Recent experimental and computational work has highlighted the fact that the attachment and detachment of ions at mineral surfaces can be a strong function of the atomic-scale surface topography. This atomic configuration depends on both crystallography and on the history of the surface. Along with information about surface structure, ultra-sensitive methods for specific element detection are needed, especially for single crystal work, because of the low number of atoms generally involved in adsorption, leaching, and cation exchange. Therefore, the capability of performing in situ, crystallographically-resolved, depth-sensitive measurements with low detection limits is desired. Glancing incidence surface analysis techniques can provide much of the required information. For these reasons, we have completed glancing incidence X-ray fluorescence and proton microprobe measurements with mineral substrates in an attempt to develop a quantitative method that will allow us to measure concentrations and depth profiles of elements attached near the surfaces of reacted minerals. The technique is non-destructive, and in certain circumstances can be applied in situ. Because such measurements are taken in the vicinity of the critical angle for total external reflection, they have been referred to in the literature as TXRF (Total external reflection X-Ray Fluorescence) or TPIXE (Total external reflection Proton Induced X-ray Emission; Van Kan, 1996), respectively. Note also that this type of chemical information can be conveniently combined with X-ray surface reflectivity measurements to provide a full chemical and structural analysis of the near surface region. Reflectivity, diffuse scatter, and TXRF measurements were completed at the Daresbury Laboratory Synchrotron Source, Beam Line 2.3. TPIXE and RBS spectra were collected at the Oxford SPM facility. Experiments to be described will include the adsorption of trace quantities of Rb onto a spodumene (010) single crystal substrate. Concentration of Rb was measured to be 0.237 µ g/cm2 (approximately 1 atom of Rb for each 6 square angstroms of surface area) with a lower limit of detection of 0.017 µ g/cm2. Power law descriptions of the evolution of the surface topography during reaction will also be presented based on the X-ray scattering results.
Van Kan J, Ph.D. Dissertation, Vrije Univ. te Amsterdam, (1996).
Zircon is a mineral known to be highly resistant to weathering. It may also survive to several geological cycles including metamorphism and anatexy. However, in some environments (hydrothermal fluids, tropical soils), zircon may loose its resistance and mass balance studies have questioned the immobility of zirconium during weathering. In addition, loss of crystal structure due to the accumulation of radiation induced defects, i.e. metamictization, strongly modifies the bulk thermodynamic properties of zircon.
Because surface properties are known to control the silicate weathering rates and may be different from bulk properties, a detailed study of the surface of zircon has been undertaken inorder to better understand the resistance of zircon to dissolution. Surface morphologies of zircons originating from various environments have been imaged using Scanning Electron Microscopy and Tapping Mode Atomic Force Microscopy. First results show contrasted surface morphologies depending on the crystallographical nature of the observed face and crystal growth conditions. Concerning the well crystallized zircon, these contrasts can be interpreted using the periodic bond chain theory which suggests that the surface structure is probably close to the bulk structure. In the case of variably metamict zircon, the metamictization degree can be determined using Raman spectroscopy and is shown to partly control the surface morphology of zircon in tropical soils.
In geologic systems, pressure solution is considered a fundamental mechanism for rock deformation, which contributes to porosity reduction. Diffusion through thin-film aqueous fluid is considered to play a significant role in transporting material on fracture surfaces and is postulated to occur between rock grains. The structure and energetics of water at the mineral interface is important for understanding microscopic processes in thin-film fluids. Recent AFM (Atomic Force Microscopy) observations (Stipp, 1994) of freshly cleaved calcite surfaces in air, revealing the atomistic structure of the surface, indicate that water present as humidity reacts with the surface to satisfy the "dangling" bonds of calcite. The formation of hillocks and holes on terraces and cleavage steps are observed with heights or depths varying between 3-30 Angstroms, on the time scale of hours. These processes are interpreted as the result of dissolution and reprecipitation within thin-film water formed on the surface.
We have studied microscopic processes occurring on and between calcite surfaces in the presence of aqueous fluid using molecular simulations with empirical force-fields. Using molecular mechanics and annealing dynamics, we have investigated the energetics and morphology of adsorbed water layers on the calcite {104} natural cleavage plane with and without steps. The adsorbed water structure from simulation is consistent with experimental observations. We observe that the adsorbed water-layer stabilizes the stepped and non-stepped surface calcite atoms. We have employed molecular dynamics to study the effect of temperature on the dynamic water structure and mobility around the steps.
Stipp SLS, Eggleston CM & Nielsen, BS, Geochemica et Cosmochimica Acta, 58, 3023-33, (1994).
Modeling the mobilization of major elements and their transport in soils requires an understanding of the dissolution/weathering kinetics of clay minerals. The dissolution of silicates under acidic conditions is ultimately related to the occurrence of protonated metal species at the surface. However, the distribution of surface species is not homogeneous over the entire mineral surface, but depends on the local structural and chemical environment of individual sites on different crystallographic surfaces. Deriving reaction rates and determining the dominating mechanisms requires the characterization of the reactive surface area. Although the specific surface area can be determined by BET techniques, the reactive surface area is not known in most cases.
