entrenchment of channels in the apex of the fan are two probable causes of deposition of coarser, more permeable material along the axis of the fan. Geochemical and isotopic parameters give indirect evidence of these more permeable zones. Ratios of Na + Ca ++ are higher off the axis of the fan, which is most likely due to longer residence times so that diffusion of saline residual pore water or ion exchange reactions create more Na+-rich groundwater. Also, 3H concentrations down the flow system are higher along the axis of the alluvial fan, a direct indication of younger groundwater age along the axis. The distribution of other minor ions, such as F and Si4+, and the redox parameters NO3 , Fetotal, SO4 2− and H2S suggest a shorter residence times along the axes. The even pattern of Na + Ca ++ and Na + K + ratios, and Si4+, NO3 , Fetotal, SO4 2−, H2S, and 3H concentrations along the flow system from wells varying 20 to 80 m in depth suggest that the aquifer–aquitard alluvial fan system is interconnected between different depths. Enriched groundwater 18O and 2H signatures in the distal end of the flow system in Area A are most likely the result of the introduction of evaporated waters from the surface, further evidence that the system is interconnected.

Ti(IV) hydrolysis constants derived from rutile solubility measurements made from 100 to 300°C by Kevin G. Knauss; Michael J. Dibley; William L. Bourcier; Henry F. Shaw (1115-1128).
Using a Au–Ir hydrothermal reaction cell and dilute buffer solutions to control pH, the isobaric solubility of rutile (TiO2) was measured over a broad pH range from pH 1 to 13 between 100 and 300°C. The solubility data were regressed to derive the cumulative and stepwise hydrolysis constants for Ti(IV) in aqueous solution. The data were accurately modeled using the hydrolysis species: Ti(OH)+ 3, Ti(OH)4(aq) and Ti(OH) 5. The following cumulative (K s3, K s4, K s5) hydrolysis constants (log K and error, reactions written with respect to the solid oxide) and stepwise (K 34, K 45) hydrolysis constants (log K and error) were determined by regression: 100°C 150°C 200°C 250°C 300°C log K s3 −5.44(0.032) −5.42(0.106) −5.28(0.055) −5.03(0.046) −4.61(0.112) log K s4 −7.72(0.150) −7.66(0.202) −7.57(0.153) −7.93(0.202) −7.92(1.11) log K s5 −17.8(0.068) −17.1(0.138) −16.4(0.132) −15.9(0.018) −15.4(0.057) log K 34 −2.28(0.153) −2.24(0.228) −2.29(0.162) −2.90(0.207) −3.31(1.12) log K 45 −10.1(0.165) −9.42(0.245) −8.86(0.202) −7.95(0.203) −7.43(1.11) These data are required to model the chemical behavior of Ti(IV) in aqueous solutions and provide a baseline for the Ti-solubility-limited dissolution of titanates.

Conditions for biological precipitation of iron by Gallionella ferruginea in a slightly polluted ground water by Erik G Søgaard; Remigijus Aruna; Joanna Abraham-Peskir; Christian Bender Koch (1129-1137).
A sand filter has been built as a pilot plant with the purpose of biological precipitation of Fe from ground water polluted with mainly chlorinated aliphatics. The ground water is pumped directly from a well in a polluted ground water aquifer in Esbjerg, Denmark. The pollution includes trichlorethylene and tetrachlorethylene together with smaller amounts of pesticides. Furthermore the best conditions for Fe precipitating bacteria were not expected to be present because of a relatively high O2 content, up to 6.7 mg/l, a low Fe content, 0.2 mg/l and a pH of ∼5 in the ground water. Added FeSO4 increased the Fe content of the ground water to about 4 mg/l. These rather extreme conditions for precipitating Fe were observed over a period of 3 months. The goal of the research was to observe the mechanism of Fe precipitation in a sand filter in the above-mentioned conditions comparative to normal conditions for biotic as well as abiotic Fe mineralization in sand filters of fresh water treatment plants. The Fe precipitating bacterium Gallionella ferrugenia was found to dominate the biotic Fe oxidation/precipitation process despite the extreme conditions. A huge amount of exopolymer from Gallionella was present. The precipitated Fe oxide was determined to be ferrihydrate. The rate of the Fe oxidation/precipitation was found to be about 1000 times faster than formerly found for abiotic physico-chemical oxidation/precipitation processes. The hydrophobic pesticides and some of their degradation products were not adsorbed in the filter. An added hydrophilic pesticide was adsorbed up to 40%. Trichlorethylene was not adsorbed in the filter. The reason for the poor adsorption of the hydrophobic compounds and trichlorethylene is due to the pronounced hydrophilic property of the exopolymers of Gallionella and the precipitated ferrihydrite.

