Complexation mechanism of Pb2+ on Al-substituted hematite: A modeling study and theoretical calculation.

Hematite nanoparticles commonly undergoes isomorphic substitution of Al3+ in nature, while how the Al-substitution-induced morphological change, defective structure and newly generated Al-OH sites affect the adsorption behavior of hematite for contaminants remains poorly understood. Herein, the interfacial reactions between Al-substituted hematite and Pb2+ was investigated via CD-MUSIC modeling and DFT calculations. As the Al content increased from 0% to 9.4%, Al-substitution promoted the proportion of (001) facets and caused Fe vacancies on hematite, which increased the total active site density of hematite from 5.60 to 17.60 sites/nm2. The surface positive charge of hematite significantly increased from 0.096 to 0.418 C/m2 at pH 5.0 due to the increases in site density and proton affinity (logKH) of hematite under Al-substitution. The adsorption amount of hematite for Pb2+ increased from 3.92 to 9.74 mmol/kg at pH 5.0 and 20 µmol/L initial Pb2+ concentration with increasing Al content. More Fe vacancies may lead to a weaker adsorption energy (Ead) of hematite for Pb2+, while the Ead was enhanced at higher Al content. The adsorption affinity (logKPb) of bidentate Pb complexes slightly increased while that of tridentate Pb complexes decreased with increasing Al content due to the presence of ≡ AlOH-0.5 and ≡ Fe2AlO-0.5 sites. Tridentate Pb complexes were dominant species on the surface of pure hematite, while bidentate ones became more dominant with increasing Al content. The obtained model parameters and molecular scale information are of great importance for better describing and predicting the environmental fate of toxic heavy metals in terrestrial and aquatic environments.

Electronic-Level Insight into the Adsorption and Surface Diffusion Kinetics of a Simplified Glyphosate Model on a Goethite Surface.

The diffusive processes that occur in minerals involve chemical and physical surface phenomena of great interest that allow for understanding the mobility of different anions of environmental importance. One of them is glyphosate, which is widely used as a pesticide. In this work, we performed Hubbard-corrected density functional theory (DFT + U) calculations to study the adsorption and surface diffusion of methylphosphonic acid (MPA), as a model of glyphosate, on the (010) plane of goethite (GOT), one of the most important Fe(III) minerals in soils and sediments. In particular, the MPA adsorption was studied at the GOT-water interface, finding a strong covalent character in the bond. We also corroborated the occurrence of double proton transfer (MPA to GOT and GOT to GOT). Finally, activation energy barriers were calculated to estimate the half-lives for molecular diffusion, showing that MPA moves almost 3000 times slower than water at the GOT surface.

Proton binding to humic nano particles: electrostatic interaction and the condensation approximation.

Proton binding to "carboxylic" and "phenolic" sites of humic nano particles (HNPs) is determined by the total proton affinity that is due to a specific and an electrostatic affinity. Both affinities are accounted for in the bi-modal Langmuir-Freundlich (bi-LF)-equation extended with a Boltzmann factor that includes the electrostatic site potential(s), y. For y → 0 the equation reduces to the bi-LF Master Curve (MC). Commonly, an electrical double layer model is used to obtain y, e.g., the bi-LF-Donnan-Vapp (monocomponent NICA-Donnan) model and bi-LF-soft-particle-Poison-Boltzmann-Theory (SPBT). A new method is presented that combines the "condensation approximation" (CA) with the MC concept (CA-MC). With the CA, the proton binding curve and MC can be transformed in, respectively, the total and specific affinity distribution. The difference at a given charge density provides the electrostatic affinity and CA-potentials vs. charge density. The MC can be obtained theoretically or by using the convention that the electrostatic interaction is negligible at 1 M salt concentration. For five HNPs CA-potentials corresponding with the bi-LF-SPBT are compared with results of the bi-LF-Donnan-Vapp model using the MC(SPBT). The bi-LF-Donnan-Vapp model fails when the Debye length > hydrated particle radius. The CA-MC(1M) method does not require characteristics of the HNPs. Combination of the bi-LF-eq. with the CA-MC(1 M) method gives the bi-LF-CA-MC(1 M) model. The CA-MC(1 M) differs from the MC(SPBT); therefore, resulting parameters can only be compared when the same method is used.

Theoretical study of the octahedral substitution effect in delaminated pyrophyllite: physicochemical properties and applications.

