Surface Concentration Of Sites Research Articles

The radicals generation in the methanol oxidation on a Pt-containing catalysts. Influence of support acidic properties

Gas phase radicals, formed during methanol oxidation, have been investigated for series of Pt-containing catalysts with different support acidic properties, using matrix isolation method. Generation rates and activation energies were determined. It was estimated that strength and concentration of active surface sites play an important role in radicals generation.

Reactive uptake of ozone by proxies for organic aerosols: Surface versus bulk processes

Uptake measurements of ozone were conducted with two types of proxies for atmospheric organic aerosols: organic liquids and self‐assembled organic monolayers. Alkanes and terminal alkenes were used. The monolayer surface was characterized, prior to and after reaction, using IR spectroscopy. Uptake experiments were conducted using a flow tube reactor coupled to a chemical ionization mass spectrometer. The reactive uptake coefficient, γ, is shown to be due to reaction with the double bond. For the monolayers, γ is composed solely of a surface reactive component and is smaller by at least an order of magnitude than values obtained for a liquid of the same chain length. Uptake by the liquids is higher due to solubility and reaction in the bulk. The phase of the atmospheric organic aerosol will determine the appropriate use of a bulk or surface uptake probability in atmospheric models. Since the aerosol surface is processed and sites are consumed, γ is time variant. We define a parameter γ as the surface uptake probability per reactive site and determine its value as 9×10−19 cm2 molecule−1. This enables the modeling of surface reactions as surface site concentrations diminish following interaction with the gaseous species.

Partitioning of base cations and sulphate between solid and dissolved phases in three podzolised forest soils

The cation and SO 4 adsorption properties of O and B horizons of three podzolised soils in Sweden and Finland were studied through analysing the soil solution at six different sampling occasions and through a set of batch experiments. High concentrations of dissolved organic carbon and cations were found in centrifuged soil solutions from the O horizon, especially during autumn. An analysis using the WHAM-S model suggested that most of the dissolved Ca, Mg, K and Mn were counter-ions, residing in the diffuse layers of dissolved fulvic acids. Hence, the solubility of these cations depended on the solubility of organic matter. Model exercises suggested that the solubility of humics in turn was influenced by seasonal differences in hydrophobicity and by the water content. Furthermore, the model results showed that only a low proportion of the B horizon organic matter was involved in calcium binding. This is probably due to the interaction between organic matter and oxide surfaces in this horizon. In the case of sulphate adsorption in the B horizons, a surface complexation model was tested for its ability to describe batch experiment data and temporal differences in soil water chemistry. This model was based on the diffuse-layer model and it was optimised by adjusting the surface site concentration and the point of zero charge. However, the interaction between Ca 2+ and SO 4 2− could not be modelled. Comparisons between the model and the field observations proved to be difficult due to a considerable soil heterogeneity.

Chromium Leaching from a Silicone Foam-Encapsulated Mixed Waste Surrogate

This study assessed chromium leaching from silicone foam-encapsulated salt waste, using a surrogate formulated after Department of Energy complex mixed waste. Two commercial formulations of silicone foam (Wacker ELEKTROGUARD 2100 and General Electric RTV-664) were evaluated as a function of waste load (28−48 wt %). Chromium leaching was formulation specific and increased with increasing waste load as measured by the Toxicity Characteristic Leaching Procedure (TCLP). Chromium release followed transport controlled dissolution at all waste loads under TCLP (cut samples) and Accelerated Leach Test (ALT) (molded samples) conditions. Aqueous and surface complexation modeling was also used to describe reduced chromium effective diffusivity due to iron oxide addition. Comparison of modeling and measured diffusivities as a function of waste load demonstrated that the total available iron surface site concentration increased with increasing waste load, consistent with pore differences measured by image analysis. These results provide a basis for further work on modeling and engineering waste encapsulation using silicone foam.