Establishing ‘reference sites’ is a difficult task and a critical factor in determining the baseline functioning of ecosystems. The information thus obtained on nutrient and contaminant background concentrations in turn subsidizes the remediation of impacted landscapes. This paper reports a study on metal (Cr, Cu, Pb, Zn, Mn, Fe and Al) and nutrient (C) distribution in sediments from Capivari River watershed (Praia do Sul Biological Reserve, Ilha Grande, Rio de Janeiro State, Brazil), an area where typical SE Brazilian coastal ecosystems are located. Contrary to what one would expect from the high rate at which these ecosystems have been deteriorating in recent decades, the study site is surprisingly well preserved. The present study was developed to assess variations in heavy metal concentrations in river basin sediments, to identify the geochemical carriers of these elements, and determine the influence of water quality and organic matter on their distribution. Results showed that heavy metal distribution has been influenced by the natural biogeochemical properties of those ecosystems found in an upland-to-lowland sequence in the watershed. Minimum and maximum total concentration were: 5 and 23 mg/kg for Cr; 4 and 29 mg/kg for Cu; 13 and 53 mg/kg for Pb; 24 and 142 mg/kg for Zn; 54 and 342 mg/kg for Mn; 0.8 and 7.2% for Fe; 0.5 and 4.9% for Al; 6.3 and 25% for C. The pH and EH are the key-parameters in explaining total metal concentration decrease in the swamp area, where dissolution processes and losses through metal transport seem important. The most important geochemical carriers are Al in the basin’s “continental” stretch and Fe in the estuarine portion. The data also provide evidence showing that organic matter is the key-parameter in Cu concentration control in the sediments through burial and accumulation processes especially in the swamp area. Heavy metal concentrations in sediments from the study area are generally lower than those found in similar regional ecosystems. Surface enrichment in heavy metal concentrations in collected sediment cores was not observed. The authors therefore conclude that this site is suitable as a ‘reference site’ for studies on the biogeochemistry and ecotoxicology of SE Brazilian coastal ecosystems.

Geochemistry of thermal springs, Alhama de Granada (southern Spain) by M. López-Chicano; J.C. Cerón; A. Vallejos; A. Pulido-Bosch (1153-1163).
The waters of the thermal springs at Alhama de Granada vary in temperature between 27 and 45°C. Temporal changes in the composition of the principal spring (Baños Viejos) indicate that a small degree of mixing may occur between deep thermal waters and shallow groundwater. Slight compositional variations also occur between the various thermal springs in the study area. These spatial variations are due to the different local hydrodynamic conditions in the springs. Towards the north in less hydraulically transmissive rocks, cooling of the rising water is more noticeable, as are ion exchange and processes of SO4 reduction. The chemical composition of the water is related to the dissolution of evaporites (SO4 and Cl salts), carbonates and silicates, and to the possible existence of sources of S within the rock. Estimates of the mean residence times have been obtained based on 14CDIC and T. The state of thermodynamic equilibrium at the spring discharge was calculated using the SOLMINEQ.88 program. The results indicate that all the samples are supersaturated with respect to quartz, chalcedony, cristobalite, calcite, aragonite and dolomite, and undersaturated with respect to gypsum, anhydrite and halite. The use of different geothermometers and modelling of saturation indices for quartz, albite and anhydrite indicate temperatures of about 110°C.

TEM-EDX investigation on Zn- and Pb-contaminated soils by Martine D Buatier; Sophie Sobanska; Françoise Elsass (1165-1177).
Lead and zinc contaminated soils from a smelter area in the northern part of France have been studied by transmission electron microscopy (TEM). This study was carried out with 4 different soils contaminated by Pb and Zn but with different chemical and physical characteristics. Two soils are tilled and have a neutral or slightly basic pH, one is a wooded soil and the last one is a meadow soil with acidic pH and high total organic content . TEM images of the soil samples have been coupled with focused energy dispersive X-ray (EDX) analyses and chemical mapping on a few micron-sized windows. This study demonstrates that TEM is a particularly efficient method to investigate metal speciation in the fine fractions of the contaminated soils. Zinc could be detected locally in sulphide minerals probably coming from the smelter emissions, but the major phases retaining Zn are Fe-oxyhydroxides and smectites. Lead could be detected in small aggregates which were characterized by EDX and selected area electron diffraction. Their structural formulae correspond to a pyromorphite-like mineral in which Pb is partly substituted by Ca and Na. Pyromorphite is present only in the wooded and meadow soils where it forms partly from amorphous Si-rich phases (slags) coming from the smelter. These results are compared with data previously obtained by spectroscopic methods on the same samples.

