Ion-Exchange Affinity of Organic Cations to Natural Organic Matter

Dec 10, 2012 - Sorption to standard soil organic matter (SOM) has been studied for a wide variety of organic cations using a flow through method with ...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/est

Ion-Exchange Affinity of Organic Cations to Natural Organic Matter: Influence of Amine Type and Nonionic Interactions at Two Different pHs Steven T. J. Droge†,* and Kai-Uwe Goss†,‡ †

Department of Analytical Environmental Chemistry, Helmholtz Centre for Environmental Research - UFZ, Permoserstrasse 15, 04318 Leipzig, Germany. ‡ Institute of Chemistry, University of Halle-Wittenberg, Kurt Mothes Str. 2, D-06120 Halle, Germany. S Supporting Information *

ABSTRACT: Sorption to standard soil organic matter (SOM) has been studied for a wide variety of organic cations using a flow through method with fully aqueous medium as eluent. SOM sorption for weak bases (pKa 4.5−7) was stronger at pH 4.5 than at pH 7, indicating that the ionexchange affinity of the cationic species to SOM was higher than the bulk partition coefficient of corresponding neutral species to SOM. In the range of pH 4.5−7, the effect of pH on the sorption coefficients for strong bases with pKa > 7 was small, within 0.3 log units. For cations with the molecular formula CxHyN, sorption was accurately predicted by a model accounting for size (increase with alkyl chain length) and type of charged group (1° amine >4° ammonium of equal size). In addition to the CxHyN-model, several empirical correction factors were derived from the data for organic cations with polar functional groups. Models based on KOW or pKa fail to explain differences in sorption affinity of the ionic species. Our data on ion-exchange affinities for 80 organic cations show many examples where specific chemical moieties, for example, CH2-units, aromatic rings or hydroxyl groups, contribute differently to the sorption coefficient as compared to bulk partitioning data of neutral compounds. Other sorption models that were evaluated to explain variation between compounds suffered from outliers of more than one log unit and did not reduce relative log mean standard errors below 0.5. A wider range of sorption coefficients and more sorption data in general are required to improve modeling efforts further.

■

INTRODUCTION A wide variety of emerging environmental contaminants, including pharmaceuticals (e.g., beta-blockers), illicit drugs (e.g., amphetamines), and biocides (e.g., benzalkonium structures), occur largely as positively charged molecules at common environmental conditions.1−5 Sorption studies with humic acid (HA) have demonstrated sorption affinities of some monovalent organic cations to be as strong as observed for common hydrophobic environmental contaminants, indicating the relevance of this sorption process. For example, cationic surfactants6 can sorb as strongly to HA as highly chlorinated PCBs,7 and a positively charged macrolide antibiotic8 sorbed as strongly as small PAHs.9,10 Environmental risk assessment (ERA) models need sorption properties of such organic cations, for example to estimate bioavailability to exposed biota, or leaching from soil to groundwater. Experimental sorption data, however, are scarce, and there are no wellvalidated models to estimate sorption properties of organic cations. The 2006 European Union technical guidance document (EU-TGD) for risk assessment of chemicals specifies no sorption model for organic cations, and takes only the © 2012 American Chemical Society

neutral fraction of organic bases into account to estimate the sorption properties at a given pH.11 Ionic compounds are also outside the applicable chemical domain of the Soil Adsorption Coefficient Program (KOCWIN) within the EPISuite v4.10 package from the U.S. Environmental Protection Agency. Ample studies have shown that sorption of organic cations occurs mainly by cation-exchange (with other cationic species present in the system) at negatively charged sorption sites in soils,12−15 sediments,16,17 clays,18,19 and purified natural organic matter.6,8,20−24 Given the wide variety of soil components with high cation-exchange capacities, that is, all organic matter and most clay minerals, adequate understanding of organic cation sorption to soils requires detailed insight in the contribution of each soil component. Currently, however, insight in the most critical parameters that affect the ion-exchange sorption processes in soil is still poor for organic cations, hampering Received: Revised: Accepted: Published: 798

August 17, 2012 November 26, 2012 December 10, 2012 December 10, 2012 dx.doi.org/10.1021/es3033499 | Environ. Sci. Technol. 2013, 47, 798−806

Environmental Science & Technology

Article

ionised at pH 7. For several included weak bases (pKa 4.5−7), we could thereby also compare bulk partitioning to SOM at pH 7 for largely neutral species, with the ion-exchange sorption affinity for largely protonated species at pH 4.5. The third aim was to compare the obtained sorption data with predictive sorption models that are currently used,11 or proposed for use,28,34 in risk assessment modeling of organic cations, and to evaluate whether new sorption models can be established using various sets of molecular descriptors.

