A noval method of predicting soil organic carbon content normalized adsorption coefficients (Koc) of polybrominated biphenyls (PBBs)

By a soil column liquid chromatograph (SCLC) method, the soil organic carbon content normalized adsorption coefficients (K(oc)) of six polybrominated biphenyls (PBB15, PBB26, PBB31, PBB49, PBB103 and PBB153) are determined. Based on the similarity between the molecular structures of PBBs and PCBs, a simple linear predictive model has been developed with the correlation coefficient R=0.9812 and standard error SE=0.19. The logK(oc) values of any PBB congeners can be predicted by using the logK(oc) values of the corresponding PCBs according to this model. Using the published data for logK(oc) values of PCB congeners, logK(oc) values of all 209 PBB congeners have been for the first time predicted. Compared with the data obtained from the experiment, the results of prediction are very accurate.     

Can we predict uranium bioavailability based on soil parameters? Part 1: effect of soil parameters on soil solution uranium concentration

Present study aims to quantify the influence of soil parameters on soil solution uranium concentration for (238)U spiked soils. Eighteen soils collected under pasture were selected such that they covered a wide range for those parameters hypothesised as being potentially important in determining U sorption. Maximum soil solution uranium concentrations were observed at alkaline pH, high inorganic carbon content and low cation exchange capacity, organic matter content, clay content, amorphous Fe and phosphate levels. Except for the significant correlation between the solid-liquid distribution coefficients (K(d), L kg(-1)) and the organic matter content (R(2)=0.70) and amorphous Fe content (R(2)=0.63), there was no single soil parameter significantly explaining the soil solution uranium concentration (which varied 100-fold). Above pH=6, log(K(d)) was linearly related with pH [log(K(d))=-1.18 pH+10.8, R(2)=0.65]. Multiple linear regression analysis did result in improved predictions of the soil solution uranium concentration but the model was complex.     

Effect of clays, metal oxides, and organic matter on rhamnolipid biosurfactant sorption by soil

Rhamnolipids produced by Pseudomonas aeruginosa have been proposed as soil washing agents for enhanced removal of metal and organic contaminants from soil. A potential limitation for the application of rhamnolipids is sorption by soil matrix components. The objective of this study is to empirically determine the contribution of representative soil constituents (clays, metal oxides, and organic matter) to sorption of the rhamnolipid form most efficient at metal complexation (monorhamnolipid). Sorption studies show that monorhamnolipid (R1) sorption is concentration dependent. At low R1 concentrations that are relevant for enhancing organic contaminant biodegradation, R1 sorption followed the order: hematite (Fe(2)O(3))>kaolinite>MnO(2) approximately illite approximately Ca-montmorillonite>gibbsite (Al(OH)(3))>humic acid-coated silica. At high R1 concentrations, relevant for use in complexation/removal of metals or organics, R1 sorption followed the order: illite>humic acid-coated silica>Ca-montmorillonite>hematite>MnO(2)>gibbsite approximately kaolinite. These results allowed prediction of R1 sorption by a series of six soils. Finally, a comparison of R1 and R2 (dirhamnolipid) shows that the R1 form sorbs more strongly alone than when in a mixture of both the R1 and R2 forms. The information presented can be used to estimate, on an individual soil basis, the extent of rhamnolipid sorption. This is important for determining: (1) whether rhamnolipid addition is a feasible remediation option and (2) the amount of rhamnolipid required to efficiently remove the contaminant.     

