Measurement of sorption isotherms with flow-through reactors

Column versus batch methods for measuring PFOS and PFOA sorption to geomedia

The objective of this study is to compare the consistency between column and batch experiment methods for measuring solid-phase sorption coefficients and isotherms for per and polyfluoroalkyl substances (PFAS). Perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) are used as representative PFAS, and experiments are conducted with three natural porous media with differing geochemical properties. Column-derived sorption isotherms are generated by conducting multiple experiments with different input concentrations (multi-C₀ method) or employing elution-front integration wherein the entire isotherm is determined from a single breakthrough curve (BTC) elution front. The isotherms generated with the multi-C₀ column method compared remarkably well to the batch isotherms over an aqueous concentration range of 3-4 orders of magnitude. Specifically, the 95% confidence intervals for the individual isotherm variables overlapped, producing statistically identical regressions. The elution-front integration isotherms generally agreed with the batch isotherms, but exhibited noise and systematic deviation at lower concentrations in some cases. All data sets were well described by the Freundlich isotherm model. Freundlich N values ranged from 0.75 to 0.81 for PFOS and was 0.87 for PFOA and are consistent with values reported in the literature for different geomedia. The results of this study indicate that column and batch experiments can measure consistent sorption isotherms and sorption coefficients for PFOS and PFOA when robust experimental setup and data analysis are implemented.

Arsenite and chromate sequestration onto ferrihydrite, siderite and goethite nanostructured minerals: Isotherms from flow-through reactor experiments and XAS measurements

Sorption isotherms remain a major tool to describe and predict the mobility of pollutants in natural and anthropogenic environments, but they are typically determined by independent batch experiments. In the present study, the sequestration of As(III), Cr(VI) and competitive As(III)-Cr(VI) on/in 6L-ferrihydrite, siderite and goethite nanostructured minerals was reinvestigated using stirred flow-through reactor experiments. Herein, sorption isotherms were particularly determined from breakthrough curves for inert and reactive tracers monitored simultaneously in a single percolation experiment. In complement, X-ray absorption spectroscopy (XAS) was used to identify As sorption sites on 6L-ferrihydrite and goethite. As expected, the minerals have high potential to remove As and Cr from water (siderite = ferrihydrite (about 60 mg/g) > goethite (20 mg/g)). As and Cr sorption isotherms were modelled with a Langmuir model, and with a sigmoidal Hill model in the case of the competitive sorption. XAS measurements have revealed that As(III) was partially oxidized (up to 22%) in the competitive system with chromate oxyanion Cr(VI). As(III) sorbed on ferrihydrite and goethite adopted edge-sharing and corner sharing complex geometries. Nowadays, a new class of adsorbing phases is being developed for wastewater treatment, including engineered nanostructured materials and nanocomposites. The use of flow through reactor experiments as a high throughput method, combined with XAS, should be considered as efficient screening methods to test their sorbing properties on various contaminants.

Examination of current adsorption models for Pb(II) and Cu(II) adsorption onto Fe3O4@Mg2Al-NO3 Layered Double Hydroxide in aqueous solution

The adsorption of Pb(II) and Cu(II) onto Fe3O4@Mg2Al-NO3 Layered Double Hydroxide (LDH) as a function of Fe3O4@Mg2Al-NO3 LDH concentration was studied. An adsorbent concentration effect ( Cs effect), namely adsorption isotherm declines as adsorbent concentration ( Cs) increases, was observed. The experimental data were fitted to the adsorption models including the classic Freundlich model, the metastable-equilibrium adsorption theory, the flocculation model, the power function model, and the surface component activity model. The results show that the Freundlich-type metastable-equilibrium adsorption equation, the power function model, and the Freundlich-surface component activity equation can adequately describe the Cs effect observed in the batch adsorption tests as all the correlation coefficients ( R2) of the nonlinear plots are higher than 0.96. In other words, their intrinsic parameters simulated from the experimental data are independent of Cs value. It is considered that the Freundlich-surface component activity equation is the best model to describe the Cs effect of the studied adsorption systems by Akaike Information Criterion evaluation criterion.

Removal and attenuation of sewage effluent combined tracer signals of phosphorus, caffeine and saccharin in soil

Contaminants in septic tank effluent (STE) are expected to be removed by the soil system before discharging to the environment. However, potential contaminants such as phosphorus (P), caffeine and artificial sweeteners do find their way to watercourses impacting aquatic eco systems. In this study, the attenuation of STE P, caffeine and saccharin were investigated in untreated soil and in soil with reduced microbial activity, in aqueous solutions and in the complex matrix of STE. Time series sorption and desorption experiments using batch equilibrium and a column experiment of STE P attenuation were conducted. The results revealed that the soil distribution coefficients (K_d) were: P 81.57 > caffeine 22.16 > saccharin 5.98 cm³/g, suggesting greater soil affinity to P adsorption. The data revealed that 80% of saccharin and 33% of caffeine attenuation was associated with microbial activities rather than adsorption processes. However, a complete removal of saccharin and caffeine did not occur during the equilibration period, suggesting their leaching potential. The dominant mechanism of P attenuation was adsorption (chemical and physical), yielding P retention of >73% and 35% for P in aqueous solution and in STE matrix, respectively, for batch equilibrium. The soil in the column acted as effluent P sink retaining 125 μg P/g soil of effluent P. The attenuation of P, caffeine and saccharin in the aqueous solution was greater than in STE, suggesting that the complex composition of STE reduced soil adsorption ability, and that other substances present in STE may be competing for soil binding sites. The data revealed that caffeine and P had similarities in the interaction with soils and thus caffeine may be considered as a STE tracer of anthropogenic source of P in receiving waters.