adequate implementation of ion-exchange sorption mechanisms in risk assessment models. The sorption affinity of organic cations for negatively charged surface sites is the result of both ionic and nonionic interactions.25,26 Our previous study focused on the influence of dissolved electrolytes on the binding affinity of organic cations to soil organic matter (SOM), and therefore largely addressed the role of ionic interactions in ion-exchange sorption processes.20 The current study follows up on that previous work, but focuses more on the role of nonionic interactions in the sorption affinity of organic cations to ionexchange sorption sites in natural SOM. While the ionic interaction between charged amine group and negatively charged group in the sorbent (e.g., carboxylic acids in natural organic matter) strongly depends on medium composition,20 the contribution of nonionic interactions to the overall sorption affinity depends on the molecular structure attached to the charged amine group. Measurements below the CEC of HA clearly showed that cationic surfactants with longer alkyl chains sorb stronger,6,23 whereas maximum sorption capacities leveled off for all surfactants at the CEC. The latter finding, as well as the observed competition with other cations,6,20 show that the hydrophobic alkyl chain is not the molecular entity that determines the preferred sorption site for organic cations. Sorption still occurs at ion-exchange moieties in the HA structure rather than by adsorption into/onto hydrophobic HA-regions, but the alkyl chain does contribute strongly to the overall binding affinity. It would be highly beneficial for predictive purposes to relate the contribution of nonionic interactions in the overall ion-exchange affinity to well-defined molecular descriptors. The octanol−water partition coefficient (KOW) is the most frequently used molecular descriptor in risk assessment to estimate sorption processes to organic phases, but several reviews and experimental papers have indicated that KOW is a poor descriptor for soil sorption of ionizable structures,27−31 especially for cationic compounds. Since validating existing sorption models for soils, as well as developing new models, is strongly limited by the current lack of high quality sorption data, new sets of consistent data must be generated for a wide variety of organic cations for relevant individual soil components. The first aim of the current study was to investigate in detail how the molecular structure of organic cations influences the overall ion-exchange affinity to SOM. In a similar setup as in previous studies,9,20,32,33 consistent SOM sorption data were generated by measuring the retention on an HPLC column filled with micronized SOM. This system allows for constant conditioning of the sorbent phase material, recovery checks, data consistency, and relatively high-throughput compared to batch sorption studies, while still ensuring sorption equilibrium in the dynamic sorption/desorption passage in the column.9 A wide variety of organic cations were tested in this study under controlled aqueous mobile phase conditions, that is, at a near neutral-pH and constant salt composition, to maintain a more or less fixed contribution of ionic interactions involved in the ion-exchange sorption process. We included several series of bases and quaternary ammonium compounds (QACs) that are composed of the simple chemical structure CxHyN, model amines with single polar groups, and a wide variety of other common organic cations that were all largely ionised at neutral pH. The second aim was to obtain additional sorption coefficients for all compounds at pH 4.5. As such, we covered an environmentally relevant pH range to study the effect of pH for the sorption affinity of strong bases that are already largely

■

MATERIALS AND METHODS Chemicals, Sorbent, And Columns. Table S1 in Supporting Information (SI) lists all 80 tested compounds (primary (1°) amines: P01−18, secondary (2°) amines: S01− 20, tertiary (3°) amines: T01−28, QACs (4°): Q01−14) with CAS-numbers, suppliers and pKa values. SI Figures S1 and S2 show structures of cationic species, other properties are listed in SI Tables S1−S6. Standard mobile phase eluents were prepared using MilliPore water ( 18.2 MΩcm) with addition of 5 mM CaCl2. For near neutral pH, acidification by ingassing CO2 was overcome by addition of 0.1 mM NaOH and aeration for several hours, resulting in average test pH of 6.8 ± 0.1. For tests with slightly acidic eluents, 10 mM HCl was used to obtain average test pH of 4.5 ± 0.1. For some compounds (strong bases P04-S05-S12-T20-T14-Q05, weak bases S01-T04-T06 and neutral 2-methylbenzofuran) tests at pH 6.8 and pH 4.5 were also performed in eluents with 15 mM NaCl instead of 5 mM CaCl2, and mobile phases were further prepared in a similar way. Pahokee peat soil was used as standardized SOM (Bulk solid source 2BS103P, IHSS, Golden, CO) with 12.7% ash content, 46.9% organic carbon (OC) content in ash-free sample, and of the same micronized batch (90% of the species are ionized. Previous studies demonstrated slightly nonlinear sorption of organic cations to organic matter.8,20 As a result of nonlinearity, DOC,IE values are concentration dependent.9 A comparison of DOC,IE between different compounds therefore requires normalization to a fixed sorbed concentration (preferably within the experimental range). This was achieved by creating sorption isotherms from DOC,IE data measured at several different concentrations.20 The representative aqueous concentration of each injected sample (Caq‑est) in the column was estimated from the maximum absorption level of the eluted peak, and, by multiplying Caq‑est with the measured DOC,IE, the corresponding sorbed concentration in the column was estimated (Cs‑est).20 Isotherms derived in this simplified manner overlapped with isotherms obtained with other approaches to estimate concentrations in dynamic column studies, for example, using absorption peak height at the statistical first moment or weighted peak averages.35,37 Further details about the HPLC setup and data evaluation from sorption isotherms are presented in SI section S1. Evaluating Structure Trends Quantitatively. Three major sorption modeling approaches were examined for the ability to describe how sorption coefficients are influenced by the nonionic structure quantitatively, (i) proportionality to octanol−water partition coefficients, with or without corrections for ionization, as recommended in current risk assessment approaches, (ii) empirical fragment contributions, by comparing the effect of specific molecular moieties with a reference value, and (iii), polyparameter approaches, to cover the most relevant molecular interactions by limited sets of specific descriptors.