Sorption of phenanthrene by soils contaminated with heavy metals

The fate of polycyclic aromatic hydrocarbons (PAHs) in soils with co-contaminants of heavy metals has yet to be elucidated. This study examined sorption of phenanthrene as a representative of PAHs by three soils contaminated with Pb, Zn or Cu. Phenanthrene sorption was clearly higher after the addition of heavy metals. The distribution coefficient (K(d)) and the organic carbon-normalized distribution coefficient (K(oc)) for phenanthrene sorption by soils spiked with Pb, Zn or Cu (0-1000 mg kg(-1)) were approximately 24% larger than those by unspiked ones, and the higher contents of heavy metals added into soils resulted in the larger K(d) and K(oc) values. The enhanced sorption of phenanthrene in the case of heavy metal-contaminated soils could be ascribed to the decreased dissolved organic matter (DOM) in solution and increased soil organic matter (SOM) as a consequence of DOM sorption onto soil solids. Concentrations of DOM in equilibrium solution for phenanthrene sorption were lower in the case of the heavy metal-spiked than unspiked soils. However, the decreased DOM in solution contributed little to the enhanced sorption of phenanthrene in the presence of metals. On the other hand, the sorbed DOM on soil solids after the addition of heavy metals in soils was found to be much more reactive and have far stronger capacity of phenanthrene uptake than the inherent SOM. The distribution coefficients of phenanthrene between water and the sorbed DOM on soil solids (K(ph/soc)) were about 2-3 magnitude larger than K(d) between water and inherent SOM, which may be the dominant mechanism of the enhanced sorption of phenanthrene by soils with the addition of heavy metals.     

Reliable QSAR for estimating Koc for persistent organic pollutants: correlation with molecular connectivity indices

Several recent studies have shown that n-octanol/water partition coefficients may not be a good predictor for estimating soil sorption coefficients of persistent organic pollutants (POPs), defined here as chemicals with log Kow greater than 5. Thus, an alternative QSAR model was developed that seems to provide reliable estimates for the soil sorption coefficients of persistent organic pollutants. This model is based on a set of calculated molecular connectivity indices and evaluated soil sorption data for 18 POPs. The chemical's size and shape, quantified by 1chi, 3chiC and 4chiC(v) indices, have a dominant effect on the soil sorption process of POPs. The developed QSAR model was rationalized in terms of potential hydrophobic interactions between persistent organic pollutants and soil organic matrix. Its high predictive power has been verified by an extensive internal and external validation procedure.     

Acetamiprid, carbendazim, diuron and thiamethoxam sorption in two Brazilian tropical soils

Sorption of acetamiprid ((E)-N1-[(6-chloro-3-pyridyl)methyl]-N2-cyano-N1-methylacetamidine), carbendazim (methyl benzimidazol-2-ylcarbamate), diuron (N-(3,4-dichlorophenyl)-N, N-dimethyl urea) and thiamethoxam (3-(2-chloro-thiazol-5-ylmethyl)-5-methyl-[1,3,5]oxadiazinan-4-ylidene-N-nitroamine) was evaluated in two Brazilian tropical soils, Oxisol and Entisol, from Primavera do Leste region, Mato Grosso State, Brazil. To describe the sorption process, batch experiments were carried out. Linear and Freundlich isotherm models were used to calculate the K(d) and K(f) coefficients from experimental data. The K(d) values were utilized to calculate the partition coefficient normalized to soil organic carbon (K(oc)). For the pesticides acetamiprid, carbendazim, diuron and thiamenthoxan the K(oc) (mL g(- 1)) values ranged in both soils from 98 - 3235, 1024 - 2644, 145 - 2631 and 104 - 2877, respectively. From the studied pesticides, only carbendazim presented correlation (r(2) = 0.82 and p < 0.01) with soil organic carbon (OC) content. Acetamiprid and thiamethoxam showed low sorption coefficients, representing a high risk of surface and ground water contamination.     

Washing of marine coastal sand in a batch reactor: sorption and desorption of BTEX

This study addresses the issues related to decontamination of marine beach sand accidentally contaminated by petroleum products. Sorption and desorption of BTEX (i.e., benzene, toluene, ethylbenzene, and xylene) onto the sand from Uran Beach, located near the city of Mumbai, India, were studied, and isotherms were determined using the bottle point method to estimate sorption coefficients. Alternatively, QSARs (i.e., quantitative structure activity relationships) were developed and used to estimate the sorption coefficients. Experiments for kinetics of volatilization as well as for kinetics of sorption and desorption in the presence of volatilization were conducted in a fabricated laboratory batch reactor. A mathematical model describing the fate of volatile hydrophobic organic pollutants like BTEX (via sorption and desorption in presence of volatilization) in a batch sediment-washing reactor was proposed. The experimental kinetic data were compared with the values predicted using the proposed models for sorption and desorption, and the optimum values of overall mass transfer coefficients for sorption (K(s)a(s)) and desorption (K(d)a(d)) were estimated. This was achieved by minimization of errors while using the sorption coefficients (Kp) obtained from either laboratory isotherm studies or the QSARs developed in the present study. Independent experimental data were also collected and used for calibration of the model for volatilization, and the values of the overall mass transfer coefficient for volatilization (K(g)a(g)) were estimated for BTEX. In these exercises of minimization of errors, comparable cumulative errors were obtained from the use of Kp values derived from experimental isotherms and QSARs.     