A Comparative Study of Surface Acid–Base Characteristics of Natural Illites from Different Origins

The acid–base characteristics of naturally occurring illites, collected from different locations, were investigated by potentiometric titrations. The experimental data were interpreted using the constant capacitance surface complexation model. Considerable release of Al and Si from illite samples and subsequent complexation or precipitation of hydroxyl aluminosilicates generated during the acidimetric forward titration and the alkalimetric back titration, respectively, were observed. Therefore, the acidimetric supernatant, rather than the neutral one, was regarded as the system blank for each illite suspension to yield the surface site concentrations. In order to describe the acid–base chemistry of aqueous illite surfaces, two surface proton-reaction models, introducing the corresponding reactions between the dissolved aluminum species and silicic acid, as well as a surface Al–Si complex on homogeneous illite surface sites, were proposed as follows:The Kf2 constant in Model II was obtained by simulating the complex formation between the dissolved aluminum species and silicic acid that occurred in acidimetric supernatant when the hydroxide was added. Additionally, the following cation exchange reaction was also considered for a special case, where a large amount of K+ is released during the corresponding acidimetric titration, in which a high concentration of protons are consumed.Optimization results indicated that both models could provide a good description of the titration behavior for all aqueous illite systems in this study. The intrinsic acidity constants for the different illites were similar in Model I, showing some generalities in their acid–base properties. Model I may be considered as a simplification of Model II, evident in the similarities between the corresponding constants. In addition, the formation constant for surface Al–Si species (complexes or precipitates) is relatively stable in this study.

Langmuir‐hinshelwood model of soil phosphorus kinetics1

A mathematical model is needed to relate the dynamics of soil phosphorus (P) chemistry in a batch reactor to the soil/solution ratio and to initial P in the reactor. The Langmuir‐Hinshelwood model appears to describe this system quite well. According to this model reversible adsorption of P from solution to the colloidal surface (Langmuir component) is followed by an irreversible reaction of the surface species (Hinshelwood component). The system is an example of heterogeneous catalysis. Adsorption follows 2nd order kinetics related to solution P concentration and concentration of available surface sites for adsorption. Desorption and reaction are assumed to follow 1st order kinetics. The system is described by three simultaneous equations, two of which are 1st order nonlinear ordinary differential equations. Numerical integration is by the Euler finite difference method. Data from the literature are used to calibrate the model and to demonstrate its characteristics for a sandy soil. Rapid drop in solution P concentration is followed by a gradual decline. The first step in calibration is to assume quasi‐equilibrium in the early phase (analogous to that in enzyme kinetics), which leads to values for the total concentration of adsorption sites, So, and the Langmuir coefficient, potassium (K). The second step then estimates the adsorption coefficient, ka, and the reaction coefficient, kr. Full simulation of the transient process over time requires great care in size of time step, t, to maintain stability and accuracy in the procedure. Analysis shows a linear correlation between total surface sites, So, and soil/solution ratio, M, as expected.

Structure, morphology and surface properties of nanostructured ZrO2 particles

Nanocrystalline zirconium dioxide particles have been produced by means of thermal decomposition of a liquid metal organic precursor in a flow reactor system. X-Ray diffraction patterns show that the powder consists of a mixture of the monoclinic and tetragonal phase and that the phase fractions depend on the experimental conditions during the preparation as well as on the subsequent thermal treatment of the sample. The mean particle diameter of the synthesized powders is ca. 40-50 A as estimated from the line width of Bragg reflections and from transmission electron micrographs. The crystallites are non-porous and exhibit a rough surface incorporating a high concentration of low-coordinated surface sites. In agreement with the small particle size, the powder has a high specific surface area (>200 m 2 g –1 ), which surmounts that of commercially available materials by a factor >3. The structure and morphology of the particles is essentially preserved during extensive thermal treatment up to 500°C. Hydroxyl groups as IR active surface probes clearly indicate that the phase fractions on the surface and in the bulk strongly differ from each other. Low-coordinated OH groups on monoclinic surface domains give rise to H/D exchange reactions with D 2 already at room temperature. They are evidenced and monitored with time by FTIR spectroscopy.