We have used a new sample preparation technique which enabled us to image the dissolution of hectorite particles in-situ at molecular scales using Atomic Force Microscopy (AFM). The hectorite particles were attached to a mica substrate via a polyethyleneimine based polymer. Hectorite is a trioctahedral smectite containing Mg and Li as the octahedral cations, and having the highest edge surface/basal surface ratio of all natural clay minerals. Particle dimensions of individual platelet-like particles (height: ~ 2 nm) representing one structural hectorite layer, and their evolution within a dissolution experiment were quantified using image analysis software.
It has been suggested that for 2:1 phyllosilicates, the edge surfaces are much more reactive in dissolution than the basal surfaces. In the case of hectorite, dissolution takes place at edge surfaces exclusively. The basal surface is completely unreactive within the investigated time scale of several hours. Based on the measured morphological changes due to dissolution, a reactive surface area normalized rate was derived. Under pH2 conditions at room temperature the dissolution rate normalized to edge surface area is 4.3 x 10-9 mole hectorite / m2 sec. This is equivalent to 1.9 x 10-10 mole hectorite / m2 sec normalized to the total surface area, which agrees well with macroscopically determined dissolution rates for many clay minerals. We also found that the reactive surface area / total surface area is not a constant value but increases during a dissolution experiment.
Our experiments provide a way of performing dissolution experiments where measured variations in the reaction rates can be attributed exclusively to variations in surface chemistry and any variations due to differences in surface area can be excluded.
Several silicate minerals show a pH dependence with a rate minimum at a certain pH. Quartz shows a rate minimum at the pH corresponding to the point of zero net proton charge (pznpc), while aluminosilicates have their minimum at a pH between the pznpc of Al2O3 and SiO2 (Walther, 1996). The pH dependence can be of three different origins; the effect of charge on a specific site, the entropy effect from charge on the surrounding water structure, or the effect, considered in the present study, of charge on the surface layer stability.
The surface layer stability was modeled using a polymerization reaction between silica chains of different length and charge at the terminating silanol groups,
pH<pznpc: +H2O-Si(OH)2-O-[Si(OH)2-O]x-Si(OH)3 + Si(OH)3-[Si(OH)2-O]x-O-Si(OH)2-H2O+,
pH>pznpc: -O-Si(OH)2-O-[Si(OH)2-O]x-Si(OH)3 + Si(OH)3-[Si(OH)2-O]x-O-Si(OH)2-O-.
As pH deviates from the pznpc the polymerization reaction is prevented by the acid-acid or base-base repulsion induced from similarly charged terminating silanol groups, while at pH close to pznpc the polymerization is instead favored by the acid-base attraction induced from neutral or opposite charged silanol groups. A stepwise increase of the silica chain length was used to model the decrease in surface charge density as pH gets closer to pznpc. The model simulations were performed using HF level calculations and the standard basis set 3-21 G, however larger basis sets and inclusion of dynamical correlation through density functional theory (DFT) using the B3LYP functional were tested but had no effect on the qualitative trends.
The use of a polymerisation reaction was a successful way of illustrating the acid-base properties created at the silica surface due to the charging of surface hydroxyl groups. The amphoteric behavior of silanol groups induced acid-acid or base-base repulsion between two chains of similar charge and attraction between neutral chains in agreement with acid-base interactions. The inductive electronic effects of substituents are known to decrease with distance, thus the minimum dissolution rate is expected at the point at which the surface charge density is at a minimum or at the point where opposite charging at the surface is large, i.e. at the pznpc. Indeed, the parabolic pH dependence as noticed during quartz dissolution was reconstructed from the reaction energy for an increase in silica chain lengths as a simulation of decrased charge density. The discussions of acid-base properties have implications also for the pH dependence of aluminosilicate dissolution. Since the pznpc of Al2O3 is at pH 8.4 and for SiO2 at 2-3 (Walther, 1996) there will be attractive forces in between these pH values from the positively charged acidic aluminum sites to the negatively charged basic silica sites, which explains the observed pH dependence of aluminosilicate dissolution as that for albite (Wollast and Chou, 1986).
Walther JV, Am. J. Science, 296, 693-728, (1996).
Wollast R & Chou L, Trans. XII Cong. Intl. Soc. Soil. Sci, (1986).
The surface chemistry of olivine (Fo92) in aqueous solutions was investigated as a function of pH, pCO2 and ionic strength using potentiometric titrations, electrokinetic measurements (streaming potential and electrophoresis techniques), and X-ray Photoelectron Spectroscopy (XPS). At pH < 9, a Mg-depleted surface layer (< 8 nm in thickness) develops due to an exchange reaction between a magnesium atom and 2 protons as confirmed by the preferential release of Mg over Si during the initial stage of olivine dissolution. At pH > 9, the surface exhibits a stoichiometric Mg/Si ratio or a slight enrichment in Mg. Electrokinetic measurements yield an isoelectric point or pHi.e.p. of 4.5 which is consistent with the dominance of silica sites at olivine surface for pH < 9. In contrast, surface titrations result in a point of zero charge or pHp.z.c. of 10 and the development of a high positive surface charge (up to 10-4 mol/m2) in acid solutions. This may be explained by penetration of H+ into the first 3-5 molecular layers of the solid as confirmed by XPS data that show a distinct increase in the half-peak-height width of the O1speak.
