Physical origin for the nonlinear sorption of very hydrophobic organic chemicals in a membrane-like polymer film

Bioconcentration factor (BCF) is often assumed to be linearly associated with the octanol-water partition coefficient K(ow) for hydrophobic organic chemicals (HOCs). However, a large amount of data has suggested that the correlation between the logBCF and logK(ow) is curvilinear for HOCs. Similar curvilinear relationship has also been noticed for sorption of HOCs into poly(dimethyl)siloxane (PDMS), a polymer with cross-linked interior structures. So far no satisfactory explanation has been given to account for the deviation. In this study, we acquired additional experimental data to show that the curvilinear relationship between the log-based PDMS-coated fiber-water partition coefficient (logK(f)) and logK(ow) for polychlorinated biphenyls (PCBs) was indeed a reflection of the sorption process occurring in PDMS film other than experimental defects. The physical origin of the nonlinearity was pinpointed based on the theory of phase partitioning for HOCs. The linear relationship is observed if the solute molecule is considerably smaller than the size of a monomer unit of PDMS in that the Gibbs free energy required for cavity formation in PDMS is comparable to that in octanol. Higher free energy of cavity formation is needed to create sufficient free volume if the PCB molecular size is comparable to or larger than the monomer unit of PDMS. On the other hand, the free energy of cavity formation in octanol remains almost constant when this occurs, resulting in the observed curvilinear relationship. The proposed model adequately explains the observed data, as well as sheds lights into the physical origin of the steric interactions of large molecular size solute with the PDMS polymer network.     

Distribution of polychlorinated dibenzo-p-dioxins and dibenzofurans in suspended sediments, dissolved phase and bottom sediment in the Houston Ship Channel

Spatial distributions of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) in the water column and bottom sediments of the Houston Ship Channel in Texas were measured three times over a 1-year period. Total TEQ concentrations in water ranged from 0.01 to 0.25 pg/l for the dissolved phase and from 0.09 to 2.91 pg/l for the suspended phase, while TEQ concentrations in bottom sediments varied from 0.9 to 139.8 ng/kg dry wt. The dissolved concentrations were lower than their respective suspended concentrations, with average dissolved/suspended ratios between 0.11 and 0.59 for individual congeners. More than 89% of the total concentration of 2378-substituted PCDD/PCDFs was attributable to OCDD but 2378-TCDD was the major contributor to total TEQ for the three sampled media. Average logs of organic carbon-normalized suspended sediment-dissolved partitioning coefficients (logK(oc)(obs)) varied between 4.92 and 8.59 l/kg-oc; while in the bottom sediment-dissolved interface, logK(oc)(obs) values ranged from 5.48 to 8.48 l/kg-oc. Observed logK(oc)values varied within a factor of 0.64-1.26 from equilibrium logK(oc) values, suggesting fluxes of PCDD/PCDFs across the interfaces. It was found that in the HSC, on average, the tendency of a compound to move from the particulate phase to the dissolved phase decreases with increasing K(ow).     

The partitioning of alkylphenolic surfactants and polybrominated diphenyl ether flame retardants in activated sludge batch tests

Polybrominated diphenyl ethers and nonylphenol polyethoxylates have been reported to be estrogenic and may enter the aquatic environment through the discharge of treated sewage effluent. Therefore, their fate during wastewater treatment processes is an important factor in determining their environmental impact. Batch tests with activated sludge from a Husmann apparatus were used to determine the effects of physico-chemical properties and sludge characteristics on the partitioning of polybrominated diphenyl ether flame retardants and nonylphenol polyethoxylate surfactants during biological wastewater treatment. Hydrophobic compounds, those with high logK(ow) values, were sorbed more rapidly and to a greater extent to the solid phase than more soluble compounds. For these hydrophobic compounds sorption may become an increasingly important removal mechanism as sludge age and therefore solids content increase. The initial rate of partitioning was greatest for the most hydrophobic compounds but all rates diminished with time as a result of progressive saturation of sorbent binding sites, a reduction of sorbate availability and as a consequence of the system reaching equilibrium. The sorption of polybrominated diphenyl ethers fit Freundlich adsorption isotherms demonstrating generally increasing adsorption capacity and efficiency with increasing hydrophobic nature. A correlation between increasing logK(ow) and increasing organic matter content was also observed for both polybrominated diphenyl ethers and nonylphenol polyethoxylates indicating the organic content of mixed liquor will also be influential in removing compounds during wastewater treatment.     