Figure 1. Experimental sorption coefficients to Pahokee peat (Log KOC for neutral species, log DOC,IE for >90% cationic species, both in L/kgoc and normalized to 1 mmol/kgoc) in 5 mM CaCl2for pH4.5 plotted against pH6.8. Open squares indicate weak bases with pKa < 7, solid dots are strong bases with pKa > 7. Orange diamond is the neutral polar compound 2-methylbenzofuran. The small insert shows data measured in 15 mM NaCl solutions at pH 4.5 and 6.8.

shows that the cation-exchange affinity of a diverse set of cations, for SOM in 5 mM Ca2+, can be as strong as bulk partitioning affinity into SOM of corresponding neutral species. It was expected that the sorption affinity of the strong bases to peat would be higher at pH 6.8 compared to pH 4.5, since more acidic SOM sites are dissociated at higher pH. Figure 1, however, shows only a small but opposite trend. Only three strong bases (pKa > 7) sorb ≥0.3 log units stronger at pH6.8 compared to pH4.5, and on average the sorption affinity of strong bases for peat is 0.22 log units weaker at pH6.8. The reason for the higher affinity at pH4.5 for the strong bases is not obvious. It may be related to the different hydrophobic state of the organic matter surrounding dissociated acidic groups at different pH. Both protonation of acidic groups39 as well as calcium bridging39−41 may lead to increased structural aggregation and therefore possibly a more dense SOM packing that allows for more favorable nonionic interactions. To study the effect of pH independently from Ca2+, we measured logDOC,IE for a small set of organic cations in eluents with 15 mM NaCl at both pH 4.5 and pH 6.8. SI Figure S6 shows the NaCl data and isotherms fitted with nF = 0.80. Figure 1 plots logDOC,IE values for pH 6.8 against pH 4.5. As noted in our earlier work, sorption was substantially stronger in 15 mM Na+ compared to 5 mM Ca2+, even though the applied ionic strength was equal.20 Just like in CaCl2 eluent, the two weakest bases still sorb considerably stronger at pH 4.5 in NaCl eluent. Most strong bases now sorb as strongly as, or stronger, at pH 6.8, as expected for SOM with a higher amount of dissociated acidic groups. The pH effect observed in NaCl eluent also corresponds to observed difference of about a factor of 2 (0.3 log units) between pH 4 and pH 7 for the 3° amine clarithromycin (pKa 8.9).8 This small effect of pH in NaCl eluents reflects the limited increase in the amount of dissociated groups observed in many types of humic and fulvic acids.42 Stronger pH effects on SOM sorption of organic cations can be expected at pH < 4.5, at which most carboxylic groups become protonated.8,43 There appears to be some consistency in the difference in sorption coefficients at pH 4.5 and pH 6.8 for strong bases: logDOC,IE values (SI Figure S6) show a systematic shift of about 0.5 when comparing NaCl data with CaCl2 data. The higher affinity of strong bases in CaCl2 solutions at pH 4.5