A comparison of the thermodynamics of metal adsorption onto two common bacteria

The cell walls of bacteria are known to adsorb a variety of metals, and thus they may control metal mobilities in many low-temperature aqueous systems. In order to quantify metal adsorption onto bacterial surfaces, recent studies have applied equilibrium thermodynamics to the specific chemical and electrostatic interactions occurring at the solution–cell wall interface. However, to date, few studies have used this approach to compare the surface properties and metal affinities of different species of bacteria. In this study, we use acid–base titrations to determine the concentrations and deprotonation constants of specific surface functional groups on Bacillus licheniformis. The cell wall displays carboxyl, phosphate and hydroxyl surface functional groups, with pK a values and 1 s errors of 5.2±0.3, 7.5±0.4 and 10.2±0.5, respectively. We perform metal– B. licheniformis adsorption experiments using Cd, Pb, Cu and Al. The average log K values for the Cd-, Pb-, Cu- and Al–carboxyl stability constants, with 1 s errors, are 3.9±0.5, 4.6±0.3, 4.9±0.4 and 5.8±0.3, respectively. Finally, we compare the surface characteristics and metal affinities of B. licheniformis to those of Bacillus subtilis, as determined by Fein et al. [Fein, J.B., Daughney, C.J., Yee, N., Davis, T., 1997. A chemical equilibrium model of metal adsorption onto bacterial surfaces. Geochim. Cosmochim. Acta 61, 3319–3328]. Our investigations indicate that these two species of bacteria have different relative and absolute concentrations of surface sites and slightly different deprotonation and metal adsorption stability constants. We relate these variations in surface properties to variations in metal affinity in order to predict metal mobilities in complex, natural systems.

Sulphite modification of galena surfaces and its effect on flotation and xanthate adsorption

The effect of sulphite interaction with galena on the mechanism of ethyl xanthate adsorption onto galena surfaces has been studied in situ using UV spectroscopy. The influence of sulphite on galena flotation has been studied. X-ray photoelectron spectroscopic (XPS) and dissolution studies have been used to identify the mechanism of interaction between sulphite and galena surfaces. Metastable ethyl monothiocarbonate is a significant derivative of ethyl xanthate adsorption on galena in the absence of sulphite. The concentration of surface sites which produce monothiocarbonate increases with the initial state of galena oxidation, which, is in turn, apparently dependent upon galena preconditioning time, temperature and oxygen concentration. Significantly different ethyl xanthate adsorption characteristics were found, after sulphite conditioning of the galena surface. In this case, ethyl xanthate adsorption rate was significantly reduced, as was the formation of monothiocarbonate. Surface complexation of lead hydroxide by adsorbed sulphite decreases the rate of ethyl xanthate adsorption onto sites which, potentially, can produce monothiocarbonate. Galena dissolution studies and XPS examination of the galena surfaces together, confirmed the formation of insoluble lead sulphite precipitates at the galena surface. The effectiveness of sulphite depression of galena flotation is enhanced by adsorbed lead hydrolysis products on galena.

Detection and sorption study of dioxouranium(VI) ions onN-(2-mercaptopropionyl)glycine-modified silver colloid by surface-enhanced Raman scattering

An aqueous silver colloid activated at 514 nm with 10-3 M N-(2-mercaptopropionyl)glycine (MPG) involves surface-active sites for the complexation of uranyl ions and displays the νs UO22+ SER spectrum at 840 cm-1, allowing uranyl detection at concentrations as low as 5×10-9 M. This level of detection competes successfully with spectrophotometric and fluorimetric determination. A quantitative study of the SER spectra as a function of total uranyl concentrations in the colloid, ranging between 5×10-9 and 1×10-5 M, shows that half of the sorption sites are occupied by uranyl for a total uranyl concentration of about 10-7 M; from this value, a 100-fold concentration increase induces only a twofold enhancement of the uranyl band at 840 cm-1, although no appreciable change of the colloid appears. As a result of an analysis of the uranyl complexation either in solution or simultaneously with several sorbed MPG molecules, the surface site concentration for uranyl sorption in the colloid used is estimated to be close to 10-8 M. This value corresponds to about one complexation site per primary particle of silver, perhaps at contacts between aggregated particles. At such sites, MPG molecules bonded to the metal surfaces through S—Ag bonds still involve amide and carboxylate groups free to coordinate an uranyl ion. The mean enhancement of the Raman cross-section of uranyl at the surface complexation sites reaches 3×105 with respect to uranyl nitrate aqueous solution. © John Wiley & Sons, Ltd.