The Cabra-Alcaide karstic massif situated in the south of Spain constitutes an important part of the so-called Natural Park of the Subbetic Sierras. This aquifer system is drained by various springs which supply a population of some 100,000 inhabitants. The feed areas of these springs show very different characteristics with respect to their geological structure, size of the drainage area, thickness of the vadose zone, elevation and degree of karstification. In addition, the carbonate rocks lie over a clayey substratum which contains large masses of intercalated evaporites. Due to these conditions, the hydrochemical composition of the springs is relatively variable. In this study a hydrogeochemical characterization of the aquifer in both space and time is undertaken and the factors that determine it are discussed. To achieve this, 19 monthly samples were taken from the 6 most significant springs of the hydrogeological system. The commencement of sampling coincided with the transition from a period of several years of severe drought and a very wet two-year period, which amplified considerably the hydrochemical and hydrodynamic response of the springs to the recharge. Identification of hydrogeochemical processes was performed by studying hydrographs, the temporal evolution of physico-chemical parameters, ionic ratios (mainly Mg/Ca) and by means of simple and multivariate statistical analyses. The saturation status was established using the WATEQF program and the mass transfer was quantified using PHREEQC. With the exception of the epikarstic subsystem (i.e. the Zarza spring), the majority of the results indicate that the aquifer exhibits a diffuse flow model, in which piston flow phenomena are seen, closely linked to the most intense precipitation. Along the direction of flow hydrochemical trends are seen as the water type changes from Ca–HCO3 to Ca–Mg–HCO3; at the same time enrichment in some ions, derived from the dissolution of evaporitic minerals of the impermeable substratum, and dedolomitization processes occur. In this way, almost 90% of the water samples were oversaturated in calcite, the majority of cases being a consequence of the dissolution of the gypsum of the substratum. Only in the epikarstic springs can it be considered that the oversaturation in calcite is due to loss of CO2 from the emerging groundwater. It is concluded that hydrodynamic aspects together with hydrogeochemical characteristics need to be taken into account to correctly explain the hydrochemical evolution of the karstic springs. Moreover, the use of both approaches permits a more accurate establishment of the degree of aquifer karstification, which in turn needs to be known in order to assess its vulnerability to contamination and to protect recharge zones.

Examination of calcareous slags from several historical smelting sites has indicated that the specific soil environment, in particular the soil pH, may have a very significant effect on the rate of weathering and metal release. A series of acid titration experiments were used to investigate whether the weathering of the calcareous slags could be increasing the buffering effect of the soils through accumulating CaCO3 in the slag-rich horizons. Such a buffering mechanism would maintain high pH levels and so limit the migration of Pb through the soil profile. Three sites were chosen; one with high levels of Ca in the soil, one with relatively low Ca levels and one with intermediate Ca levels. Analysis of metal concentrations was determined using ICP-AES. The results support the hypothesis that, while the soil pH remains between 8 and 5, the CaCO3 provides an effective buffer against the mobilisation of Pb. Between soil pH 5 and 4 it is suggested that both CaCO3 and PbCO3 participate in the buffering reaction, which slows down with a further drop in pH. However, this reaction ultimately releases Pb into the soil solution, although at a much slower rate than would be the case in an unbuffered soil. An important implication of these findings is that migration rates of metals in soils cannot be assumed to be constant over time, if such buffering mechanisms are operational.

Iron speciation in modern sediment from Erhai Lake, southwestern ChinaRedox conditions in an ancient environment by Zheng Guodong; Bokuichiro Takano; Akihito Kuno; Motoyuki Matsuo (1201-1213).
In a core of sub-aquatic sediment from Erhai Lake, southwestern China, 4 Fe species were identified as paramagnetic Fe3+, superparamagnetic Fe3+, hematite Fe3+, and paramagnetic high-spin Fe2+ using Mössbauer spectroscopy. The 120 cm core has a distinct lithological boundary at a depth of about 70 cm. Each Fe species has a distinctive distribution with depth. These results represent the redox conditions within the sediment, and also probably reflect the primary sedimentary environments. With increasing burial depth, hematite (Fe2O3) decreased, especially below depths greater than 25 cm, and finally disappeared at around 95 cm. The summed paramagnetic Fe3+ (superparamagnetic Fe3++paramagnetic Fe3+) did not change as much, only exhibiting a slight decrease at depths greater than 75 cm, about 5 cm beneath the lithological boundary within the core. The intensity of paramagnetic high-spin Fe2+ increased with depth. These vertical variations were in harmony with organic geochemical parameters such as TOC concentration, H-index and O-index, indicating that reducing conditions are strongly intensified in the sediment below 70 cm. The geological, organic geochemical and 14C data combined with the present Mössbauer spectroscopic study give a strong indication that the redox environment of Erhai Lake probably shifted rather rapidly from a deep reducing to a shallow oxic state at about 2 ka ago.