Partitioning of chloroaromatic compounds between the aqueous phase and dissolved and particulate soil organic matter at chlorophenol contaminated sites

The retention and mobility of hydrophobic organic contaminants (HOCs) in soil is mainly determined by hydrophobic partitioning to dissolved and particulate organic matter (DOM and POM, respectively). The aqueous phase, DOM, and POM fractions were extracted and separated from soils at three sites contaminated with technical chlorophenol formulations. Concentrations of chlorophenols (CP), polychlorinated phenoxyphenols (PCPP), polychlorinated diphenyl ethers (PCDE) and polychlorinated dibenzo-p-dioxins and furans (PCDD/F) were determined. The partitioning to POM, in relation to DOM, increased in all three soils with increasing hydrophobicity in the order CP < PCPP ~ PCDE ~ PCDF < PCDD. Differences in partitioning to DOM (logK(DOC)) and POM (logK(POC)) could not be explained by differences in gross organic C chemistry. Black carbon did not contribute significantly to the sorption of PCDDs, whereas >70% wood fibre in one soil resulted in a decrease of logK(POC) of 0.5 units for CPs and PCDDs. We conclude that logK(OC) for both DOM and POM need to be explicitly determined when the retention and mobility of HOCs is described and modelled in soils.     

Transfer functions for solid-solution partitioning of cadmium for Australian soils

To assess transport and ecotoxicological risks of metals, such as cadmium (Cd) in soils, models are needed for partitioning and speciation. We derived regression-based “partition-relations” based on adsorption and desorption experiments for main Australian soil types. First, batch adsorption experiments were carried out over a realistic range of dissolved Cd concentrations in agricultural soils in Australia. Results showed linear sorption relationships, implying the adequacy of using Kd values to describe partitioning. Desorption measurements were then carried out to assess in-situ Kd values and relate these to soil properties The best transfer functions for solid-solution partitioning were found for Kd values relating total dissolved Cd concentration to total soil Cd concentrations, accounting for the variation in pH, SOM contents and DOC concentrations. Model predictions compared well with measurements of an independent data set, but there was a tendency to underestimate dissolved Cd concentrations of highly polluted soils.     

Reverse quantitative structure-activity relationship for modelling the sorption of esfenvalerate to dissolved organic matter. A multivariate approach

The sorption of the pyrethroid, esfenvalerate, to the dissolved and/or dispersed fraction of eight different natural humic compounds has been investigated. The dissolved organic matters (DOMs) included in this study originate from ground water, soil pore water, and surface waters. Sorption was modelled at DOM concentration levels where equilibrium partitioning of esfenvalerate between DOM and the aqueous bulk phase prevails. The inherent characteristics of the eight different humic materials, quantified in the preceding paper by Thomsen et al. (2002, this issue (PII: S0045-6535(02)00335-1)), have been used as explanatory variables for modelling this equilibrium partitioning. Using a reverse QSAR approach based on by projection-into-latent-structure regression (PLS-R) inherent sorbent properties determining for the sorption affinity of esfenvalerate to DOM were analysed. For all humic substances a decrease in the DOM-normalised equilibrium-partitioning coefficient, K_DOM, with increasing concentration of DOM was observed. Significant variations in K_DOM values, as function of the inherent characteristics of the individual humic substances, were found at DOM concentrations of 75 and 100 ppm, respectively. The latter is a strong indication of variations in sorption mechanisms of esfenvalerate to DOM of varying inherent properties. Groupings in the principal property space quantifying DOMs may indicate that separate models are needed for quantifying the equilibrium partitioning to different classes of DOM.     