Evaluating the long-term performance of low-cost adsorbents using small-scale adsorption column experiments

This study investigated a novel method of predicting the long-term phosphorus removal performance of large-scale adsorption filters, using data derived from short-term, small-scale column experiments. The filter media investigated were low-cost adsorbents such as aluminum sulfate drinking water treatment residual, ferric sulfate drinking water treatment residual, and fine and coarse crushed concretes. Small-bore adsorption columns were loaded with synthetic wastewater, and treated column effluent volume was plotted against the mass of phosphorus adsorbed per unit mass of filter media. It was observed that the curve described by the data strongly resembled that of a standard adsorption isotherm created from batch adsorption data. Consequently, it was hypothesized that an equation following the form of the Freundlich isotherm would describe the relationship between filter loading and media saturation. Moreover, the relationship between filter loading and effluent concentration could also be derived from this equation. The proposed model was demonstrated to accurately predict the performance of large-scale adsorption filters over a period of up to three months with a very high degree of accuracy. Furthermore, the coefficients necessary to produce said model could be determined from just 24 h of small-scale experimental data.

Ion adsorption components in liquid/solid systems

Experiments on Zn2+ and Cd2+ adsorptions on vermiculite in aqueous solutions were conducted to investigate the widely observed adsorbent concentration effect on the traditionally defined adsorption isotherm in the adsorbate range 25-500 mg/L and adsorbent range 10-150 g/L. The results showed that the equilibrium ion adsorption density did not correspond to a unique equilibrium ion concentration in liquid phase. Three adsorbate/adsorbent ratios, the equilibrium adsorption density, the ratio of equilibrium adsorbate concentration in liquid phase to adsorbent concentration, and the ratio of initial adsorbate concentration to adsorbent concentration, were found to be related with unique values in the tested range. Based on the assumption that the equilibrium state of a liquid/solid adsorption system is determined by four mutually related components: adsorbate in liquid phase, adsorbate in solid phase, uncovered adsorption site and covered adsorption site, and that the equilibrium chemical potentials of these components should be equalized, a new model was presented for describing ion adsorption isotherm in liquid/solid systems. The proposed model fit well the experimental data obtained from the examined samples.

Proton buffering and metal leaching in sandy soils

Recent developments in acidification research focused on the leaching of metals from contaminated soil. In this paper the buffering of sandy soils upon acidification is studied in relation to the release of major (Al, Ca, Mg) and trace metals (Cu, Cd, Ni, Zn) from the soil reactive surface. The buffering process and the (de)sorption of metals are described with a mechanistic multisurface model, expressing the sorption onto different soil surfaces (organic matter, clay, Fe (hydr)oxides). The pH of sandy soil samples is predicted upon proton addition in combination with the behavior of major and trace metals. Acidification of contaminated sandy soil samples, with different pH levels and metal contents, is performed in a flow-through reactor by flushing the samples with acid solution. Acidification has taken place in successive steps of proton addition and followed by sampling. Prediction of pH upon acidification with a multisurface model gives satisfying results for all samples studied. The pH is modeled reasonably well between pH 6 and 4. Below pH 4 the predicted pH values are slightly too low, probably due to the buffering by Al-containing minerals (e.g., Al hydroxide), which are not included in the model. Desorption of major and trace metals upon pH decrease is, in general, predicted well, within a factor of 1-5 on a linear scale. Overall prediction of proton buffering in combination with desorption of metals in sandy soil samples, over a wide pH range and metal content, is done quite well for the studied metals with the multisurface model.

Reactive transport of 85Sr in a chernobyl sand column: static and dynamic experiments and modeling