The dissolution rates of Fo92 were measured at 25°C in a mixed-flow reactor as a function of pH (3 to 12), (sum) CO2 (0 to 0.05 M) and aqueous silica concentration (0 to 0.001 M). In CO2-free solutions, the rates decrease with pH at 3<pH<8 with a slope close to 0.5. At pH>=8 the rates decrease with a smaller slope of about 0.2 and become pH independent for pH>=10. Olivine dissolution rates are inversely proportional to aqueous carbonate and silica activities at pCO32- <4 and pH>8.5, respectively. These results allow to elucidate the mechanisms of forsterite dissolution. In acid to weakly alkaline solutions, the dissolution is controlled by the decomposition of a silica-rich, Mg-free, protonated surface complex formed by a fast exchange of 2 H+ for one Mg2+, followed by the rate limiting sorption of 0.5 H+ on each exchanged site. At pH > 9, it is the hydration of surface Mg sites with the formation of >MgOH2+ species which controls dissolution. Preferential Si release, observed at the initial dissolution stage in alkaline solutions, results in the formation of a brucite-like layer on the surface. The breaking of Mg-O-Mg bonds is thus the critical step for forsterite dissolution at these conditions. The results of this study demonstrate that olivine dissolution is controlled by two parallel reactions occuring on silica-rich and hydrated Mg sites. Based on the results of this study, the influence of dissolved atmospheric CO2 on dissolution rate cannot be neglected when modeling the weathering of mafic minerals by alkaline surficial waters.
Despite their abundance in natural and man-made environments, little is known of the structure and structural development of amorphous sulfides. These materials form the link between dissolved species and crystalline solids and are, therefore, an important stage in the global cycling of metals. The structural changes in amorphous compounds can be (uniquely) followed by investigating the local environment of both the metal and sulfur using XAS. The development of X-ray absorption spectroscopy (XAS) has the unique ability to provide element specific structural information on non-crystalline solids. By freezing (77K) sulfide precipitates formed in (oxygenated) aqueous solutions at various times (and after by ageing at various temperatures), the structural evolution an be monitored.
In one study, the evolution of CuS precipitates was examined and a clearly defined structural change was recorded involving a metastable primitive structure that ages to a structure with the characteristics of amorphous covellite. EXAFS analysis of the primitive structure demonstrates the presence of disulfide (S2-groups) and a Cu-S interaction of 2.8 Å, the latter is not found in covellite. Copper K-edge XANES indicates the dominance of 3- over 4-co-ordinate copper in the primitive phase, while Cu L3-edge spectra reveals only Cu(I) to be present in all precipitates formed. On ageing, the primitive structure transforms to one with the characteristics of covellite and this transformation involves the reordering of the S2- and Cu3S-CuS3-layers. Development of the primitive phase from either a wurtzite-like structure or planar Cu3S-CuS3-layers is possible, with the structural evolution driven by the antipathy of Cu(II) for tetrahedral co-ordination and anomalous electron densities in the metastable structures.
In the study of the formation of HgS, a transition in the first few seconds was recorded. After 2 seconds a cinnabar like linear structure forms with the Hg bonded to 2 S atoms at 2.37 Å, while after 10 minutes the Hg is surrounded by 4 S atoms at 2.53 Å similar to metacinnabar; in many natural surface anoxic environments cubic HgS has been recognised as the stable phase. Future experiments will utilise the energy dispersive EXAFS which can produce structural information on the scale of 100 ms (and good data in >1 sec.); at this time scale the intermediate stages involved will be revealed.
Using the constant addition technique the mechanism of calcite crystal growth in seawater has been evaluated by the results of a set of rate experiments in which the influence of the solution composition and specific seawater components were measured at the temperature of 298K. Assuming the overall reaction rate model, the partial reaction order with respect to the carbonate ion concentration is equal to 3 in artificial seawater solutions and remains constant in absence of the specific inorganic constituent such as magnesium, sulphates, phosphates and related seawater complex ions as well dissolved organic matter. Their presence however decreases the apparent kinetic constant of more then one order of magnitude. This suggest that the mechanism of calcite crystal growth from seawater does not change significantly as a results of adsorption and incorporation of seawater inhibitors as magnesium, sulphates, phosphates and dissolved organic matter. Atomic Force Microscopy observations of calcite crystal growth through fluid cell indicates that speed and length of steps are critically controlled by the supersaturation state of the solution that energetically controls the growth morphology for a given activity ratio of the calcium and carbonate ions. Inorganic and organic calcite inhibitors reducing the density of the actives sites increase the width of the terraces and influence length and directions of the steps. The results of this two-scale investigations allows us to define a calcite crystal growth model from seawater where the surface roughness coefficient is a function of the ratio between the initial and the non-reacted surface area. Growth on the perfect cleavage rhombohedron face produces an initial increase of the surface roughness that later decreases to reach a constant value. When the spiral mechanism is dominant the roughness asymptotically reaches the maximum value of 1.5. The contribution of the specific seawater inhibitors can be estimated associating a geometric factor compatible with the maximum and minimum value of the roughness coefficient.