Soot sorption of non-ortho and ortho substituted PCBs

Field-observations of distribution coefficients well above expectations from bulk organic-matter partitioning for several chlorinated aromatic compound classes have lead to the hypothesis that enhanced affinity to soot may not be limited to polycyclic aromatic hydrocarbons but may extend as a significant process for a wider range of hydrophobic organic compounds. This suggestion was here tested in soot-column sorption experiments with a series of ortho- and non-ortho substituted polychlorinated biphenyls (PCBs), using diesel particulate matter (NIST standard reference material SRM-1650) as model soot sorbent. For congeners of similar hydrophobicity, considerably higher affinities toward the soot sorbent were observed for the non-ortho substituted PCBs. Mono- to tetra-ortho substituted PCBs exhibited log-based soot-water distribution coefficients (K(sc)) from 5.25 to 5.51 l/kg(sc) at solute concentrations corresponding to 1-13 microg/l. In contrast, biphenyl, mono- and dichloro- non-ortho substituted PCBs yielded logK(sc) values between 5.09 and 6.35 l/kg(sc). These results are 20-50, and 75-110 times higher, respectively, than the corresponding K(ow)-predicted K(oc) numbers. This strong interaction with soot, particularly of non-ortho substituted PCBs, may fundamentally affect their environmental distribution and bioavailable exposure.     

Study on fluoride emission from soils at high temperature related to brick-making process

Characteristics of fluoride emission from 12 soils at temperatures of 400-1,100 degrees C related to the brick-making process were studied. The results obtained in this study indicate that fluoride emission as gaseous HF and SiF4 was related to the firing temperature, soil total fluoride content, soil composition and calcium compounds added to soils. Soils began to release fluoride at temperatures between 500 and 700 degrees C. Marked increases of the average fluoride mission rate from 57.2% to 85.4% of soil total fluoride were noticed as the heating temperature was increased from 700 to 1,100 degrees C. It was found that the major proportion (over 50%) of the soil total fluoride was emitted from soils at approximate 800 degrees C. The amount of fluoride released into the atmosphere when heated depended on the total fluoride contents in the soils. Correlation analysis showed that the soil composition, such as cation exchange capacity, exchangeable calcium and CaCO3, had some influence on fluoride emission below 900 degrees C, but had no influence at temperatures above 900 degrees C. Addition of four calcium compounds (CaO, CaCO3, Ca(OH)2, and CaSO4) at 1.5% by weight raised the temperature at which fluoride began to be released to 700 degrees C. The greatest decrease in fluoride emission among the four calcium compound treatments was found with CaCO3.     

Speciation of zinc in contaminated soils

The chemical speciation of zinc in soil solutions is critical to the understanding of its bioavailability and potential toxic effects. We studied the speciation of Zn in soil solution extracts from 66 contaminated soils representative of a wide range of field conditions in both North America and Europe. Within this dataset, we evaluated the links among the dissolved concentrations of zinc and the speciation of Zn(2+), soil solution pH, total soil Zn, dissolved organic matter (DOM), soil organic matter (SOM) and the concentrations of different inorganic anions. The solid-liquid partitioning coefficient (K(d)) for Zn ranged from 17 to 13,100Lkg(-1) soil. The fraction of dissolved Zn bound to DOM varied from 60% to 98% and the soil solution free Zn(2+) varied from 40% to 60% of the labile Zn. Multiple regression equations to predict free Zn(2+), dissolved Zn and the solid-liquid partitioning of Zn are given for potential use in environmental fate modeling and risk assessment. The multiple regressions also highlight some of the most important soil properties controlling the solubility and chemical speciation of zinc in contaminated soils.     

Predicting organic contaminant concentrations in sediment porewater using solid-phase microextraction