Effect of the cationic composition of sorption solution on the quantification of sorption-desorption parameters of heavy metals in soils

We obtained the sorption isotherms of Cd, Cu, Pb and Zn in clay, clay saline and organic soils. The distribution coefficients (K(d)) were determined in 0.02 eq l(-1) CaCl(2) and in a solution that simulated the soil solution cationic composition. The K(d) values greatly varied with the composition of the sorption solution and the initial metal concentration. The sorption experiments were complemented with the quantification of the extractable metal, to estimate the reversibility of metal sorption. The extraction yields depended on the metal-soil combination, and the initial metal concentration, showing no correlation with previous K(d) values. The effect of the solution composition in mobility predictions was estimated through a Retention Factor, defined as the ratio of the K(d) versus the extraction yield. Results showed that risk was over- or underestimated using the CaCl(2) medium in soils with a markedly different soil solution composition.     

Sorption of herbicides in relation to soil variability and landscape position

Using the soil-water sorption partitioning coefficient (Kd), this study quantified the spatial variation of 2,4-D sorption by soil in an undulating-to-hummocky terrain landscape near Minnedosa, MB, Canada. Herbicide sorption was most strongly related to soil organic matter content and slope position, with greatest sorption occurring in lower landscape positions with greater soil organic matter content. The relation between sorption and slope position was more pronounced under conventional tillage (CT) than under long-term zero-tillage (ZT). Using multivariate regression and three independent variables (soil organic matter content, soil clay content and soil pH), the prediction of herbicide sorption by soil was very good for CT (R2 = 0.89) and adequately for ZT (R2 = 0.53).     

Estimating octanol-air partition coefficients with octanol-water partition coefficients and Henry's law constants

The octanol-air partition coefficient (K(OA)) is useful for predicting the partitioning behavior of organic compounds between air and environmental matrices such as soil, vegetation, and aerosol particles. At present, experimentally determined K(OA) values are available for only several hundred compounds. Therefore, the ability to estimate K(OA) is necessary for screening level evaluation of most chemicals. Although it is possible to estimate K(OA) from the octanol-water partition coefficient (K(OW)) and Henry's law constant (HLC), various concerns have been raised in regard to the usability of this estimation methodology. This work examines the accuracy and usability of K(OW) and HLC in application to a comprehensive database set of K(OA) values for screening level environmental assessment. Results indicate that K(OW) and HLC can be used to accurately predict K(OA) even when estimated K(OW) and HLC values are used. For an experimental dataset of 310log K(OA) values for different compounds, the K(OW)-HLC method was statistically accurate as follows: correlation coefficient (r2): 0.972, standard deviation: 0.526, absolute mean error: 0.358 using predominantly experimental K(OW) and HLC values. When K(OW) and HLC values were estimated (using the KOWWIN and HENRYWIN programs), the statistical accuracy was: correlation coefficient (r2): 0.957, standard deviation: 0.668, absolute mean error: 0.479.     

Adsorption and desorption processes of MCPA in Polish mineral soils

Studies on the adsorption and desorption of MCPA (4-chloro-2-methylophenoxyacetic acid) were performed in soil horizons of three representative Polish agricultural soils. The Hyperdystric Arenosol, the Haplic Luvisol and the Hypereutric Cambisol were investigated in laboratory batch experiments. Initially, both the adsorption and desorption proceeded rapidly, and either the equilibrium was reached after approximately 30 min or the process slowed down and continued at a slow rate. In the latter case, the equilibrium was reached after 8 hours. Data on the adsorption/desorption kinetics fitted well to the two-site kinetic model. The measured sorption and desorption isotherms were of L-type. The sorption distribution coefficients (K(ads) (d)) were in the range of 0.75--0.97 for Ap soil horizons and significantly lower in deeper soil layers. The corresponding desorption coefficients (K(des) (d)) were higher and ranged from 1.02 to 2.01. Both the adsorption and desorption of MCPA in all soil horizons was strongly and negatively related to soil pH. It appears that hydrophobic sorption plays a dominant role in the MCPA retention in topsoils whereas hydrophilic sorption of MCPA anions is the dominant adsorption mechanism in subsoils.     