■

RESULTS AND DISCUSSION Sorption Data Quality. SI section 2 presents details on sorption data quality. Briefly, measured recoveries were in the range of 102 ± 8% (SI Table S7), background retention on SiC was less than 3% of total measured retention, local sorption equilibrium in the column can be assumed (SI Figure S4A), and data from different columns and detectors were consistent (SI Figure S5). Sorption isotherm nonlinearity (Freunlich exponent nF) was on average 0.81 ± 0.08 (SI Tables S8−S9), and all isotherms were refitted with a fixed nF of 0.8 to obtain logDOC,IE (SI Table S10A,B), for reasons also discussed in our previous paper.20 For 140 out of 157 obtained logDOC,IE values in 5 mM CaCl2 eluent (at both pH 4.5 and 6.8), more than 90% of the compound was in its cationic form (SI Table S1). The Effect of pH on the Sorption Affinity of Organic Cations. Figure 1 plots logDOC,IE values obtained at pH4.5 against those obtained at pH6.8 (both with 5 mM CaCl2). There is no significant effect of pH in this range for the neutral polar compound 2-methylbenzofuran. The tested weak bases (S01, T01-T10) sorbed stronger or to a similar extent at pH 4.5 compared to pH 6.8, while the fraction ionised varied from >90% at pH 4.5 to 2° > 4°, suggesting that the ionic interactions between organic cation and SOM are also influenced by amine type. (2) The regressions on most series of analogue amines or ammonium structures follow fairly similar slopes, indicating constant contribution of increasing nonionic interactions with longer alkyl chain lengths. (3) The slopes in Figure 3A,C represent the

from 7) to derive empirical correction factors that apply to the pH-range 4.5−7 (SI Table S12). A new “average” VxNAi803

dx.doi.org/10.1021/es3033499 | Environ. Sci. Technol. 2013, 47, 798−806

Environmental Science & Technology

Article

predicted logKOC based on the VxNAi-model combined with the 16 correction factors for the 46 polar amines. Nearly all organic cations are predicted within a factor of 3 in this way, even for four out of five multifunctional “test compounds”. The test-outlier T24 (as well as correction factors based on only one compound), however, suggest that this is a rather superficial way to predict cation sorption to SOM, since many more correction factors can be expected to be required if the diversity of (multi)functional groups included in the data set is increased. Polyparameter Models with Different Sets of Molecular Descriptors. Extending the KOW-approach or direct KOC,N-estimates with the charge density parameter A+, to additionally account for the type of charged amine group, does not substantially improve the modeling of the total logDOC,IE data set (SI Figure S11A−D), although CxHyN amines align better. The reason for the wide scatter for polar amines between the bulk partition parameter (KOW) and ion-exchange affinities (logD OC,IE ) is most likely the difference in contribution factors of specific chemical moieties for these two sorption processes, as discussed in detail above, for example, CH2-units, (poly)aromatic rings and hydroxyl groups. As described in more detail in text accompanying SI Figure S11, several other polyparameter approaches were evaluated to model organic cation sorption to SOM, but these either showed very poor correlations (R2 < 0.2) or, as discussed for the MCIapproach, did not adequately reduce the variation observed in logDOC,IE for the charged polar amines (nearly all sy.x > 0.5). For example, logKOC,N values estimated with the Bronner-Goss ppLFER for Pahokee peat,9 applying ABSOLV estimated descriptors (V-E-S-A-B) for all neutral bases (no QACs), scattered widely with logDOC,IE (Figure 2G). Abraham and Acree47 have recently suggested to use specific descriptors for ionic species in a ppLFER approach (Vi-Ei-Si-Ai-Bi, SI Table S4), which are estimated from neutral ppLFER descriptors, and in addition apply a sixth parameter specifically for cations (J+, partially based on NAi). This approach was followed in this study using ABSOLV descriptors. Multiple linear regression of the logDOC,IE(pH4.5) data with the six cation descriptors (SI eq S4 below SI Table S9B, sy.x 0.40) resulted in only five outliers of >0.5 log units (Figure 2H). From all evaluated models, this Abraham/Acree-approach fitted the data with lowest error margins, appearing as good as the correction factor approach. However, the Abraham/Acree-approach deals with free ion partitioning to neutral phases and was not intended as such for ion-exchange processes. Furthermore, there are such uncertainties in each cation descriptor, especially for multifunctional structures (as briefly mentioned in ref 47), that it remains questionable how adequate this ppLFER-approach is for organic cations outside the tested data set. Sorption for the macrolide antibiotic clarithromycin (pKa∼8.9, C38H69NO13, 6 ether, 4 hydroxyl, 1 ketone, 1 ester) has been accurately documented to Elliot soil humic acid.8 As explained in more detail in SI Section 5, the reported clarithromycin logKDOC(1 mmol/kg) was calculated to be 3.9 in medium with 5 mM Ca2+. The Abraham/Acree-approach predicts a logKOC of 1.7 (2.2 log units underestimation). The VxNAi-correction factor approach (SI eq S3 with Table 1) would predict a logKOC of 4.4 (logKOC(only VxNAi) = 9.1), so 0.5 above the reported logKOC. COSMOtherm can also directly calculate sorption coefficients for ionic species (e.g., KOC,i) using deprotonated HA model structures (e.g., HA-model structure with one deprotonated