Because of its cost and time saving features, solid-phase microextraction (SPME) is a leading candidate as a biomimic technique in assessing the bioavailable fraction of hydrophobic organic contaminants (HOCs) in sediment porewater. However, no predictive modeling framework in which to systematically address the effect of key parameters on SPME performance for this application exists. In this study, we derived two governing equations to predict (1) the minimum sediment volume (V(s)min) required to achieve non-depletive conditions, and (2) dissolved phase HOC porewater concentrations (C(pw)) as functions of HOC- and sediment specific characteristics in a conceptual three compartment system. The resulting model predicted that V(s)min was independent of HOC concentrations both in sediment and porewater, but did vary with hydrophobicity (characterized by logK(ow)), the fraction of sediment porewater (f(pw)), and the volume (V(f)) of the SPME sorbent phase. Moreover, the effects of these parameters were minimized (i.e., V(s)min reached plateaus) as logK(ow) approached 4-5. Model predictions of C(pw), a surrogate for SPME-based detection limits in porewater, decreased with increasing sediment volume (V(s)) at low V(s) values, but rapidly leveled off as V(s) increased. A third result suggested that the sediment HOC concentration required for SPME is completely independent of K(ow). These results suggest that relatively small sediment volumes participate in exchange equilibria among sediment, porewater and the SPME fiber, and that large sediment HOC reservoirs are not needed to improve the detection sensitivity of SPME-based porewater samplers. The ultimate utility of this modeling framework will be to assist future experimental designs and help predict in situ bioavailability of sediment-associated HOCs.     

Uptake rates of alkylphenols, PAHs and carbazoles in semipermeable membrane devices (SPMDs) and polar organic chemical integrative samplers (POCIS)

Passive sampling devices provide a useful contribution to the monitoring of contaminants in the aquatic environment. However, calibration data needed for the calculation of water concentrations from sampler accumulations are restricted to a limited number of compound classes. Thus uptake of a range of alkylated phenols (AP), polycyclic aromatic hydrocarbons (PAH) and carbazoles was determined for semipermeable membrane devices (SPMDs) and polar organic chemical integrative samplers (POCIS) using a flow through exposure system. Sampling rates ranged from 0.02 to 0.26 l d(-1) for POCIS and 0.02 to 13.83 l d(-1) for SPMDs. Observed SPMD uptake was also compared to that predicted by an empirical model including the use of performance reference compounds (PRCs). Predicted sampling rates did not differ by more than a factor of 1.3 from experimental values for PAH, providing further evidence that the PRC approach can be successfully used to determine in situ sampling rates for these compounds. Experimental sampling rates for AP in SPMDs were, however, much lower than predicted. This discrepancy was too large to be explained by small uncertainties in the calibration system or in the calculations. Based on these data we conclude that while hydrophobic AP are accumulated by SPMDs their partitioning cannot be predicted from their logK(ow) using current methods. Due to this lower than expected uptake, sampling rates were only higher in SPMDs than POCIS in the range of logK(ow)>5.0. Simultaneous deployment of both sampler types allows the study of compounds with a broad range of physicochemical properties.     

The effects of particle size, organic matter content, crop residues and dissolved organic matter on the sorption kinetics of atrazine and isoproturon by clay soil

Reliable predictions of the fate and behaviour of pesticides in soils is dependent on the use of accurate ‘equilibrium’ sorption constants and/or rate coefficients. However, the sensitivity of these parameters to changes in the physicochemical characteristics of soil solids and interstitial solutions remains poorly understood. Here, we investigate the effects of soil organic matter content, particle size distribution, dissolved organic matter and the presence of crop residues (wheat straw and ash) on the sorption of the herbicides atrazine and isoproturon by a clay soil. Sorption K_d's derived from batch ‘equilibrium’ studies for both atrazine and isoproturon by <2 mm clay soil were approximately 3.5 L/kg. The similarity of K_oc's for isoproturon sorption by the <2 mm clay soil and <2 mm clay soil oxidised with hydrogen peroxide suggested that the sorption of this herbicide was strongly influenced by soil organic matter. By contrast, K_oc's for atrazine sorption by oxidised soil were three times greater than those for <2 mm soil, indicating that the soil mineral components might have affected sorption of this herbicide. No significant differences between the sorption of either herbicide by <2 mm clay soil and (i) <250 μm clay soil, (ii) clay soil mixed with wheat straw or ash at ratios similar to those observed under field conditions, (iii) <2 mm clay soil in the presence of dissolved organic matter as opposed to organic free water, were observed.     