We have studied, using DFT calculations, some geometrical and electronic properties of delaminated pyrophyllite (D-P) and the corresponding layers that result from three isomorphic substitutions on octahedral sheets (Mg2+, Fe2+ and Fe3+). Bond lengths, layer thickness (dL), band gap (Eg), work function (WF), magnetic moment (µ), density of states and charge distributions are reported. These properties are important to control the behaviour of electronic devices. The results show that the layer thickness increases according to the ionic radius of the considered substituent. In the case of the three substitutions a reduction of the forbidden band is observed. Mg2+ induces a decrease in the Eg value of about 16.5% with respect to D-P, whereas for Fe this reduction is more significant due to the presence of trap states in the forbidden zone. For Fe2+ (Fe3+) the reduction in the Eg is around 62% (51%). Regarding the WF, our results showed that there is a decrease in its value after substitution. D-P has the highest WF value (8.15 eV), whereas the delaminated clay with Fe2+ has the lowest value (2.22 eV). Finally, D-P and D-P substituted with Mg2+ have a diamagnetic behaviour (µ = 0), whereas the presence of Fe2+ and Fe3+ induces a paramagnetic behaviour. The computed magnetic moment is 4 µB and 1 µB for D-P substituted with Fe2+ and Fe3+, respectively.

Insights of competitive adsorption on activated carbon of binary caffeine and diclofenac solutions.

A commercial activated carbon (AC), obtained from peanut shells, was characterized and tested as adsorbent for the removal of the pharmaceutical products caffeine (CF) and diclofenac (DIC), which were used as model emerging contaminants. Nitrogen adsorption, XRD, SEM, FT-IR spectroscopy, and chemical analyses were typical of ACs, and Boehm titrations, calculations of surface sites distributions and zeta potential measurements indicated that reactions of deprotonable oxygenated groups at the AC surface lead to an isoelectric point of 3.2. A theoretical equation derived from the Langmuir isotherm is proposed to explain the adsorption percentage or adsorbed fraction (fads) as a function of the adsorbent dose (D, adsorbent "concentration"). Good fittings of the fads vs. D curves and the normal adsorption isotherms were obtained with the same Langmuir parameters. An important and practical application of this new equation is to permit a straightforward calculation of the solid dose needed to achieve a required adsorption percentage. With the aim of describing the adsorption processes of CF and DIC and their competition for surface sites under an ample range of concentrations, the adsorption of the emerging contaminants was investigated in single adsorbate experiments and with binary mixtures, and the competitive Langmuir model was applied. CF adsorption was high and independent of pH, whereas DIC adsorption was high between pH 4 and 6 and showed a continuous decrease from pH 6 to 10.5. The use of the competitive Langmuir isotherm for binary mixtures indicated that there was no pure competition between CF and DIC for surface sites. Instead, there was influenced competition, meaning that the presence of one substance at the surface modified the adsorption parameters of the other, either through lateral interaction forces or by changing the molecular orientation at the surface. In both cases, one substance favored the adsorption of the other, compared to pure competition.

Determining rate coefficients for ion adsorption at the solid/water interface: better from desorption rate than from adsorption rate.

One of the most common approaches in the adsorption kinetics literature is to compare the fitting performance of several empirical or non-empirical equations (pseudo-first order, pseudo-second order, Elovich, parabolic diffusion, etc.) with the aim of selecting the equation that best describes the experimental data. This is normally a futile fitting exercise that leads to the determination of ambiguous rate parameters, without providing insights into the behaviour of the studied system. A more realistic approach is to treat it as a combination of mass transport and chemical reaction under controlled conditions, and thus actual adsorption-desorption rate parameters are readily estimated. This article applies a simple and realistic physicochemical model to describe and understand the adsorption-desorption kinetics of ions at the solid/water interface. The model is applied to an ATR-FTIR study of phosphate adsorption-desorption on goethite, which is a very well-known and reference system, ideal for testing the performance of a physicochemical treatment that combines transport and reaction. Always the same phosphate species (monodentate mononuclear protonated) was present at the goethite surface during adsorption-desorption. There was an excellent agreement between theory and experiments at a variety of phosphate concentrations and surface coverages for adsorption kinetics, desorption kinetics and equilibrium situations, employing just one set of rate coefficients. The use of rate vs. adsorption curves permitted easy detection of conditions of transport- and reaction-controlled kinetics. The phosphate-goethite system is a fast-adsorbing/slow-desorbing system, with an adsorption rate constant k = 1.26 × 103 s-1 and a desorption rate constant kd = 1.66 × 10-5 s-1. Therefore, adsorption was transport-controlled and desorption was reaction-controlled. The half-life of the desorption reaction is 41 700 s (11.6 h) but for adsorption it would take only a few seconds in the absence of transport control. For this kind of system, which is ubiquitous in nature and technological processes, it is easier to determine rate constants from desorption than from adsorption experiments.