model (SI) eq S3was created (SI Table S12A). As listed in Table 1, 41 polar amines were used to derive 16 empirical Table 1. Empirical Correction Factors for Polar Fragments in Addition to the Average VxNAi-Model (SI eq S3) average VxNAi-model for pH range 4.5−7 LogDOC,IE = 1.53·Vx+0.32·NAi-0.27 (eq.S3)a functional group

VxNAi-model correction factors(in log units)b

phenyl, or 3xF, or -ΞN, or Sc −Cl polycyclic aromatic ring pyridine NH+ aniline-NH2 −CNC− (HBDon) −CNC− (HBAcc or neutral) benzimidazole −C−NH2 (aniline) −OH −C(O)N-phenyl −C(O)NC− −C(O)OC− −COC− −C(O)C− −C(O)NH2 Internal HB

0 +0.5 (3) +0.7 (2) +0.7 (6) +0.55 (1) +0.6 (2) −0.1 (3) +1.7 (3) +1.2 (1) −0.1 (5) −1.4 (4) −0.4 (1) −0.8 (3) −0.6 (12) +0.1 (2) −0.65 (1) −1.3 (1)

a

32 CxHyN compounds, Q03+Q04 excluded. b41 compounds used, nr. of moieties (e.g., 4 ethers for verapamil) used in parentheses. More details on error margins in SI Table S10. cPhenyl was found to have negligible influence already in the VxNAi-model, was confirmed by dibenzylamine (S18). See SI Table S12B for details.

correction factors, as shown in detail in SI Table S12B. Five polar amines that contained different polar moieties were used to “test” the applicability of some of the correction factors together with the VxNAi-model: four were correctly predicted within 0.2 log units, while clonidine (T24) was now 2.3 log units off. This may be related to a specific influence of charge delocalization over two N atoms in T24 on binding, as shown, for example, by the acetate-complex in Figure 2. Compared to the VxNAi-model, several observations for the polar amines stand out when combining plotted sorption data in SI Figure S8, the correction factors in Table 1, and Turbomole-structures in SI Figure S2. First, additional heterocyclic N-atoms in charged imidazole and indole structures are hydrogen-bond donors (HBD-N, blue in SI Figure S2, for example, P17 and T03), and especially imidazoles were found to enhance sorption. Other heterocyclic N-atoms were HB-acceptors (HBA-N, T25) or their polarity was delocalized by neighboring aromatic rings (e.g., T18), and these on average slightly decreased sorption. Second, polyaromatic structures strongly increased sorption, as observed for the imidazoles (incl. HBD-N), quinoline (T01, incl. pyridine charged N) and propranolol (S04, relative to other beta-blockers). Third, additional OH-groups do not lower the sorption affinity relative to VxNAi-model predictions, whereas they decrease bulk partition coefficients for neutral alcohols by >1 log unit compared to analogue alkanes.9,44 The capacity for internal hydrogen bonding (adrenaline, S14) seems to strongly decrease the sorption affinity. Finally, additional ester-, ether-, or amide-groups decrease the sorption affinity, and the reducing effect of amide groups seems to be stronger for amides next to a benzene ring. Figure 2F shows the result of 804