Sorption–desorption of imidacloprid onto a lacustrine Egyptian soil and its clay and humic acid fractions

Sorption–desorption of the insecticide imidacloprid 1-[(6-chloro-3-pyridinyl)-methyl]-N-nitro-2-imidazolidinimine onto a lacustrine sandy clay loam Egyptian soil and its clay and humic acid (HA) fractions was investigated in 24-h batch equilibrium experiments. Imidacloprid (IMDA) sorption–desorption isotherms onto the three sorbents were found to belong to a non-linear L-type and were best described by the Freundlich model. The value of the IMDA adsorption distribution coefficient, Kd_ads, varied according to its initial concentration and was ranged 40–84 for HA, 14–58 for clay and 1.85–4.15 for bulk soil. Freundlich sorption coefficient, Kf_ads, values were 63.0, 39.7 and 4.0 for HA, clay and bulk soil, respectively. The normalized soil Koc value for imidacloprid sorption was ～800 indicating its slight mobility in soils. Nonlinear sorption isotherms were indicated by 1/n_ads values <1 for all sorbents. Values of the hysteresis index (H) were <1, indicating the irreversibility of imidacloprid sorption process with all tested sorbents. Gibbs free energy (ΔG) values indicated a spontaneous and physicosorption process for IMDA and a more favorable sorption to HA than clay and soil. In conclusion, although the humic acid fraction showed the highest capacity and affinity for imidacloprid sorption, the clay fraction contributed to approximately 95% of soil-sorbed insecticide. Clay and humic acid fractions were found to be the major two factors controlling IMDA sorption in soils. The slight mobility of IMDA in soils and the hysteresis phenomenon associated with the irreversibility of its sorption onto, mainly, clay and organic matter of soils make its leachability unlikely to occur.     

Partitioning of organochlorine pesticides from water to polyethylene passive samplers

The mass transfer rates and equilibrium partitioning behaviour of 14 diverse organochlorine pesticides (OCP) between water and polyethylene (PE) passive samplers, cut from custom made PE sheets and commercial polyethylene plastic bags, were quantified. Overall mass transfer coefficients, k_O, estimated PE membrane diffusion coefficients, D_PE, and PE-water partitioning coefficients, K_PE-water, are reported. In addition, the partitioning of three polycyclic aromatic hydrocarbons (PAHs) from water to PE is quantified and compared with literature values. K_PE-water values agreed mostly within a factor of two for both passive samplers and also with literature values for the reference PAHs. As PE is expected to exhibit similar sorption behaviour to long-chain alkanes, PE-water partitioning coefficients were compared to hexadecane-water partitioning coefficients estimated with the SPARC online calculator, COSMOtherm and a polyparameter linear free energy relationship based on the Abraham approach. The best correlation for all compounds tested was with COSMOtherm estimated hexadecane-water partitioning coefficients.     

Organic compounds adsorption onto activated carbon: the effect of association between dissolved humic substances and pesticides

The quantitative determination of pesticide binding to dissolved humic substances is relevant to both water treatment operation using activated carbon adsorption process and the application of transport models that predict the environmental distribution patterns of a given hydrophobic contaminant. In this study and in a first set of experiments, the extent of binding between (i) three pesticides of environmental concern, aldicarb, lindane and pentachlorophenol, and (ii) dissolved commercial humic acid and soil extracted fulvic acid, was determined using dialysis experiments and water solubility enhancement tests. In a second set of experiments, the influence of dissolved humic substances or pesticide on the retention of the other co-adsorbate onto activated carbon was investigated in binary systems. It was found that association was negligible for aldicarb and that the pesticide sorption onto activated carbon was not affected by humic acid (8.5 mg liter(-1) DOC). The association constants K for lindane and pentachlorophenol were identical in the presence of fulvic acid (logK=4.1) but lower than that observed with humic acid. In the presence of humic acid, binding affinity for pentachlorophenol (logK=4.6) was higher than the one observed for lindane (logK=4.4), despite its much higher water solubility. This observation suggests that the aromatic character of the pentachlorophenol molecule contributes to association interactions with humic acid. From co-adsorption experiments onto activated carbon it was found that fulvic acid (7.7 mg litre(-1) DOC) slightly enhances sorption kinetics of pentachlorophenol. Lindane (1 mg litre(-1)) does not affect sorption kinetics for fulvic acid but markedly enhances both the sorption kinetics and adsorptive capacity for humic acid. Activated carbon retention of dissolved humic substances or pesticide appears to be enhanced by the association potential that exists between these co-adsorbates in some binary systems.     