The Effect of the Ionic Strength on the Adsorption Isotherms of Nickel on Silica

Isotherms of adsorption of Ni(II) on silica from 0.01MNaClO4are almost linear with a log–log slope of 0.95, and the percentage of uptake at a given pH is insensitive to the initial Ni concentration. At higher ionic strengths (≥0.1 mol dm−3) two kinds of behavior are observed. For low Ni concentrations the adsorption is approximately equal to that observed at lower ionic strengths. For higher Ni concentrations (but still much lower than the concentration of surface sites) the Ni adsorption from 0.1 and 0.3 mol dm−3NaClO4is lower than the adsorption from 0.01MNaClO4by a factor of 2 (pH 9) to 3 (pH 8). When water is replaced by a mixed solvent (e.g., 2% aqueous DMSO, THF, methanol, glycerol) the Ni adsorption isotherms obtained in the presence of 0.1 mol dm−3NaClO4are linear with a log–log slope of 0.95. These adsorption isotherms are rather insensitive to the nature and concentration of the organic cosolvent. It is impossible to explain the Ni adsorption curves over the entire studied range of initial concentrations and ionic strengths in terms of surface complexation model, unless the Boltzmann factor is properly corrected.

Diamond growth chemistry: Its observation using real time in situ molecular beam scattering techniques

The chemical reactive properties of deuterated methyl radicals on CVD diamond surfaces have been studied with thermal desorption, reactive molecular beam scattering and stochastic simulation techniques, both in the absence and presence of an H-atom flux. The thermal decomposition of adsorbed methyl is shown to be strongly influenced by the concentration of free surface sites. The probability of abstraction of D from adsorbed CD3 by atomic H is 0.05 whilst the probability for removal in the form of methane (CD3H) is 0.01. The implications of the results to diamond CVD are discussed. (C) 1995 Elsevier Science S.A.

Sorption of trace metals (Cu, Pb, Zn) by suspended lake particles in artificial (0.005 M NaNO 3) and natural (Esthwaite Water) freshwaters

The sorption of Cu, Pb, and Zn onto natural lake particles, suspended in 0.005 M NaNO 3 solution and in a natural lake water (Esthwaite Water, Cumbria, UK), was studied as a function of pH and time in a series of laboratory experiments, under environmentally realistic conditions. The sorption of all three metals increased with increasing pH and reaction time (2 h and 7 days). In 0.005 M NaNO 3 solution, the well-defined sorption edges spanned 2–2.5 pH units for Cu and Pb, and ≈ 4 pH units for Zn. In the natural lake water, the Cu sorption edge was broader and both Cu and Zn were less strongly sorbed. The binding stability decreased in the order Pb>Cu>Zn. Competitive adsorption onto surface sites appeared to be the main factor determining the observed sorption behaviour. Application of a macroscopic metal exchange model to the 7 day NaNO 3 results enabled the surface site concentration to be estimated as 0.79 ± 0.07 mmol g −1. The modelling exercise suggested that an observed shift in the sorption edge of Zn, in the presence of Pb and Cu, was due to competition for surface sites. The experimental data are in good general agreement with field observations of trace metal behaviour in Cumbrian lakes. The almost total sorption of Pb by lake particles throughout the in-situ pH range is compatible with previous field measurements including trace metal budgets and residence times. Dissolved Zn concentrations in the lake are lower than predicted by the sorption experiments, but the lower lake concentrations are consistent with the previously observed scavenging of Zn by planktonic algae. Both the decreased sorption of Cu in the experiments with natural lake water, compared to that in NaNO 3 solution, and the relatively small-scale removal of dissolved Cu by particles in the lake itself can partially be explained by humic complexation.

Kinetics of Halogenated Organic Compound Degradation by Iron Metal

A combination of new and previously reported data on the kinetics of dehalogenation by zero-valent iron (Fe0) has been subjected to an analysis of factors effecting contaminant degradation rates. First-order rate constants (kobs) from both batch and column studies vary widely and without meaningful correlation. However, normalization of these data to iron surface area concentration yields a specific rate constant (kSA) that varies by only 1 order of magnitude for individual halocarbons. Correlation analysis using kSA reveals that dechlorination is generally more rapid at saturated carbon centers than unsaturated carbons and that high degrees of halogenation favor rapid reduction. However, new data and additional analysis will be necessary to obtain reliable quantitative structure−activity relationships. Further generalization of our kinetic model has been obtained by accounting for the concentration and saturation of reactive surface sites, but kSA is still the most appropriate starting point for design calculations. Representative values of kSA have been provided for the common chlorinated solvents.