Fractionation and leaching of Cd, Cu, Fe, Pb, and Zn from smelter residues of an old environmental liability in Argentina.

An integrated chemical and mineralogical characterization approach was applied to smelter wastes collected from 50-year-old dump sites in Argentina. Characterization included pseudo-total element concentrations, acid generation/neutralization potential, sequential extractions, pH-dependent leaching kinetics, and mineralogical analysis of all residues. These analyses provided detailed information on the reactivity of the minerals in the waste material and associated metal release. Cadmium and Zn were the elements of greatest environmental concern due to their high mobility. On average, the release of Zn and Cd in pH-dependent leaching essays reached 17.6% (up to 5.24 mg g-1) and 52.7% (up to 0.02 mg g-1) of the pseudo-total content, respectively. Moreover, Cd and Zn were also the metals that showed the higher proportions of labile fractions associated to the adsorbed and exchangeable fraction (60-92% for Cd and 19-38% for Zn). Since Cd and Zn concentrations in the residue are not high enough to form their own minerals, a large proportion of these elements would be weakly adsorbed on Fe oxyhydroxides. In contrast, the low release of Cu, Pb and Fe would be associated with these elements being incorporated into the crystalline structure of insoluble or very poorly soluble minerals. Lead is incorporated into plumbojarosite and anglesite. Copper was mainly in association with Fe oxyhydroxides and may also have been incorporated into the plumbojarosite structure. The latter could act as a sink especially for Pb under the acidic conditions of the smelter residue. Despite the elevated concentrations of Pb observed in the residue, it showed a very low mobility (≈0.1%), indicating that it is mostly stabilized. Nevertheless, the smelter residue is a continuous source of metals requiring remediation.

Measuring the isoelectric point of the edges of clay mineral particles: the case of montmorillonite.

The isoelectric point (IEP) of the edge surface of a montmorillonite sample was determined by using electrophoretic mobility measurements. This parameter, which is fundamental for the understanding of the charging behavior of clay mineral surfaces, was never measured so far because of the presence of permanent negative charges within the montmorillonite structure, charges that mask the electrokinetic behavior of the edges. The strategy was to block or neutralize the structural charges with two different cations, methylene blue (MB(+)) and tetraethylenepentaminecopper(II) ([Cu(tetren)](2+)), so that the charging behavior of the particles becomes that of the edge surfaces. Adsorption isotherms of MB(+) and [Cu(tetren)](2+) at different ionic strengths (NaCl) were performed to establish the uptakes that neutralize the cation exchange capacity (CEC, 0.96 meq g(-1)) of the sample. At high adsorptive concentrations, there was a superequivalent adsorption of MB(+) (adsorption exceeding the CEC) and an equivalent adsorption of [Cu(tetren)](2+) (adsorption reaching the CEC). In both cases, structural charges were neutralized at uptakes very close to the CEC. Zeta potential (ζ) vs pH data at different ionic strengths of montmorillonite with adsorbed MB(+) allowed to estimate an upper limit of the edge's IEP, 5.3 ± 0.2. The same kind of data obtained with adsorbed [Cu(tetren)](2+) provided a lower limit of the IEP, 4.0 ± 0.2. These values are in agreement with previously informed IEP and point of zero charge of pyrophyllite, which is structurally analogous to montmorillonite but carries no permanent charges. The importance of knowing the IEP of the edge surface of clay minerals is discussed. This value characterizes the intrinsic reactivity of edges, that is, the protonating capacity of edge groups in absence of any electric field generated by structural charges. It also allows us to correct relative edge charge vs pH curves obtained by potentiometric titrations and to obtain the true edge charge vs pH curves at different electrolyte concentrations.

The two-species phosphate adsorption kinetics on goethite.