Comparison of nonideal sorption formulations in modeling the transport of phthalate esters through packed soil columns

Sorption of dimethyl phthalate (DMP), diethyl phthalate (DEP) and dipropyl phthalate (DPP) to two soil materials that vary in organic matter content was investigated using miscible displacement experiments under saturated flow conditions. Generated breakthrough curves (BTCs) were inversely simulated using linear, equilibrium sorption (LE), nonlinear, equilibrium sorption (NL), linear, first-order nonequilibrium sorption (LFO), linear, radial diffusion (LRD), and nonlinear, first-order nonequilibrium sorption (NFO) models. The Akaike information criterion was utilized to determine the preferred model. The LE model could not adequately describe phthalate ester (PE) BTCs in higher organic matter soil or for more hydrophobic PEs. The LFO and LRD models adequately described the BTCs but a slight improvement in curve-fitting was gained in some cases when the NFO model was used. However, none of the models could properly describe the desorptive tail of DPP for the high organic matter soil. Transport of DPP through this soil was adequately predicted when degradation or sorption hysteresis was considered. Using the optimized parameter values along with values reported by others it was shown that the organic carbon distribution coefficient (K(oc)) of PEs correlates well with the octanol/water partition coefficient (K(ow)). Also, a strong relationship was found between the first-order sorption rate coefficient normalized to injection pulse size and compound residence time. A similar trend of timescale dependence was found for the rate parameter in the radial diffusion model. Results also revealed that the fraction of instantaneous sorption sites is dependent on K(ow) and appears to decrease with the increase in the sorption rate parameter.     

Nonlinear binding of phenanthrene to the extracted fulvic acid fraction in soil in comparison with other organic matter fractions and to the whole soil sample

Fractions of soil organic matter in a natural soil were extracted and sorption (or binding) characteristics of phenanthrene on each fraction and to the whole sample were investigated. The organic carbon normalized single point sorption (or binding) coefficient followed lipid > humin (HM) > humic acid (HA) > fulvic acid (FA) > whole soil sample, while the nonlinear exponent exhibited lipid > FA > HA > whole soil sample > HM. FA showed nonlinear binding of phenanthrene as it often does with other fractions. HM and HA contributed the majority of organic carbon in the soil. The calculated sorption coefficients of the whole soil were about two times greater than the measured values at different equilibrium phenanthrene concentrations. As for phenanthrene, the sorption capacity and nonlinearity of the physically mixed HA-HM mixtures were stronger as compared to the chemically reconstituted HA-HM composite. This was attributed to (besides the conditioning effect of the organic solvents) interactions between HA and HM and acid-base additions during fractionation.     

Mobility of chloroaromatic compounds in soil: case studies of Swedish chlorophenol-contaminated sawmill sites

This paper summarizes recent studies on the environmental fate of chloroaromatic compounds in chlorophenol (CP)-contaminated soil and groundwater at Swedish sawmill sites. Relative proportions of CPs, polychlorinated phenoxy phenols (PCPPs), polychlorinated diphenyl ethers (PCDEs), polychlorinated dibenzo-p-dioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs) were determined in preservatives, particulate organic matter (POM), dissolved organic matter (DOM), groundwater, and particles filtered from groundwater. All compound classes were found in the different compartments. The fraction of PCPPs, PCDEs, PCDDs, and PCDFs had increased in the soil samples relative to the proportions in the preservatives. This increase showed correlation with the hydrophobicity, that is, PCDDs had the largest increase. Similar correlation was found between hydrophobicity and the importance of partitioning to POM over DOM. The more water soluble compound group, CP, was found equally distributed between POM and DOM. For PCPPs, PCDEs, PCDDs, and PCDFs, the relative partitioning to POM increased with increased hydrophobicity. Despite the relative partitioning towards POM, compared with DOM, cotransport with DOM and suspended colloidal fractions was found to substantially increase the transport of these compounds in the groundwater samples.