Microbial Reduction of Crystalline Iron(III) Oxides: Influence of Oxide Surface Area and Potential for Cell Growth

Quantitative aspects of microbial crystalline iron(III) oxide reduction were examined using a dissimilatory iron(III) oxide-reducing bacterium (Shewanella alga strain BrY). The initial rate and long-term extent of reduction of a range of synthetic iron(III) oxides were linearly correlated with oxide surface area. Oxide reduction rates reached an asymptote at cell concentrations in excess of ≈1 × 109/m2 of oxide surface. Experiments with microbially reduced goethite that had been washed with pH 5 sodium acetate to remove adsorbed Fe(II) suggested that formation of a Fe(II) surface phase (adsorbed or precipitated) limited the extent of iron(III) oxide reduction. These results demonstrated explicitly that the rate and extent of microbial iron(III) oxide reduction is controlled by the surface area and site concentration of the solid phase. Strain BrY grew in media with synthetic goethite as the sole electron acceptor. The quantity of cells produced per micromole of goethite reduced (2.5 × 106) was comparable to that determined previously for growth of BrY and other dissimilatory Fe(III)-reducing bacteria coupled to amorphous iron(III) oxide reduction. BrY reduced a substantial fraction (8−18%) of the crystalline iron(III) oxide content of a variety of soil and subsurface materials, and several cultures containing these materials were transferred repeatedly with continued active Fe(III) reduction. These findings indicate that Fe(III)-reducing bacteria may be able to survive and produce significant quantities of Fe(II) in anaerobic soil and subsurface environments where crystalline iron(III) oxides (e.g., goethite) are the dominant forms of Fe(III) available for microbial reduction. Results suggest that the potential for cell growth and Fe(II) generation will be determined by the iron(III) oxide surface site concentration in the soil or sediment matrix.

The role of sorbed humic substances on the distribution of organic and inorganic contaminants in groundwater

Mineral-bound humic substances modify inorganic surfaces in subsurface sediments, changing the nature and number of complexation sites for contaminants. Because of adsorptive enrichment, the reactive surface area or site concentration contributed by mineral-bound humic substances can exceed that of dissolved or colloidal humic substances by two orders of magnitude. Mineral-bound humic materials may, therefore, provide a major sink for the removal of contaminants in groundwater. The reactivity of the humic substance is primarily determined by the structural and bulk chemical properties of the humic substance and the aqueous solution chemistry. Organic and inorganic contaminants sorb readily to mineral-bound humic substances. The sorption of hydrophobic organic compounds increases as ionic strength decreases, is enhanced by divalent cations, and displays non-linear isotherms and competitive adsorption behavior. Collectively, these results suggest that hydrophobic adsorption, rather than phase partitioning, is the primary sorption mechanism for neutral organic molecules on these particle coatings. Mineral-bound humic substances augment, rather than change, the intrinsic complexation properties of mineral surfaces for metal cations. The degree of sorption enhancement promoted by mineral-bound organic material varies strongly with pH and depends on the magnitude of the stability constants between the metal cation and the humic substance, the strength and magnitude of adsorption of the humic substance by the mineral surface, and the extent of aqueous complex formation between the non-sorbed humic substance and metal. The simplest sorption model for humate-modified surfaces is the linear additivity model (LAM). Sorption data for certain hydrophobic organic compounds and metal cations appear to conform to this model.

Laser Temperature Jump Relaxation Measurements of Adsorption/Desorption Kinetics at Liquid/Solid Interfaces

The iodine laser temperature jump method is used to study adsorption/desorption kinetics at a methylated silica/solution interface. A suspension of C1-derivatized fumed silica is used for the kinetic measurements. The colloidal silica does not significantly change the attenuation of near-IR radiation from the iodine laser and allows the surface site concentration to be varied so that adsorption and desorption rates can be determined. The temperature jump relaxation method was used to investigate the effect of electrolyte on adsorption of a charged solute (ANS) on a C1 silica surface. Adsorption equilibrium conditions were optimized to observe a maximum relaxation signal. Without electrolyte, the relaxation signal is biexponential, which is also reflected in a broad chromatographic peak shape and a two-site sorption isotherm. When electrolyte is added, the relaxation signal is primarily single exponential, which agrees with the linear adsorption isotherm. The adsorption rate and equilibrium constant were found to increase significantly with added electrolyte, which showed that adsorption kinetics can influence both band broadening and retention. 28 refs., 7 figs., 4 tab.