Evaluation of impacts of soil fractions on phenanthrene sorption

Phenanthrene sorption to soils and soil fractions was investigated using two contrasting soils with different clay mineral and organic carbon (OC) contents in an attempt to evaluate the contribution of each soil fraction to phenanthrene sorption and the applicability of the carbon-normalized distribution constant (K(OC)) in soils. Sorbents were characterized using surface analysis, solid-state (13)C NMR analysis, and glass transition temperature (T(g)) analysis to gain a insight into the chemical nature of OC in soils. Dissolved organic carbon (DOC) in the soil solution impeded the phenanthrene sorption, while humins accounted for the predominant phenanthrene sorption in soils. The contribution of OC to phenanthrene sorption in soil would be overestimated if only a K(OC)-approach was adopted, since clay minerals could account for much of the sorption, especially when OC was low in soils. Nitrogen gas was shown to be inappropriate for probing non-polar sorption capacity. The results obtained highlight the importance of clay minerals in governing the sorption of phenanthrene in soil, and emphasize the inapplicability of the carbon-normalized distribution coefficient K(OC) in soils.     

Correlation relationships between physico-chemical properties and gas chromatographic retention index of polychlorinated-dibenzofurans

Correlation relationships between physico-chemical properties including vapor pressures (P), water solubilities (S), Henry's law constants (H(c)), n-octanol-water partition coefficients (K(ow)), sediment-water partition coefficient (K(pw)) and biotic lipid-water partition coefficient (K(bw), bioconcentration factor) of polychlorinated-dibenzofurans (PCDFs) and their gas chromatographic retention indices (GC-RIs) were established. A model equation between GC-RIs (= RI) and these physico-chemical properties (K) of PCDFs was in a form of log K = aRI2 + bRI + c with correlation coefficients (R2) greater than 0.94, except H(c). These equations were derived from six experimental data (five experimental data for log K(bw)) in each physico-chemical properties of PCDFs reported previously. The values of log P, log S, log H(c), log K(ow), log K(pw) and log K(bw) of PCDFs predicted by these equations based on their GC-RIs in the present study derviated from those calculated by the solubility parameters for fate analysis method in a previous study by 0.49, 0.32, 0.11, 0.34, 0.14 and 0.22 log units, respectively.     

Development of a continuously stirred flow cell for investigating sorption mass transfer

This paper introduces a new reversible-flow design for a continuously stirred reactor used to study sorption mass transfer in soil and solvent systems. The stirred reactor has potential advantages over conventional packed column or batch reactors because it isolates intraparticle sorption rate limitations from advective-dispersive transport, yet allows changes to flux through the reactor for analysis of sorption kinetics under dynamic conditions. Previously, stirred reactors have often failed due to clogging of sediment on the effluent frit. The reverse-flow backwashing design allows longer life and higher confidence in maintaining mixed conditions than previous designs. Mass transfer 'rate coefficients estimated from stirred and column experiments are compared; both techniques produced results consistent with a published correlation. The data also show that fitted sorption mass transfer coefficients can be strongly dependent on the choice of equilibrium partition coefficient (i.e. batch or first-moment derived values), and that the conventional two-site sorption kinetics model fails to accurately predict sorption mass transfer in the presence of changing solvent velocity through the reactor.     

Comparison of screening methods for the estimation of adsorption coefficients on soil

The HPLC screening method for estimating soil sorption coefficients was compared to other QSAR approaches based on log P_ow, connectivity indices, molar refraction and molecular fragment approaches. For the data set under consideration (66 compounds from different chemical classes) only the HPLC screening method and, to a lesser extend, the log P_ow method proved to be suitable for predicting soil sorption coefficients with acceptable accuracy. The HPLC method and the regression with log P_ow were then cross-validated.     