Studying the adsorption-desorption kinetics of ions and molecules is crucial to understand the mobility of nutrients and pollutants in the environment. This article reports the adsorption-desorption kinetics of phosphate on goethite, as measured by ATR-FTIR spectroscopy at pH 4.5, 7.0 and 9.5. The system phosphate-goethite has become a model system to test new experimental setups and theories to understand the behavior of pollutants with phosphonic or phosphinic moieties such as glyphosate or glufosinate. One of the main difficulties in the analysis of ATR-FTIR spectra in adsorption-desorption kinetics is to calibrate the equipment to convert absorbance vs. t curves into adsorption vs. t curves, and thus the methodology to achieve a good calibration using spectroscopic data in combination with adsorption isotherms is clearly described. The time evolution of the different surface species was monitored simultaneously during adsorption and desorption at different pH, showing the advantages of this spectroscopy over traditional adsorption methods that only quantify total adsorption. Results were analysed in terms of a simple adsorption-desorption model that takes into account transport, attachment, detachment and surface transformation of the adsorbed species. The same rate parameters at a given pH could predict well the adsorption-desorption kinetics of the two formed surface species and the corresponding adsorption isotherm, giving new insights into the dynamics of phosphate on the surface of goethite. It was found that phosphate desorbed faster from goethite at low pH than at high pH, which is counterintuitive, but has good practical and environmental applications.

Ciprofloxacin in Layered Double Hydroxides: Looking for the Best Synthesis Method.

It is important to develop new methods of release to improve pharmacokinetic parameters of drugs, especially antibiotics, whose plasmatic concentration is determinant to ensure an effective treatment. Layered double hydroxides (LDH) are inorganic and biocompatible materials with high drug intercalation capacity and release properties that can be tuned by controlling the pH value. These materials can be an excellent choice to achieve a sustained release and an optimal drug concentration in plasm. In this work, LDH were synthesized with intercalated ciprofloxacin (CIP) by three different methods: coprecipitation, reconstruction and ion exchange. LDH-CIP complexes were characterized by XRD, TG-DSC, TEM, SEM, FTIR, electrophoretic mobilities, and drug release and dissolution kinetics in NaCl solutions and under physiological conditions. The coprecipitation and reconstruction methods lead to the formation of ill-defined products, whereas the ion exchange method rendered the best intercalation results. CIP release was controlled by dissolution at pH<3 and by desorption and ion exchange at intermediate and high pH. In comparison with a commercial formulation, the LDH-CIP complex prepared by ion exchange presented a slower release profile. The fast dissolution at gastric pH raises the need of developing some type of coating for protecting LDH materials.

Magnetic Mesoporous Silica Nanoparticles for Drug Delivery Systems: Synthesis, Characterization and Application as Norfloxacin Carrier.

Mesoporous silica nanoparticles, with and without the inclusion of a magnetic core, were hydrothermally synthesized and employed as carrier of the antibiotic norfloxacin (NFX). The antibiotic-loaded materials were prepared by wet impregnation. Differences in drug content (and in further release profile) were directly related to changes in surface area, particle aggregation and hydrophobicity of the solids. The kinetics of NFX release has been studied in batch experiments. In all cases, more than 55% of the antibiotic was quickly desorbed during the first 5 min due to the localization of NFX on the external surface of the nanoparticles. The rest of the drug (situated inside the mesopores) was released through a diffusion-controlled transport and the rate was strongly dependent of the pH, reaching its minimum value at neutral pH. The calculated activation energy confirmed that the release was controlled by a diffusion process. Breaking of H-bonds and electrostatic and hydrophobic interactions appear to be responsible for NFX desorption from the solid surface. Such interactions increase, however, the thermal stability of the drug when the NFX and the carriers are combined. The antimicrobial activities of the drug loaded nanoparticles and the free antibiotic were compared and discussed.

Synthesis of Layered Double Hydroxides Intercalated With Drugs for Controlled Release: Successful Intercalation of Ibuprofen and Failed Intercalation of Paracetamol.

This work examines the effect of drug structure and ionization degree on the formation and properties of biocompatible layered double hydroxides (LDH) intercalated with ibuprofen and paracetamol. Ibuprofen (pKa = 5.3) is in its anionic form, whereas paracetamol (pKa 9.4) is only partially ionized at the synthesis pH (9.0), and thus intercalation is expected to be different in the two cases. Chemical analyses, X-ray diffraction, electron microscopy, infrared spectroscopy and thermal analyses were applied to characterize the materials. Dissolution kinetics and drug release kinetics were also investigated, in an ample range of pH (3.0-9.0) in NaCl solutions, and in physiological buffers (1.2, 4.5 and 6.8). All characterization techniques showed that an efficient intercalation of ibuprofen took place, resulting in a material with 30% of its weight corresponding to the drug. On the contrary, all techniques revealed a very poor intercalation of paracetamol (1.2%). The dissolution kinetics of LDHs was highly pH-dependent, being higher as pH decreased. The drug release kinetics, conversely, increased as pH increased. In physiological buffers the release rate depended not only on the pH but also on the type of buffer. This last behavior is useful to control the release in different parts of the digestive system.