Oxidation rates of low sulphide (<0.5 wt.%) gneissic waste rock from the Cluff lake U mine, northern Saskatchewan, Canada were determined using 3 independent methods: O2 consumption rates in kinetic cells, SO4 measurements of kinetic cell effluent and humidity cell SO4 release rates. The O2 consumption measurements demonstrated that the oxidation of pyrite was strongly dependent on grain size and moderately dependent on water content, temperature and microbiology. Oxygen consumption rates were highest at water contents of 5–10 wt.% (12–25% saturation). Measured SO4 release rates (3.1–91 mg SO4 kg−1 wk−1) for the kinetic cells were comparable to rates calculated from the O2 consumption values (6.9–70 mg SO4 kg−1 wk−1). Sulphate release rates determined from humidity cells were generally higher than those obtained from the kinetic cells, ranging from 6 to 64 mg SO4 kg−1 wk−1 for the coarsest and finest fraction, respectively. These differences were attributed to sample heterogeneity.

Evaporative chemical evolution of natural waters at Yucca Mountain, Nevada by N.D Rosenberg; G.E Gdowski; K.G Knauss (1231-1240).
The authors report results from experiments on the evaporative chemical evolution of the two major types of natural waters present at Yucca Mountain, Nevada. The first is represented by J13 well water, a dilute Na–HCO3–CO3 water similar to saturated horizons in volcanic tuffs across the western United States. The second is represented by Ca–C1–SO4-rich pore water that has a higher dissolved ion content, from the unsaturated zone above the repository horizon. Data include anion and cation analysis and qualitative mineral identification for a series of open system experiments, with and without crushed tuff present, conducted at sub-boiling temperatures (75–85°C). This work is motivated by a need to characterize the chemistry of concentrated aqueous films that might form on engineered components at the potential high-level, nuclear-waste repository at Yucca Mountain, Nevada and lead to electrochemical corrosion. The experiments indicate the evolution of high pH, Na–HCO3–CO3 brines from J13-like waters and the evolution of near-neutral pH brines from the pore water compositions. The minerals recovered after complete evaporation of the J13 water alone experiments include amorphous silica, aragonite, calcite, halite, niter, thermonatrite and, possibly, gypsum, anhydrite and hectorite. Tachyhydrite and gypsum were observed to form in equivalent experiments with pore waters from the unsaturated zone.

A two-column reactor was designed to remove dissolved As and Cd from contaminated water. The reactor functions by equilibrating the targeted water with CO2 and directing it via saturated flow through a column of crushed siderite. This results in siderite dissolution and an increase in dissolved Fe(II). The feedwater is then directed into the top of a second, aerated column of crushed limestone, where it passes by unsaturated flow. The Fe2+ ion oxidizes quickly to Fe3+ and precipitates as Fe(III) oxyhydroxide, which is an effective sorbent of AsO4 3−. The aeration that occurs in the second column also removes dissolved CO2 from the feedwater. This causes precipitation of Ca and Cd carbonates. Together, the two processes reduce As and Cd concentrations from 1 and 3 mg/l, respectively, to below detection (respectively <0.005 and <0.01 mg/l). A time-limited reduction in Cr concentration also occurred. Much of the As was removed in the first column of the reactor, because Fe(III) oxyhydroxides also formed there. This was due to oxidation of Fe(II) by Cr(VI) and other oxidants present in the input wastewater. Although As is removed in the reactor columns by a sorption mechanism, the sorbent responsible, Fe(III) oxyhydroxide, is continuously produced during the operation of the reactor. Thus, unlike attenuation in a system with a fixed amount of sorbent, breakthrough of the As contaminant should never occur.