The effects of nonlinear sorption and competition with major cations present in the soil solution on radioactive strontium transport in an eolian sand were examined. Three laboratory techniques were used to identify and quantify the chemical and hydrodynamic processes involved in strontium transport: batch experiments, stirred flow-through reactor experiments and saturated laboratory columns. The major goal was to compare the results obtained under static and dynamic conditions and to describe in a deterministic manner the predominant processes involved in radioactive strontium transport in such systems. Experiments under dynamic conditions, namely flow-through reactor and column experiments, were in very good agreement even though the solid/liquid ratio was very different. The experimental data obtained from the flow-through reactor study pointed to a nonlinear, instantaneous and reversible sorption process. Miscible displacement experiments were conducted to demonstrate the competition between stable and radioactive strontium and to quantify its effect on the 85Sr retardation factor. The results were modeled using the PHREEQC computer code. A suitable cation-exchange model was used to describe the solute/soil reaction. The model successfully described the results of the entire set of miscible displacement experiments using the same set of parameter values for the reaction calculations. The column study revealed that the stable Sr aqueous concentration was the most sensitive variable of the model, and that the initial state of the sand/solution system had also to be controlled to explain and describe the measured retardation factor of radioactive strontium. From these observations, propositions can be made to explain the discrepancies observed between some data obtained from static (batches) and dynamic (reactor and column) experiments. Desorbed antecedent species (stable Sr) are removed from the column or reactor in the flow system but continue to compete for sorption sites in the batch system. Batch experiments are simple and fast, and provide a very useful means of multiplying data. However, interpretation becomes difficult when different species compete for sorption sites in the soil/solution system. A combination of batches, flow-through reactor and column experiments, coupled with hydrogeochemical modeling, would seem to offer a very powerful tool for identifying and quantifying the predominant processes on a cubic decimeter scale (dm3) and for providing a range of radioactive strontium retardation factor as a function of the geochemistry of the soil/solution system.

Kinetic study on the sorption of dissolved natural organic matter onto different aquifer materials: the effects of hydrophobicity and functional groups

The subsurface sorption of Suwannee River fulvic acid (SRFA) and humic acid (SRHA) onto a synthetic aquifer material (iron-oxide-coated quartz) and two natural aquifer materials (Ringold sediment and Bemidji soils) was studied in both batch and column experiments. The hypothesis that hydrophobic effects followed by ligand exchange are the dominant mechanism contributing to the chemical sorption happening between dissolved natural organic matter (NOM) and the mineral surfaces is supported by observations of several phenomena: nonlinear isotherms, faster sorption rates versus slower desorption rates, phosphate competition, a solution pH increase during NOM sorption, and functional groups and aromaticity-related sorption. In addition, high-pressure size exclusion chromatography (HPSEC) and carboxylic acidity showed that lower molecular weight NOM components of SRHA are preferentially sorbed to iron oxide, a result in contrast to that for SRFA. Phosphate increased the desorption of sorbed NOM as well as soil organic matter. All of these trends support ligand exchange as the dominant reaction between NOM and the iron oxide surfaces; however, if the soil surface has been occupied by soil organic matter, then the sorption of NOM is more due to hydrophobic effect.

Sorption and mobility of dithiopyr in golf course greens rooting medium

Sorption and mobility of dithiopyr in golf course greens rooting medium (RM) were studied. The sorption increased from 20 to 27 degrees C at 24 h after treatment (HAT) and reached equilibrium in 48 HAT at 20 degrees C. The sorption isotherms had Freundlich values (KF) of 1122, 27.44 to 35.16, and 0.053 to 0.168 for peat moss, the RM, and quartz sand, respectively, and solid to aqueous phase partition coefficients (Kd) of 470 to 1706 L/kg, 14.61 to 84.4 L/kg, and 0.07 to 0.29 L/kg for peat moss, RM, and quartz sand, respectively. Generally, higher dithiopyr concentration in the aqueous solution and the reduced pH of the solution corresponded to the higher Kd values. The average values for dispersion (D, cm2/min), retardation coefficient (R), beta, and omega parameters for solute transport in the RM lysimeter; obtained from CXTFIT curve fitting of Br- breakthrough curves; were 0.95, 1.01, 1, and 93.89, respectively. After elutriation by 18 L of aqueous KNO3 (10 mM), greater than 90% of the added dithiopyr remained in the top 10 cm of the RM lysimeter and no detectable dithiopyr was present at depths beyond 35 cm. The lysimeter effluent contained dithiopyr at concentrations less than 3.5 microg/L. The R value obtained from CXFIT curve fitting is 38.5. Results from both sorption and mobility experiments indicated that dithiopyr is quite immobile in golf course greens RM and has minimal potential for movement into surface water drainage or ground water.

Reaction-based model describing competitive sorption and transport of Cd, Zn, and Ni in an acidic soil

Predicting the mobility of heavy metals in soils requires models that accurately describe metal adsorption in the presence of competing cations. They should also be easily adjustable to specific soil materials and applicable in reactive transport codes. In this study, Cd adsorption to an acidic soil material was investigated over a wide concentration range (10(-8) to 10(-2) M CdCl2) in the presence of different background electrolytes (10(-4) to 10(-2) M CaCl2 or MgCl2 or 0.05 to 0.5 M NaCl). The adsorption experiments were conducted at pH values between 4.6 and 6.5 A reaction-based sorption model was developed using a combination of nonspecific cation exchange reactions and competitive sorption reactions to sites with high affinity for heavy metals. This combined cation exchange/specific sorption (CESS) model accurately described the entire Cd sorption data set. Coupled to a solute transport code, the model accurately predicted Cd breakthrough curves obtained in column transport experiments. The model was further extended to describe competitive sorption and transport of Cd, Zn, and Ni. At pH 4.6, both Zn and Ni exhibited similar sorption and transport behavior as observed for Cd. In all transport experiments conducted under acidic conditions, heavy metal adsorption was shown to be reversible and kinetic effects were negligible within time periods ranging from hours up to four weeks.