Equilibrium mono- and multicomponent adsorption models: From homogeneous ideal to heterogeneous non-ideal binding.

Aqueous sorption processes play an important role in, for example, pollutant binding to natural nanoparticles, colloid stability, separation and enrichment of components and remediation processes. In this article, which is a tribute to Hans Lyklema, models of localized (ad)sorption of molecules and ions from aqueous solution on homogeneous and heterogeneous nanoparticles are presented. The discussed models range from ideal monocomponent sorption on homogeneous (Langmuir) and heterogeneous sites, to multicomponent ideal sorption on homogeneous and heterogeneous sites, multicomponent multisite ion complexation with charge distribution (CD-MUSIC) and non-ideal competitive adsorption on heterogeneous sites (NICA). Attention is also paid to lateral interaction, site-induced aggregation, binding stoichiometry and multilayer formation. Electrical double layer models are discussed in relation to ion binding on impermeable and permeable nanoparticles. Insight in models that can describe sorption of molecules and ions on nanoparticles leads to awareness of the limitations of using simple models for complex systems and is needed for the selection and application of an appropriate model for a given system. This is relevant for all practical sorption processes and for a better understanding of the role of natural nanoparticles in the binding of nutrients and pollutants.

On the dissolution kinetics of humic acid particles. Effect of monocarboxylic acids.

The dissolution kinetics of humic acid particles has been studied in batch experiments, and the effects of monocarboxylic (formic, acetic, and propionic) acids are reported. The dissolution rate of the particles is significantly affected by the presence of monocarboxylic acids in the pH range 4-10. At pH 7, for example, propionic acid increases 30 times this dissolution rate. The capacity of increasing the dissolution rate is in the order formic acid

An ATR-FTIR study of different phosphonic acids in aqueous solution.

An ATR-FIR study of the vibrational spectra of 1-hydroxyethane-1,1'-diphosphonic acid (HEDP), nitrilotris(methylenephosphonic acid) (NTMP) and N,N-bis(2-hydroxyethyl)aminomethylphosphonic acid (BHAMP) in aqueous solution is presented. The study was performed in the range of pH from 5 to 9, and bands assignments are given in the 2000-890 cm(-1) range. However, as phosphonates display bands due to the PO stretching vibration mainly in the 900-1200 cm(-1) range, the study is focused in this midinfrared region, which shows important changes as the pH changes, specially the nu(POH) at approximately 925 cm(-1) and nu(PO(3)(2-)) at approximately 970 cm(-1) vibrations. IR analyses give also evidences for the zwitterionic nature of BHAMP and NTMP in solution with a strong indication that the zwitterion in both compounds remains intact throughout the pH range investigated. The successive protonation steps with the decrease of pH were evidenced in the IR spectra of the three studied phosphonates.

Surface speciation of phosphate on goethite as seen by InfraRed Surface Titrations (IRST).

Phosphate adsorption at the metal oxide-water interface has been intensely studied, and the system phosphate-goethite in aqueous media is normally used as a model system with abundant information regarding adsorption-desorption under very different conditions. In spite of this, there is still discussion on whether the main inner-sphere surface complexes that phosphate forms on goethite are monodentate or bidentate. A new spectroscopic technique, InfraRed Surface Titration (IRST), is presented here and used to systematically explore the surface speciation of phosphate on goethite in the pH range 4.5-9.5 at different surface coverages. IRST enabled to construct distribution curves of surface species and distribution curves of dissolved phosphate species. In combination with the CD-MUSIC surface complexation model it was possible to conclude that surface complexes are monodentate. Very accurate distribution curves were obtained, showing a crossing point at pH5.5 at a surface coverage of 2.0µmolm-2, with a mononuclear monoprotonated species predominating at pH>5.5 and a mononuclear diprotonated species prevailing at pH<5.5. On the contrary, at the low surface coverage of 0.7µmolm-2 there is no crossing point, with the mononuclear monoprotonated species prevailing at all pH. IRST can become a powerful technique to investigate structure, properties and reactions of any IR-active surface complex at the solid-water interface.

Adsorption kinetics of phosphate and arsenate on goethite. A comparative study.