The process of Cd2+ uptake during carbonate crystallisation is of considerable interest in geochemistry and environmental sciences. The determination of the Cd/Ca ratios within sedimentary carbonates has been used to estimate ancient climatic and biological conditions of marine environments (Marchitto et al., 1998) and sorption and desorption of Cd2+ by calcite crystals is a phenomenon directly related to contamination problems of groundwater (Nriagu, 1981). However, the mechanisms of Cd2+ incorporation on crystallising surfaces are controversial and numerous models invoking adsorption, formation of intermediate phases and subsequent recrystallisation, etc. have been recently reported (Davis et al., 1987; Stipp et al., 1992). Some models propose the direct precipitation from solution of the (Cd,Ca)CO3 solid solution on calcite as a possible way of Cd2+ incorporation from aqueous solution (Chiarello, 1994) .
Here we present in situ AFM observations of the crystallisation of (Cd,Ca)CO3 solid solution on calcite cleavage surfaces, using aqueous solutions with concentrations ranging from [Ca2+] = 0 µ mol/l ; [CO32-] = 5.10-11 µ mol/l; [Cd2+]= 10-6 µ mol/l to [Ca2+]= 125 µ mol/l; [CO32-] = 125 µ mol/l and [Cd2+] = 10-4 µ mol/l. Under such experimental conditions the rates of advancement of both monomolecular steps (~3Å high) and two dimensional nucleation rates have been measured.
Crystallisation of (Cd,Ca)CO3 on calcite surfaces from different supersaturated Cd2+ - Ca2+ - CO32- aqueous solutions is interpreted in terms of the relationship between supersaturation and the operating growth mechanism (spiral growth, two dimensional nucleation or continuous growth). For the case of solid solutions it is necessary to consider that both the degree of supersaturation (ß) and the transitional supersaturation values (ß* and ß**) between growth mechanisms are not single parameters, but are functions of the solid composition (Prieto et al., 1993). As a result, the intersection of the calculated supersaturation function for a given aqueous solution (ß(Cd,Ca)CO3), and the change of transitional supersaturation values for different solid solution compositions defines regions of different growth mechanism in supersaturation-solid composition diagrams. AFM observations and kinetic data obtained (step velocities and nucleation rates) are consistent with the existence of such regions, also revealing a clear influence of the calcite surface structure on the crystallisation process. Our observations suggest that even at very low supersaturations of the aqueous solution relative to the solid solution, Cd2+ is incorporated into the carbonate by direct growth of a (Cd,Ca)CO3 solid solution, either by step advancement or two dimensional nucleation.
Chiarello RP & Sturchio NC, Geochim. et Cosmochim Acta, 58, 1467-1474, (1994).
Davis JA, Fuller CC & Cook AD, Geochim. et Cosmochim. Acta, 51, 1477-1490, (1987).
Marchitto JrTM, Curry WB & Oppo DW, Nature, 393, 557-560, (1998).
Nriagu JO, Cadmium in the Environment, Wiley-Interscience, 1981
Prieto M, Putnis A & Fernández-Díaz L, Geol. Mag, 130, 289-299, (1993).
Stipp SL, Hochella MF Jr, Parks GA & Leckie JO, Geochim. et Cosmochim Acta, 56, 1941-1954, (1992).
The basic principle of an Atomic Force Microscope (AFM) is that a sharp probe mounted on a flexible cantilever is in (intermittent) contact with the sample surface. A laser beam focussed on the cantilever and reflected onto a photodiode detector monitors the deflection due to the variations in surface topology at (sub)nanometric scale.
The existence of two different {110} surfaces on natural melanite from Rudny (Kazakhstan) could be visualized for the first time. The Hartman-Perdok Theory predicts the possibility that two energetically different interfaces are possible, but they were never observed. The absolute height difference between these two surfaces is about 0.4 nm, i.e., half the theoretical thickness of an elementary growth layer. Moreover, the relation between the growth spirals that are present together with dissolution etch pits, will be discussed.
The F character of {110} on experimentally grown spessartine (Mn3Al2(SiO4)3) could be proven by the presence of growth hillocks. The typical {100} surface of hydrothermally grown spessartine is definitely due to corrosion since dissolution features and etch pits are observed.
The interaction with the sharp probe can reconstruct the surface of hydrothermally grown crystals completely. Since this occurs only on hydrothermally grown spessartine, the presence of an amorphous layer together with the capillary bonded water on the surface could be responsible for this effect. Reconstruction can be avoided by using the intermittent contact mode.