Sediment samples were taken along the West, North, and East rivers of the Pearl River system at 28 locations in 1998, and a total of 49 elements were determined by ICP–AES, ICP–MS and INAA. The probability features of the datasets were studied, and the average concentrations of these elements in sediments of the three rivers were calculated. Significant differences in element concentrations among the three rivers were observed and the results were confirmed by statistical tests including analysis of variance (ANOVA), Kruskal–Wallis test, and t-test. Spatial distribution maps of element concentrations were produced using a geographical information system (GIS). The immobile trace elements (such as Sc, Ti, V, Cr, Mn, Fe, Co, and Ni) are enriched in the West River where limestone dominates the rock types in the watershed. Because of the strong weathering, immobile trace elements are enriched and reside in secondary minerals of the weathering products. All three rivers have high concentrations of rare earth elements (REEs) because of strong weathering, but relatively higher concentrations of REEs are observed in sediments of the East River where granite dominates the rock type. Granite contains high concentrations of REEs and the sediments have inherited this feature from their bedrock. Alkaline element (Li, Na, K, Rb, and Cs) concentrations are elevated in sediments of the East River, these may reside in granitic primary minerals. Relatively high concentrations of alkaline earth elements (Mg, Ca, Sr, and Ba) are observed in the West River, inherited from the limestone bedrock. High Pb and Bi concentrations are found in the North River and are caused by Pb mineralization and the discharge of a smelter in the upper reaches of the river. However, statistical tests did not indicate a significant difference between Pb concentrations in the North River and the other two rivers, which suggests that statistical results should be carefully used and explained.

Reaction with periodic acid-Schiff reagent (PAS), a modified histochemical staining procedure, was used as an index of carbohydrate content of natural waters. Material derived from the standard method for extraction of soluble humic and fulvic acids (HFA) from natural waters, namely adsorption onto and subsequent elution from DEAE cellulose and XAD resin, showed 5–31% of its C content to be carbohydrate. On passage of pure polysaccharide through a DEAE column, 88% of the loading was retained and 50% was subsequently eluted into what would be classified as the HA fraction. Clearly DEAE extraction cannot be regarded as specific for HFA. Similarly, samples were analysed from 62 lentic (non-flowing) waters of varied chemistry and trophic status in NW England. 22%±5% (mean±S.D.) of dissolved organic C (DOC) was PAS+. The proportion of the DOC that reacted positively to the PAS test exceeded 30% of total DOC in sites where DOC concentration was low (<7 mg/l). There was no correlation in the concentrations of DOC and PAS+C suggesting the components of the former have independent sources. Aluminium and Si concentration correlated negatively with PAS+ concentration whilst exhibiting no correlation with DOC. Consequently, prediction of metal speciation in natural waters requires separate consideration of metal interaction with PAS+ and other components of DOC, such as HFA.

Chemistry of major and minor elements, 87Sr/86Sr, δD, and δ18O of oilfield waters, and 87Sr/86Sr of whole rock were measured from Paleozoic strata in the Central Tarim basin, NW China. The aim is to elucidate the origin and migration of formation water and its relation to petroleum migration. High salinity oilfield waters in Carboniferous, Silurian and Ordovician reservoirs have maintained the same Na/Cl ratio as seawater, indicative of subaerially evaporated seawater. Two possible sources of evaporitic water are Carboniferous (CII) and Cambrian, both of which contain evaporitic sediments. Geographic and stratigraphic trends in water chemistry suggest that most of the high salinity water is from the Cambrian. Strontium, H and O isotopes as well as ion chemistry indicate at least 3 end member waters in the basin. High-salinity Cambrian evaporitic water was expelled upward into Ordovician, Silurian and Carboniferous reservoirs along faults and fractures during compaction and burial. Meteoric water has likely invaded the section throughout its history as uplift created subaerial unconformities. Meteoric water certainly infiltrated Silurian and older strata during development of the CIII unconformity and again in recent times. Modern meteoric water enters Carboniferous strata from the west and flows eastward, mixing with the high salinity Cambrian water and to a lesser degree with paleometeoric water. The third end member is highly radiogenic, shale-derived water which has migrated eastward from the Awati Depression to the west. Enrichment of Ca and Sr and depletion of K, Mg, and SO4 relative to the seawater evaporation trajectory suggest waters were affected by albitization of feldspars, dolomitization, illitization of smectite, and SO4 reduction. The mixing of meteoric water occurred subsequently to seawater evaporation, main water-rock interactions, and brine migration. The direction of brine migration is consistent with that of petroleum migration, suggesting water and petroleum have followed the same migration pathways.

Introduction to Ground Water Geochemistry by Yousif K. Kharaka (1285-1286).

Santorini Volcano, Geological Society Memoir Number 19 by Nicholas J.G Pearce (1286-1287).