The adsorption kinetics of phosphate and arsenate on goethite is studied and compared. Batch adsorption experiments were performed at different adsorbate concentrations, pH, temperatures and stirring rates. For both oxoanions the adsorption rate increases by increasing adsorbate concentration, decreasing pH and increasing temperature. It does not change by changing stirring rate. The adsorption takes place in two processes: a fast one that takes place in less than 5 min and a slow one that takes place in several hours or more. The rate of the slow process does not depend directly on the concentration of phosphate or arsenate in solution, but depends linearly on the amount of phosphate or arsenate that was adsorbed during the fast process. Apparent activation energies and absence of stirring rate effects suggest that the slow process is controlled by diffusion into pores, although the evidence is not conclusive. The similarities in the adsorption kinetics of phosphate and arsenate are quantitatively shown by using a three-parameters equation that takes into account both the fast and the slow processes. These similarities are in line with the similar reactivity that phosphate and arsenate have in general and may be important for theoretical and experimental studies of the fate of these oxoanions in the environment.

Mechanisms of soil humic acid adsorption onto montmorillonite and kaolinite.

To explore the adsorption mechanisms of a soil humic acid (HA) on purified kaolinite and montmorillonite, a combination of adsorption measurements, attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy and isothermal titration calorimetry (ITC) was employed at pH 4.0, 6.0 and 8.0. The adsorption affinities and plateaus of HA on the two clays increased with decreasing pH, indicating the importance of electrostatic interaction. The effects were more significant for kaolinite than for montmorillonite. The substantial adsorption at pH 8.0 indicated hydrophobic interaction and/or H-bonding also played a role. The ATR-FTIR results at pH 8.0 showed that the Si-O groups located at basal faces of the two clays were involved in the adsorption process. For kaolinite, at pH 4.0 and 6.0, HA adsorption occurred via OH groups on the edge faces and basal octahedral faces (both positively charged), plus some adsorption at Si-O group. The exothermic molar adsorption enthalpy decreased relatively dramatically with adsorption up to adsorption values of 0.7µmol/g on montmorillonite and 1.0µmol/g on kaolinite, but the decrease was attenuated at higher adsorption. The high exothermic molar enthalpy of HA binding to the clays was ascribed to ligand exchange and electrostatic binding, which are enthalpy-driven. At high adsorption values, JGHA adsorption by hydrophobic attraction and H-bonding also occurs.

On the mechanisms of dissolution of montroydite [HgO(s)]: dependence of the dissolution rate on pH, temperature, and stirring rate.

The dissolution behavior of montroydite (HgO) has been studied using a fully automated system. Dissolution data under equilibrium conditions are in agreement with previously published data and indicate that HgO solubility is relatively high and constant between pH 4 and 10.1 and increases markedly at pH<4. The dissolution rate also has similar behavior: it is relatively high and constant between pH 4 and 10.1 and increases sharply at pH<4. The dissolution process obeys a three-dimensional contraction or attrition mechanism. The dissolution rate increases with increasing temperature and stirring rate and is the result of mixed transport and reaction control. The rate of HgO dissolution is considerably higher than that of other divalent metal oxides at low pH. This high rate is due to the ability of Hg(II) to rapidly exchange its ligands. Data suggest that montroydite will only occur in nature in highly contaminated sites and indicate that Hg oxidation products that are formed at the liquid Hg/water interface may dissolve rapidly.

Kinetics of phosphate adsorption on goethite: comparing batch adsorption and ATR-IR measurements.

The adsorption kinetics of phosphate on goethite has been studied by batch adsorption experiments and by in situ ATR-IR spectroscopy at different pH, initial phosphate concentrations and stirring rates. Batch adsorption results are very similar to those reported by several authors, and show a rather fast initial adsorption taking place in a few minutes followed by a slower process taking place in days or weeks. The adsorption kinetics could be also monitored by integrating the phosphate signals obtained in ATR-IR experiments, and a very good agreement between both techniques was found. At pH 4.5 two surface complexes, the bidentate nonprotonated (FeO)(2)PO(2) and the bidentate protonated (FeO)(2)(OH)PO complexes, are formed at the surface. There are small changes in the relative concentrations of these species as the reaction proceeds, and they seem to evolve in time rather independently. At pH 7.5 and 9 the dominating surface species is (FeO)(2)PO(2), which is accompanied by an extra unidentified species at low concentration. They also seem to evolve independently as the reaction proceeds. The results are consistent with a mechanism that involve a fast adsorption followed by a slow diffusion into pores, and are not consistent with surface precipitation of iron phosphate.