Growth spirals on prism faces of quartz crystals which are grown on spessartine, establish the F character of these faces.
Smectites are highly reactive minerals of superficial environments. This surface reactivity arises from their layered and defective structure. Both cation exchange on basal plane sites, and pH-dependent sorption, have been reported to occur on smectites, but the molecular mechanism of this second sorption process, and its possible interference with the first one have not been investigated yet. These issues have been addressed by combining spectroscopic and chemical kinetic approaches to cobalt sorption on hectorite, a magnesian smectite.
Contrasted kinetics of cobalt adsorption on hectorite were observed at high (I=0.3 M) and low (I=0.01 M) ionic strengths. At I=0.01 M, a high amount of Co sorbed within the first 5 minutes of reaction (30.9 µ mol/g hectorite). This rapid uptake is indicative of Co sorption on cation exchange sites. For I=0.3 M, this initial adsorption was much less important (3.6 µ mol/g hectorite for 5 mn reaction), but sorption became significant after 5 days of reaction (37.5 µ mol/g hectorite), and coincided with proton release in the supernatant. As cation exchange is hindered at I=0.3 M, Co presumably adsorbed on pH-dependent sites at high ionic strength.
P-EXAFS experiments were performed on a self-supported film of Co-sorbed hectorite prepared at I=0.3 M to determine the mechanism of pH-dependent sorption. The shape of P-EXAFS spectra strongly varied with the experimental angle between the clay film and the electric field vector e. This angular dependence indicates that the environment of Co is anisotropic, and suggests a direct bonding of the sorbate on the clay sorbent. Combination of geometrical considerations and spectral simulations substantiate the conclusion that Co was surrounded by 1.6 ± 0.4 Mg at 3.03 Å and 2.2 ± 0.5 Si at 3.27 Å in the parallel and perpendicular orientations, respectively. These distances are similar to Mg-Mg and Mg-Si distances in hectorite. The orientation of the two cationic shells with respect to the film plane further confirms that Co sorbed at the edges of hectorite platelets, in the continuity of the magnesian octahedral sheet.
EXAFS spectra of similar shapes were obtained at I=0.01 M or 0.3 M, and for reacting times higher than 6 h, which indicates that a significant amount of Co also sorbed on layer edge sites at low ionic strength. This result is unexpected because solution chemistry suggests adsorption of Co on basal plane exchange sites at I=0.01 M. To account for this duality, chemical data were fitted by a kinetics model, which considers the occurrence of both cation exchange and edge adsorption. This model shows that even at I=0.01 M, and after a short reacting time, a significant amount of Co is transferred from exchange sites to edge sites. This sorption mechanism reconciles EXAFS and chemical results.
The binding energy (BE) of Cr(III) sorbed to micas splits into a high and low BE field, ~0.3 eV, after reaction in NaCl and KCl aqueous solutions, respectively. At the experimental conditions, Na+ and K+ differentially block permanent charge sites but not variably charged edge sites, and the question is whether or not this small BE difference reflects a difference in the dominant sorption site for Cr. Although XPS is a useful tool for determining the chemical state and bonding environment of species at the near-surface of minerals, interpreting small shifts in BEs such as this is a complex issue owing to potential variations in (I) dominant sorption sites, (II) sorption density, (III) speciation of the sorbed cation, and (IV) surface composition. Silicates are electrical insulators, which precludes an absolute BE scale. Differential dissolution of a silicate substrate could shift the BE of elements used as BE references, and surface charging can degrade analytical precision. Alternatively, surface composition could affect the BE of the sorbate. In order to rule out these complicating factors, we (1) used solutions with similar ionic strength, pH, aCl-, and aCr which limited differences in sorbed Cr speciation; (2) used both Si and C as BE references-Si and C have complementary properties and, for example, their combined use can help determine if BE shifts are due to differences in surface composition; (3) analysed both Cr2 p and Cr3 p spectra to gather information on Cr depth distribution and reduce the risk of artifacts due to charge shifting; and (4) compared BEs over a range of sorption density. Given these procedures, we interpret the difference in Cr(III) BE between the NaCl and KCl experiments to reflect a shift in the dominant sorption site and conclude that small BE shifts for cations sorbed to silicates can be interpreted with confidence if strict protocols are employed.
We are using ab initio electronic structure calculations to study the binding of organic molecules on silicate surfaces, focusing particularly on pollutant molecules such as PCBs and dioxins. This is a challenging problem, and we will report preliminary calculations using chlorinated aromatic molecules on a hydrated silica surface. We will report on the technical strategies, such as the relative merits of plane wave basis sets vs localised atomic orbitals, as well as the results of the preliminary calculations.
The hydroxy (oxide) and sulfide minerals of iron occur as very fine particle materials in sedimentary systems, spanning an important redox boundary between oxidised surface sediments and the reduced environments dominated by bacterial sulfate reduction. Because of their relative abundance, combined with highly reactive and large area surfaces, these minerals play a key role in the cycling of elements at the surface of the earth.