dx.doi.org/10.1021/es3033499 | Environ. Sci. Technol. 2013, 47, 798−806

Environmental Science & Technology

Article

carboxylic acid with equal mol % hydrated Na+ and 60 mol % water). Disappointingly, the specific COSMOtherm KOC,i tested showed no relation with logDOC,IE (Figure 2I), not even for CxHyN amines, and does not improve sufficiently when combined with A+ or when a different carboxylic acid in the HA model was deprotonated (SI Figure S11E,F). Creating new polyparameter LFERs by multiple regression for logDOC,IE,(pH4.5) data with ABSOLV descriptors or various COSMOtherm sigma-moments, with or without A+, did not seem to result in adequately predictive models (SI Figure S11G−L). Among the consistently underestimated outliers in these models are mostly polyaromatic structures, but also amides and procaine (T16, ester and aniline NH2). Evaluation of the Modeling Approaches. The relatively poor predictive power of most modeling approaches suggests that the sorption process of organic cations may need much more refined insight in the sorption sites where interactions between compounds and SOM actually occur. In bulk partitioning processes, the molecules are fully immersed in the solvent system and the solvent can effectively interact at all surfaces of the dissolved molecule. Ion-exchange sites on strongly hydrated natural organic matter are most likely directed toward the medium, rendering the interaction site as a surface phenomenon. This may explain the relatively low CH2-unit increment for CxHyN compounds, and also why none of the multiparameter regressions seem to adequately explain the variation in the whole data set for polar organic cations. Flat molecular surfaces may exhibit more effective sorption interactions, as suggested by the positive correction factor for polyaromatic moieties in the VxNAi-model approach. Another unresolved issue is how the cationic moiety of each compound actually binds to the SOM ion-exchange site. The Turbomoleoptimized structures for acetate-complexes (SI Figure S2) are merely first approximations of actual (nonhydrated) interaction. In the absence of accurate 3D-modeling of the sorption interactions, as recommended in other work,48 current improvement in risk assessment may rather rely on empirical contributions for the various functional groups present in our test set (SI Figure S11A,B). A wider range of SOM sorption coefficients are required to improve the empirical correction factors. Sorption data to other soil phases are also needed to improve ERA modeling development for the sorption of organic cations to natural soils.

■

group of CEFIC (European Chemical Industry Council). We thank Satoshi Endo and anonymus reviewers for helpful comments.

■

(1) Alder, A. C.; Schaffner, C.; Majewsky, M.; Klasmeier, J.; Fenner, K. Fate of [beta]-blocker human pharmaceuticals in surface water: Comparison of measured and simulated concentrations in the Glatt Valley Watershed, Switzerland. Water Res. 2010, 44 (3), 936−948. (2) Zuccato, E.; Castiglioni, S.; Bagnati, R.; Chiabrando, C.; Grassi, P.; Fanelli, R. Illicit drugs, a novel group of environmental contaminants. Water Res. 2008, 42 (4−5), 961−968. (3) Martínez-Carballo, E.; Sitka, A.; González-Barreiro, C.; Kreuzinger, N.; Fürhacker, M.; Scharf, S.; Gans, O. Determination of selected quaternary ammonium compounds by liquid chromatography with mass spectrometry. Part I. Application to surface, waste and indirect discharge water samples in Austria. Environ. Pollut. 2007, 145 (2), 489−496. (4) Martínez-Carballo, E.; González-Barreiro, C.; Sitka, A.; Kreuzinger, N.; Scharf, S.; Gans, O. Determination of selected quaternary ammonium compounds by liquid chromatography with mass spectrometry. Part II. Application to sediment and sludge samples in Austria. Environ. Pollut. 2007, 146 (2), 543−547. (5) Li, X.; Brownawell, B. J. Quaternary ammonium compounds in urban estuarine sediment environments - A class of contaminants in need of increased attention? Environ. Sci. Technol. 2010, 44 (19), 7561−7568. (6) Ishiguro, M.; Tan, W.; Koopal, L. K. Binding of cationic surfactants to humic substances. Colloids Surf. A 2007, 306 (1−3 SPEC. ISS.), 29−39. (7) Durjava, M. K.; ter Laak, T. L.; Hermens, J. L. M.; Struijs, J. Distribution of PAHs and PCBs to dissolved organic matter: High distribution coefficients with consequences for environmental fate modeling. Chemosphere 2007, 67 (5), 990−997. (8) Sibley, S. D.; Pedersen, J. A. Interaction of the macrolide antimicrobial clarithromycin with dissolved humic acid. Environ. Sci. Technol. 2008, 42 (2), 422−428. (9) Bronner, G.; Goss, K.-U. Sorption of organic chemicals to soil organic matter: influence of soil variability and pH dependence. Environ. Sci. Technol. 2011, 45 (4), 1307−1312. (10) Ter Laak, T. L.; Durjava, M.; Struijs, J.; Hermens, J. L. M. Solid phase dosing and sampling technique to determine partition coefficients of hydrophobic chemicals in complex matrixes. Environ. Sci. Technol. 2005, 39 (10), 3736−3742. (11) EU Technical Guidance Document on Risk Assessment, Part II (Chapter 3) Environmental Risk Assessment; European Chemicals Bureau: Ispra, 2006. (12) Bi, E.; Schmidt, T. C.; Haderlein, S. B. Environmental factors influencing sorption of heterocyclic aromatic compounds to soil. Environ. Sci. Technol. 2007, 41 (9), 3172−3178. (13) Beaulieu, J. J.; Tank, J. L.; Kopacz, M. Sorption of imidazoliumbased ionic liquids to aquatic sediments. Chemosphere 2008, 70, 1320− 1328. (14) MacKay, A. A.; Seremet, D. E. Probe compounds to quantify cation exchange and complexation interactions of ciprofloxacin with soils. Environ. Sci. Technol. 2008, 42 (22), 8270−8276. (15) Stepnowski, P.; Mrozik, W.; Nichthauser, J. Adsorption of alkylimidazolium and alkylpyridinium ionic liquids onto natural soils. Environ. Sci. Technol. 2007, 41 (2), 511−516. (16) Brownawell, B. J.; Chen, H.; Collier, J. M.; Westall, J. C. Adsorption of organic cations to natural materials. Environ. Sci. Technol. 1990, 24 (8), 1234−1241. (17) Ramil, M.; El Aref, T.; Fink, G.; Scheurer, M.; Ternes, T. A. Fate of beta blockers in squatic-sediment systems: sorption and biotransformation. Environ. Sci. Technol. 2009, 44 (3), 962−970. (18) Polubesova, T.; Nir, S. Modeling of organic and inorganic cation sorption by Illite. Clays Clay Miner. 1999, 47 (3), 366−374.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information file includes chemical structures and properties, details on methods and data quality, fitted isotherm parameters and plotted sorption isotherms, and more detailed model evaluations. This information is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: +49-341-2351411; fax +49-341-2351443; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by a grant of APAG (The European Oleochemicals & Allied Products Group), a Sector 805