Chemical Relaxation and Double Layer Model Analysis of Boron Adsorption on Alumina

AbstractBoron is a plant nutrient essential for adequate plant growth, yet the range between B deficiency and toxicity levels is rather narrow. Boron adsorption reactions with soil components, particularly sesquioxides, most often regulate the amount of B in the soil solution. The reaction mechanisms of B adsorption on oxides have not been fully characterized, however. Pressure‐jump relaxation experiments were conducted to measure the rates and determine the reaction mechanism for B adsorption on an alumina (γ‐Al2O3) surface from B(OH)3‐B (OH)−4 solutions. Relaxation times (τ) were measured from pH 7.0 to 9.7 in alumina suspensions with 0.012 mol L−1 total B. A plot of τ−1 vs. B(OH)−4 plus surface site concentration obtained from the triple layer model (TLM) assuming inner sphere B(OH)−4 adsorption yielded an adsorption rate constant (kintf) of 3.3 × 105 L mol−1 s−1 and a desorption rate constant (kintr) of 1.8 × 10−3 L mol−1 s−1. The ratio kintf/kintr yielded an equilibrium constant (log KintKIN) of 8.26, in agreement with the intrinsic equilibrium constant for B(OH)−4 adsorption (log KintBIS = 7.69) obtained from adsorption isotherms. Four additional surface complexation models were tested for their ability to model both the equilibrium and kinetic data simultaneously: the constant capacitance model, the diffuse layer model, a Stern model variant, and the TLM assuming outer sphere B(OH)−4 adsorption. Only the TLM, assuming both B(OH)3 and B(OH)−4 were adsorbed via ligand exchange on neutral surface sites, was successful. The TLM indicated that B(OH)−4 is the predominant adsorbed species throughout the pH range 7.0 to 10.8.

Elastomer–carbon black interaction: Influence of elastomer chemical structure and carbon black surface chemistry on bound rubber formation

AbstractThe contribution of elastomer polarity and reactivity to bound rubber formation has been investigated. In this study, a number of elastomers of different chemical nature have been tested. The surface of the carbon black (N110) has also been modified by nitric acid oxidation in order to increase the concentration of surface functional groups. The experimental results have shown that the bound rubber formation is barely related to the polarity of the polymers. It is the reactive sites in both the elastomer and carbon black which are mainly responsible for bound rubber formation. It therefore appears that the elastomer/carbon black interaction leading to the formation of bound rubber is essentially a chemical process involving primary bond formation between elastomer and carbon black. The oxidized carbon black exhibits a higher surface activity which may be due to an increased concentration of oxygen‐containing reactive surface sites, namely, phenolic hydroxyl, carboxyl, lactone, and quinone groups. © 1995 John Wiley & Sons, Inc.

Stability constants for silicate adsorbed to ferrihydrite

AbstractIntrinsic surface acidity constants (Kalintr, Ka2intr) and surface complexation constant for adsorption of orthosilicate onto synthetic ferrihydrite (Ksi for the complex ≡FeOSi(OH)3) have been determined from acid/base titrations in 0.001-0.1 M NaClO4 electrolytes and silicate adsorption experiments in 0.01 M NaNOi electrolyte (pH 3-6). The surface equilibrium constants were calculated according to the two-layer model by Dzombak ' Morel (1990). Near equilibrium between protons/hydroxyls in solution and the ferrihydrite surface was obtained within minutes while equilibration with silicate required days-weeks, both reactions probably being diffusion controlled. Applying the values for specific surface area and site densities for ferrihydrite used by Dzombak ' Morel (1990) (600 m2 g–1, 3.4 μmole m–2) the constants pKalintr = 6.93 ± 0.12, pKa2intr = 8.72 ± 0.17 and log Ksi = 3.62 were calculated by using the FITEQL optimization routine. Use of the specific surface area actually measured (269 m2 g-1) gave a poorer fit of the experimental data. Due to the slow adsorption of silicate and hence long shaking times, changes in the surface characteristics of the ferrihydrite seem to take place, probably a decrease in the concentration of surface sites. Adsorption isotherms calculated using the derived equilibrium constants showed that approximately twice the amount of silicate was adsorbed at pH 5 compared with pH 3.Infrared spectroscopy of silica adsorbed to ferrihydrite showed Si-O stretching absorption maxima in the range 940-960 cm-1. The shift of the absorption maximum to higher wavenumbers with increasing amount of silicate adsorbed is probably due to an increase in the frequency of Si-O-Si bonds between orthosilicate adsorbed at adjacent sites. Small amounts of goethite were identified in the adsorption products.