Our recent studies have involved measuring the uptake from aqueous solution of a range of elements representative of major geochemical and environmental groups (Cu, Cd, As, La, H, Tc) through interactions with the surfaces of major oxidic and sulfidic iron minerals. In addition to bulk chemical analysis to determine uptake behaviour as a function of key variables, X-ray absorption spectroscopies (EXAFS and XANES) have been used to probe the mechanisms of interaction between the fine particle mineral surface and the species removed from solution. These studies show a rich variety of behaviour at the molecular level, including outer sphere and inner sphere surface complexation ("physisorption" and "chemisorption") and precipitation or replacement reactions that may include redox processes.
Such fundamental knowledge of the mechanisms and rates of these interactions is required to inform our understanding of the controls on element cycling at scales ranging from molecular to global, and to undertake predictive modelling of geochemical cycles.
The B horizon of a podzolic soil profile is a heterogeneous mixture of primary minerals, silicate clays, precipitated iron and aluminium oxides and humic materials. There are three major sources of surface charge in these horizons: (1) the permanent negative charge of silicate clays created by the isomorphic substitution of Si and Al by ions of lower valence in the crystal lattice, (2) the pH dependent positive/negative surface charge on secondary oxides due to protonation/de-protonation of surface hydroxyls and (3) the pH dependent negative charge of humic material formed by the dissociation of (mainly) carboxylic groups. In addition to the heterogeneity within the soils there is also a considerable variation between soils with varying organic matter, clay and oxide contents. In order to understand exchange processes at the solid/water interface in these soils under changing conditions it is necessary to quantify the origin of the surface charge. We have estimated the origin of surface charge for different podzol B horizons by combing information from determinations of specific surface area (BET), Cs adsorption, ion exclusion and oxalate soluble Fe, Al and Si. The results show that the pH dependent surface charge largely determines the charge characteristics of the B horizon. However, the variation in the magnitude of surface charge originating from either oxides or humic material is considerable between soils. The surface charge of these materials were also strongly influenced by the presence of specifically bound ions like sulphate (oxides) and aluminium (humic matter). An attempt was made to model the development of surface charge over a pH range by a mechanistic model taking into account the protonation of surface hydroxyls, dissociation of carboxylic groups and quantitatively important complexation processes. The results shows that the effect of a change in pH on the surface charge is considerable and we therefore conclude that the assumed constant permanent negative charge of the B horizon, as used in several soil chemical models, has to be reconsidered.
Ferrihydrites (FeOOH, or "HFO") are poorly ordered iron hydroxides that have been proposed to play an important role in the geochemistry of trace elements because of their reactivity and large specific surface area (Manceau et al, 1997). In tropicals countries, soils are very often constituted of iron oxi-hydroxides which can fix weighted elements like Au (Greffié et al., 1996). X-ray Absorption Fine Structure (XAFS) spectroscopy was used to determine the speciation of gold at the iron hydroxide-water interface as a function of pH, and chloride concentration of the initial Au-solution. Samples of Au(III) adsorbed onto ferrihydrites have been prepared from pure ferrihydrites and various gold(III) chloride solutions (0.0028 M Au; 1.2 and 0.0056 M Cl and pH=4-10). The gold concentration at the beginning of the experiment and after equilibrium has been mesured by ICPAES in order to calculate the gold amount sorbed (0.7 g for 100 g HFO). Environmental Scanning Electron Microscopy (gun voltage=30 kV, secondary ion detector, chamber pressure=1/100 of the atmospheric pressure) analyses show that Au is homogeneously distributed. No evidence for metallic precipitates was found at the scale of observation (50 to 0.5 µ m ). XAFS spectra have been collected at the SSRL (Stanford, USA) facility at 293 K and at the LURE facility (Orsay, France) at 10 K, using the fluorescence mode (solid state Ge-detector) and Si(220) double-crystal monochromator. The Au-LIII edge XAFS spectra for Au adsorbed onto HFO's are close to these measured for Au(III)-bearing model compounds and aqueous solutions. These XANES spectra show the presence of multiple-scattering features characteristic of Au(III) located in a square planar geometry. The analysis of the XAFS oscillations suggests the presence of AuO4, AuCl3O or AuCl2,O2, environments (depending on synthesis conditions) with Au-O and Au-Cl distances of =2.00 and 2.28 Å. At low pH and high Cl-concentration, AuCl3O or AuCl2,O2 moieties are observed while, at low pH or low Cl concentration, AuO4 appear dominant. No clear evidence for Au nor Fe second neighbors were observed (as well as multiple-scattering features that are much weaker when O replaces Cl according to FEFF ab-initio calculations). Collecting XAFS data at low temperatures do not resulted in drastically different XAFS spectra. These results show that the redox state of Au remains unchanged during adsorption in contrast to the short range environment of Au which is strongly sensitive on the pH and the Cl molality of the solution during adsorption.
Greffié C, Benedetti M, Parron A & Amouric M, Geochimica Cosmochimica Acta, 60, 1531-1542, (1996).