dx.doi.org/10.1021/es3033499 | Environ. Sci. Technol. 2013, 47, 798−806

Environmental Science & Technology

Article

(19) Polubesova, T.; Rytwo, G.; Nir, S.; Serban, C.; Margulies, L. Adsorption of benzyltrimethylammonium and benzyltriethylammonium on montmorillonite: Experimental studies and model calculations. Clays Clay Miner. 1997, 45 (6), 834−841. (20) Droge, S. T. J.; Goss, K.-U. Effect of sodium and calcium cations on the ion-exchange affinity of organic cations for soil organic matter. Environ. Sci. Technol. 2012, 46 (11), 5894−5901. (21) Karthikeyan, K. G.; Chorover, J. Humic acid complexation of basic and neutral polycyclic aromatic compounds. Chemosphere 2002, 48 (9), 955−964. (22) Richter, M. K.; Sander, M.; Krauss, M.; Christl, I.; Dahinden, M. G.; Schneider, M. K.; Schwarzenbach, R. P. Cation binding of antimicrobial sulfathiazole to Leonardite humic acid. Environ. Sci. Technol. 2009, 43 (17), 6632−6638. (23) Matsuda, M.; Kaminaga, A.; Hayakawa, K.; Takisawa, N.; Miyajima, T. Surfactant binding by humic acids in the presence of divalent metal salts. Colloids Surf. A 2009, 347 (1−3), 45−49. (24) Chen, Y.; Droge, S. T. J.; Hermens, J. L. M. Analyzing freely dissolved concentrations of cationic surfactant utilizing ion-exchange capability of polyacrylate coated solid-phase microextraction fibers. J. Chromatogr., A 2012, 1252, 15−22. (25) Cheng, S. L.; Yan, H. S.; Zhao, C. Q. The synergistic effect between hydrophobic and electrostatic interactions in the uptake of amino acids by strongly acidic cation-exchange resins. J. Chromatogr., A 2006, 1108 (1), 43−49. (26) Murakami, F. Quantitative structure-retention relationships in cation-exchange chromatography. J. Chromatogr., A 1980, 198 (3), 241. (27) Tolls, J. Sorption of veterinary pharmaceuticals in soils: A review. Environ. Sci. Technol. 2001, 35 (17), 3397−3406. (28) Franco, A.; Trapp, S. Estimation of the soil-water partition coefficient normalized to organic carbon for ionizable organic chemicals. Environ. Toxicol. Chem. 2008, 27 (10), 1995−2004. (29) Holten-Lutzhoft, H.-C.; Vaes, W. H. J.; Freidig, A. P.; HallingSorensen, B.; Hermens, J. L. M. Influence of pH and other modifying factors on the distribution behavior of 4-quinolones to solid phases and humic acids studied by ″negligible-depletion″ SPME-HPLC. Environ. Sci. Technol. 2000, 34 (23), 4989−4994. (30) Arnold, C. d. G.; Ciani, A.; Müller, S. R.; Amirbahman, A.; Schwarzenbach, R. P. Association of triorganotin compounds with dissolved humic acids. Environ. Sci. Technol. 1998, 32 (19), 2976− 2983. (31) Hörsing, M.; Ledin, A.; Grabic, R.; Fick, J.; Tysklind, M.; Jansen, J. l. C.; Andersen, H. R. Determination of sorption of seventy-five pharmaceuticals in sewage sludge. Water Res. 2011, 45 (15), 4470− 4482. (32) Bronner, G.; Goss, K.-U. Predicting sorption of pesticides and other multifunctional organic chemicals to soil organic carbon. Environ. Sci. Technol. 2011, 45 (4), 1313−1319. (33) Tülp, H. C.; Fenner, K.; Schwarzenbach, R. P.; Goss, K.-U. pHDependent sorption of acidic organic chemicals to soil organic matter. Environ. Sci. Technol. 2009, 43 (24), 9189. (34) Trapp, S.; Franco, A.; Mackay, D. Activity-based concept for transport and partitioning of ionizing organics. Environ. Sci. Technol. 2010, 44 (16), 6123−6129. (35) Schenzel, J.; Goss, K.-U.; Bucheli, T.; Schwarzenbach, R. P.; Droge, S. T. J. Experimentally determined soil organic matter-water sorption coefficients for different classes of natural toxins and comparison with estimated numbers. Environ. Sci. Technol. 2012, 46 (11), 6118−6126. (36) Lyon, W. G. Swelling of peats in liquid methyl, tetramethylene and propyl sulfoxides and in liquid propyl sulfone. Environ. Toxicol. Chem. 1995, 14 (2), 229−236. (37) Gustafsson, Ö .; Bucheli, T. Soot sorption of non-ortho and ortho substituted PCBs. Chemosphere 2003, 53, 515−522. (38) Ter Laak, T. L.; Gebbink, W. A.; Tolls, J. The effect of pH and ionic strength on the sorption of sulfachloropyridazine, tylosin, and oxytetracycline to soil. Environ. Toxicol. Chem. 2006, 25 (4), 904−911.