Estimating K(oc) for persistent organic pollutants: limitations of correlations with K(ow)

The n-octanol/water partition coefficient (K(ow)) is commonly used to predict the soil or aquatic particle water partition coefficient normalized to organic carbon (K(oc)). Many correlations are available covering several chemical classes and ranges of hydrophobicity. This work indicates the K(ow) may not be a strong predictor for persistent organic pollutants (POPs) which are defined here as chemicals with logK(ow) > 5.0. In addition, the correlation developed in this work for POPs will still result in a predicted value which is of by a factor of 15. Accordingly, care must be taken when applying K(oc) estimations using K(ow) for POPs until more suitable correlations are developed.     

Correlations of nonlinear sorption of organic solutes with soil/sediment physicochemical properties

The estimation of solute sorptive behaviors is essential when direct sorption data are unavailable and will provide a convenient way to assess the fate and the biological activity of organic solutes in soil/sediment environments. In this study, the sorption of 2,4-dichlorophenol (2,4-DCP) on 19 soil/sediment samples and the sorption of 13 organic solutes on one sediment were investigated. All sorption isotherms are nonlinear and can be described satisfactorily by a simple dual-mode model (DMM): q(e)=KpCe+Q0 . bCe/(1+bCe), where Kp (mlg(-1)) is the partition coefficient; Ce (microgml(-1)) is the equilibrium concentration; Q0 (microgg(-1)) is the maximum adsorption capacity; Q0 . b (mlg(-1)) is the Langmuir-type isotherm slope in the low concentration (Henry's law) range and b (mlmicrog(-1)) is a constant related to the affinity of the surface for the solute. Based on these nonlinear sorption isotherms and similar other nonlinear isotherms, it is observed that, for both polar 2,4-DCP and nonpolar phenanthrene, Kp, Q0 and Q0 . b are linearly correlated with soil/sediment organic carbon content (f(oc) in the range of 0.118-53.7%). The results indicate that the nonlinear sorption of organic solutes results primarily from interactions with soil/sediment organic matter. The K*oc K*oc=Kp/f(oc)), Qoc (Qoc=Q0/f(oc)), Loc (Loc=Q0 . b/f(oc)) and b for a given organic solute with different soils/sediments are largely invariant. Furthermore, logK*oc, logb and logLoc for various organic solutes are correlated significantly with the solute logKow or logSw (logKow in the range of 0.9 to 5.13 and logSw in the range of -6.176 to -0.070). A fundamental empirical equation was then established to calculate approximately the nonlinear sorption from soil/sediment f(oc) and solute Sw for a given solute equilibrium concentration.     

Fate and transport of monoterpenes through soils. Part I. Prediction of temperature dependent soil fate model input-parameters

Monoterpenes are C10H(n)O(n') compounds of natural origin and are potentially environmentally safe substitutes for traditional pesticides. Still, an assessment of their environmental behaviour is required. As a first step in a theoretical study focussing on monoterpenes applied as pesticides to terrestrial environments, soil fate model input-parameters were determined for 20 monoterpenes with widely different structural characteristics. Input-parameters are the water solubility (S(W)), vapour pressure (P), n-octanol-water partition coefficient (K(OW)), atmospheric air and bulk water diffusion coefficients (D(A)air and D(W)water), first order biodegradation rate constants (k), and their temperature dependence. Values for these parameters were estimated or taken from previous experimental work. The quality of the estimations was discussed by focussing on their statistics and by comparison with available experimental data. From these properties, the air-water partition coefficient (K(AW), Henry's Law constant), the interface-water partition coefficient (K(IW)) and the organic matter-water partition coefficient (K(OM)) could be estimated with varying levels of accuracy. In general, little experimental data turned out to be available on biodegradation rate constants and on the temperature dependence of physico-chemical parameters.     

Investigations on the sorption of phenols to dissolved organic matter by a QSAR study

The sorption of various phenols to Aldrich-HA and BSA was investigated by solid phase microextraction (SPME). The Aldrich-HA sorption with log K(DOC)-values between 2 and 3 was determined, whereas the sorption to BSA with log K(DOC)-values between 2 and 6 was much stronger. To enable an estimation of sorption constants a QSAR model was investigated. The linear free energy relationship (LFER) model showed a good correlation between the sorption constants and the log K(OW)-values with correlation coefficients of R = 0.910 and R = 0.878 for Aldrich-HA and BSA, respectively.