Manceau A, & Gates WP, Clays and Clay Minerals, 45, 448-460, (1997).
Understanding the inhibition of crystal growth by organic additives has wide application both in natural mineralisation in sedimentary environments and in industrial processes. Phosphonate inhibitors have been used for some years to prevent the build-up of barite scale in off-shore oil-wells. However, the mechanism of inhibition is largely unknown. In this paper we demonstrate that it is possible to study the growth of barite and celestite at molecular resolution in situ in a fluid cell of an atomic force microscope (AFM) and hence to make inferences about the attachment of inhibitor molecules on specific surface sites.
In the absence of inhibitor, the advancement of molecular steps as well as two-dimensional nucleation can be observed (Pina et al., 1998). Specific features of the process, such as the shape of the nuclei and the growth rate in different crystallographic directions can be measured by direct observation. This provides the basis for understanding modifications to the growth process introduced by the presence of inhibitor molecules in the solution.
The phosphonic acids described here include small molecules such as HEDP (Hydroxyethane diphosphonic acid) which have been used as inhibitors for some years, as well as new 'molecular-engineered' molecules with a larger number of binding motifs, and hence presumably a more efficient inhibition under higher supersaturation conditions. One such example is a macrocyclic aminomethylphosphonate, (hexaza-18-crown-6). All phosphonic acids, when added to the supersaturated sulphate solution in the fluid cell, first attach to the molecular-height cleavage steps on the surface. As the concentration of the inhibitor increases, the organic molecules also attach to the surface and build a complete monomolecular coating. With small phosphonate molecules such as HEDP it is not possible to observe this directly, but the attachment of molecules on kink sites can be inferred from the development of serrated step morphologies. Larger phosphonate molecules can be imaged directly by changes in the height of steps.
HEDP inhibits growth by attaching to the growth steps. Further growth is inhibited unless the supersaturation of the sulphate solution is increased, or the inhibitor is removed. The effectiveness of the inhibitor depends on its concentration, and the supersaturation of the fluid. Passing an inhibitor-free solution over the surface can remove HEDP, indicating a rather weak attachment to the surface. On the other hand the larger macrocyclic molecules with up to 6 binding phosphonate motifs attach very strongly to step edges and surfaces and completely inhibit growth, even at low inhibitor concentrations where HEDP is less effective. This inhibitor cannot be removed by passing inhibitor-free fluid over the crystal surface, and continues to be effective at high supersaturations.
The effectiveness of the inhibitors and the mechanism of inhibition will be defined as a function of both inhibitor concentration and fluid supersaturation.
Pina CM, Becker U, Risthaus P, Bosbach D & Putnis A, Nature, 395, 483-486, (1998).
Pressure solution creep has for a long time been considered an important ductile rock deformation mechanism in diagenesis of sedimentary rocks. There is however, still considerable debate about the dissolution and transport mechanisms on the microscopic grain boundary scale. The principal models are 1) The "thin film model" (Weyl 1959) where the "contact" surfaces are assumed smooth with a fluid film between the mineral grains supporting the stress and allowing diffusional matter transport. 2) The "dynamic island channel" model (Raj 1982) where there are islands of actual mineral contact that dissolve on the perimeter and matter is transported in the channels between the islands. When one island has dissolved new islands take over the stress giving a dynamically changing geometry of the interface. 3) The "microcrack model" (Gratz 1991) which assumes astatic island channel geometry caused by initial microcracking of the surfaces with a thin fluid film to transport matter through the mineral "contacts" to the channels. The different models lead to different qualitative grain boundary geometries, rate laws and order of magnitude differences in pressure solution creep rates. Carefully performed experiments should therefore enable us to discern under which circumstances a particular mechanism is the predominant.
We present the results of a number of surface deformation experiments on synthetic and natural halite close to ambient temperature and fluid pressure. We chose halite as our model system because the creep rates are very fast compared to other naturally ocurring minerals and the pressure solution is diffusion controlled thus facilitating comparison of the models. We have applied capacitative micrometry with Angstrom resolution during the experiments to measure the creep rate of single contacts under highly controlled conditions. The contact surface width was constant during each experiment and between experiments it was varied from 10 to 1000 micrometers. In order to avoid changes in the saturation that would otherwise result in gross flow of sodium chloride between the solution and the crystal we have used water volumes of 2 to 4 microliters around the contacts in a fluid sealed enclosure and controlled the temperature to ± 0.2 mK. The grain contact pressure has been varied from 0.2 to 20 MPa. The measured creep rates at varying pressures and contact widths are compared to the different models indicating possible ranges of validity of each. The contact surfaces are studied by optical and atomic force microscopy after the experiments and are found to be structured with typical island sizes of 1 to 10 micrometers. Differences between results on natural and laboratory grown halite is discussed with relation to the influence of pre-existing faults as opposed to microcracks and plastic deformation induced by the applied stresses.
Weyl PK, Journal of Geophysical Research, 64, 2001-2025, (1959).
Raj R, Journal of Geophysical Research, 87, 4731-4739, (1982).