(39) Schimpf, M. E.; Petteys, M. P. Characterization of humic materials by flow field-flow fractionation. Colloids Surf. A 1997, 120 (1−3), 87−100. (40) Kalinichev, A. G.; Kirkpatrick, R. J. Molecular dynamics simulation of cationic complexation with natural organic matter. Eur. J. Soil Sci. 2007, 58 (4), 909−917. (41) Siripinyanond, A.; Worapanyanond, S.; Shiowatana, J. Field-flow fractionation-inductively coupled plasma mass spectrometry: An alternative approach to investigate metal-humic substances interaction. Environ. Sci. Technol. 2005, 39 (9), 3295−3301. (42) Milne, C. J.; Kinniburgh, D. G.; Tipping, E. Generic NICADonnan model parameters for proton binding by humic substances. Environ. Sci. Technol. 2001, 35 (10), 2049−2059. (43) Nielsen, T.; Siigur, K.; Helweg, C.; Jørgensen, O.; Hansen, P. E.; Kirso, U. Sorption of polycyclic aromatic compounds to humic acid as studied by high-performance liquid chromatography. Environ. Sci. Technol. 1997, 31 (4), 1102−1108. (44) Endo, S.; Grathwohl, P.; Haderlein, S. B.; Schmidt, T. C. LFERs for soil organic carbon-water distribution coefficients (Koc) at environmentally relevant sorbate concentrations. Environ. Sci. Technol. 2009, 43 (9), 3094−3100. (45) Trapp, S.; Horobin, R. W. A predictive model for the selective accumulation of chemicals in tumor cells. Eur. Biophys. J. 2008, 34, 959−966. (46) Franco, A.; Wenjing, F. U.; Trapp, S. Influence of soil pH on the sorption of ionizable chemicals: Modeling advances. Environ. Toxicol. Chem. 2009, 28 (3), 458−464. (47) Abraham, M.; Acree, W. E., Jr Equations for the transfer of neutral molecules and ionic species from water to organic phases. J. Org. Chem. 2010, 75 (4), 1006−1015. (48) Aristilde, L.; Sposito, G. Molecular modeling of metal complexation by a fluoroquinolone antibiotic. Environ. Toxicol. Chem. 2008, 27 (11), 2304−2310.

806

dx.doi.org/10.1021/es3033499 | Environ. Sci. Technol. 2013, 47, 798−806