Gratz AJ, Geology, 19, 901-904, (1991).
During lixiviation of the French inactive analog of high activity nuclear glass, complex alteration layers are described for long term experiments (several months). These layers are a mixture of crystallized and amorphous mineral compounds which are known to trap heavy metals (for example Zr and Fe). In order to understand the first steps of glass alteration, nuclear glass monoliths were altered during short leaching time (3 hours to 7 days). Alterations were conducted under static flow, at 90°C, with deionized water or silicon bearing solution (86 ppm of Si). Structural changes of the monolith surface were investigated at the Zr-L2,3 and Fe-L2,3 edges at LURE/SuperACO (Orsay, France) facility. In such beam lines, total yield detection allowed to probe few hundreds Å at the energies considered. XANES spectra of reference compounds with known Zr and Fe structural environments were also recorded. Surface of altered monoliths were observed by SEM before spectroscopic measurements. No precipitates were observed for leaching experiments in Si-bearing solution, and in deionized water for 3 and 8 hours.
Zr-L2,3 XANES spectra yield identical signatures between pristine glass and altered monoliths obtained with the silicon bearing solution whatever the alteration time was. Comparison with model compounds, indicate that Zr is 6-fold coordinated to oxygen. For alteration performed with deionized water, the local environment of Zr is modified till 3 h and strongly suggests an increase of Zr coordination number. No more change is observed for longer alteration times. At Fe-L2,3 edge, XANES spectra evidence that, whatever the silicon content and alteration time are, the local structure around Fe is modified during leaching. Fe is octahedrally coordinated to oxygens in altered glass monoliths, while it is 4-fold coordinated in pristine glass.
Zr- and Fe- L2,3 XANES using soft X-ray and total yield detection are sensitive probes of surface modifications during short term alteration of the nuclear glass. These changes in local structure are different for Fe and Zr. Iron is highly reactive while Zr is stabilized when Si-content in solution is higher.
Minerals of the phyllomanganate group, of which birnessite (Bi) is one of the predominant members, play a pivotal role in redox and adsorption processes which occur in soil, groundwater, oceanic and aquatic systems. Results obtained from complementary structural techniques on the sorption mechanism of transition metals (Me) on Bi will be presented and discussed in light of the recent structural birnessite model presented by Drits et al. (1997) and Silvester et al. (1997). One of the primary aims of this study is to obtain structural models for Me-sorbed phyllomanganate, which can be used to assist in the determination of the speciation of trace elements in natural systems.
Me-sorbed birnessites (MeBi) were prepared at different surface coverages by equilibrating a Na-exchanged buserite suspension in the presence of the aqueous metal at pH 4. Me/Mn atomic ratio were varied from 0.5% to 13.4% for Zn, from 0.1 to 6.1% for Pb, and was equal to 16.3% for Cu. Selected area electron diffraction (SAED) coupled with energy dispersive analysis reveal that Me Bi are chemically and structurally heterogeneous. Some particules display super-reflection networks, as well as commensurable and incommensurable satellite reflections resulting from the long range ordered, or semi-ordered, distribution of interlayer Me. X-ray diffraction (XRD) show that ZnBi and PbBi particules have a one-layer hexagonal unit cell whereras CuBi has a hexagonal layer symmetry and a monoclinic layer stack. All MeBi contain variable amounts of stacking faults.
In all MeBi a well defined 2nd shell of Mn atoms was identified by EXAFS at 3.4 - 3.5 Å; for Zn and Cu, and at 3.7 - 3.8 Å; for Pb, which indicates that Me sorb above and/or below layer Mn vacancies. Metal surface complexes are bonded to three oxygens of the Bi anionic framework and have 6 nearest neighbor Mn atoms as in chalcophanite (Zn2+Mn4+3O7.9H2O), where Zn form triple-corner (TC) surface complexes. MeBi however differ from chalcophanite by four principal short range order features. First, they have a weaker EXAFS signal as a consequence of the higher structural disorder as indicated by SAED and XRD. Second, Me-Me TC pairs across vacant layer sites were identified at all surface coverages in PbBi, but not in ZnBi nor in CuBi. This stands in contrast with chalcophanite where interlayer Zn systematically form Zn-Zn TC pairs above and below layer vacant sites. Third, Zn-Olayer and Zn-Mn distances are shorter than in chalcophanite indicating that interlayer Zn atoms are closer to the phyllomanganate layer. The displacement of Zn in direction of the MnO6 layer can be accounted for by the absence, or scarcity, of Zn-Zn TC pairs. Fourth, short Zn-O (1.97 - 1.99 Å) and Zn-Mn (3.35 Å) distances were detected at low Zn concentration (Zn/Mn = 0.5% and 0.8%). These distances are compatible with the formation of IVZnO4 TC-sharing surface complexes at layer vacancy sites. These complexes supposedly form on most undersaturated surface sites where Mn3+ layer prevail. It will be shown in conclusion that several of the surface complexes described on synthetic birnessite have been identified very recently in soils.
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications