Article pubs.acs.org/est
Partitioning of Organic Chemicals to Storage Lipids: Elucidating the Dependence on Fatty Acid Composition and Temperature Anett Geisler,‡,† Satoshi Endo,‡ and Kai-Uwe Goss*,‡,§ ‡
Helmholtz Centre for Environmental Research UFZ, Permoserstr. 15, D-04318 Leipzig, Germany University of Halle-Wittenberg, Institute of Chemistry, Kurt Mothes Str. 2, D-06120 Halle, Germany
§
S Supporting Information *
ABSTRACT: Lipids serve as important compartments in partitioning of neutral organic chemicals into organisms. Storage lipids, made up of triglycerides with various fatty acids, are among the major classes of lipids. Here, we present experimental equilibrium partition data for diverse chemicals in fish oil, linseed oil, and goose fat at 37 °C. These data, in combination with data from the literature for olive oil and milk fat, show that the fatty acid composition of triglycerides has no significant influence on the partition coefficient. This result allows the derivation of a general predictive model for partitioning into storage lipids. We have collected storage lipid/water partition coefficients for 247 compounds to calibrate polyparameter linear free energy relationships (pp-LFERs) for 37 °C, which achieved a model fit with a root mean squared error of 0.20 log units. To extend the applicability of this model toward the aquatic food chain, we also measured fish oil partition data at 7 °C. The resulting enthalpies were used to calibrate an additional pp-LFER for the temperature dependence of storage lipid/water partitioning. This model allows us to estimate partition coefficients at desired temperatures that occur under typical ambient conditions.
■
INTRODUCTION Lipids are an important, and often even dominant, compartment for bioaccumulation of organic chemicals. Two structurally different classes of lipids can be distinguished: phospholipids in membranes and triacylglycerides as storage lipids. The work from Endo et al.1 suggested that the properties of phospholipids as partitioning phases significantly differ from those of olive oil, a model for storage lipid. In the literature, there are numerous experimental partition coefficients for selected chemicals and lipids2−5 but only a few data allow for a systematic comparison between different types of storage lipids. The available data generally give an indication that differences in partition coefficients into different types of storage lipids may be small. For example, Mayer et al.6,7 found that polycyclic aromatic hydrocarbons (PAHs) partition similarly to fish oil, olive oil, rapeseed oil, and sunflower oil, and polychlorinated biphenyls (PCBs) to olive oil, tobis fish oil, and seal oil. However, this work did not cover any organic chemicals with functional groups. In more recent work, we showed for polar and nonpolar chemicals that different types of milk did not differ in their partition properties and that milk fat resembled olive oil with regard to partitioning.8 Niimi9 reported similar solubilities of various, mainly hydrophobic organic chemicals in triolein and cod liver oil. However, a systematic investigation is still lacking. Recently, van der Heijden and Jonker reported a significant interspecies variability in partitioning of PCBs to homogenates of aquatic worms and sticklebacks when normalized to total lipid content.10 On the basis of this observation, Jonker11 noted that field-observed biomagnification through a food chain could be explained by differing © 2012 American Chemical Society
accumulation capacities of lipids in organisms. This notion emphasizes a need for further studies on the variability of accumulation properties of lipids. Lipid/water partition coefficients are typically measured at 37 °C for the relevance to humans and other mammals, or around 25 °C for experimental convenience. However, for assessment of bioaccumulation in fish and other aquatic organisms it is necessary to predict partitioning to lipids at lower temperatures. Considering the temperature variability in the environment, it is desirable to establish a model for the temperature dependence of lipid partitioning. The main goal of the present work was to test the variability of storage lipids from different animals and plants with regard to their capacity to accumulate organic chemicals. To this end, a diverse set of polar and nonpolar chemicals were tested for their partitioning into various storage lipids that cover a wide range of fatty acid compositions. In a next step, these new partitioning data were combined with existing data in the literature to yield a comprehensive data set at 37 °C, which was used to refine a model for predicting partitioning to storage lipids in general. Finally, to extend the applicability of the model equation to other temperatures we determined partition coefficients into fish oil at 7 °C. Enthalpies were calculated from the partition coefficients at 7 and 37 °C and were then Received: Revised: Accepted: Published: 9519
May 20, 2012 July 23, 2012 July 31, 2012 July 31, 2012 dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
Environmental Science & Technology
Article
Table 1. Typical Fatty Acid Compositions of Investigated Animal and Plant Oils and Fats as Mass Percent of Total Lipids fatty acid olive oil15 linseed oil15 milk fat16 fish oil17 goose fat15
4:0
3.3
6:0
1.6
8:0
1.3
10:0
12:0
3.0
3.6
14:0
16:0
16:1
18:0
18:1
18:2
18:3
20:0
1.5
2.5 3.5 14.6 2.1 6.5
75.5 18.0 29.8 22.4 58.0
7.5 14.0 2.4 2.0 9.5
1.0 58.0 0.8 1.0 2.0
0.5
9.5 3.7 0.5
11.5 6.5 26.3 9.7 21.0
2.3 8.5 2.5
20:1
20:4
20:5
22:1
22:5
22:6
14.4
0.9
9.5
7.5
1.3
12.7
Table 2. Selected Chemicals from Various Compound Classes with Their Experimental Storage Lipid/Water Partition Coefficients and the Corresponding Average Values exp. log Kstorage lipid/water substance
olive oil ref 14
milk fat ref 8
fish oil
linseed oil
goose fat
average ± SD
1-chloropentane 1-nitropropane pentafluorobenzene propyl acetate 1-nonanol propylbenzene toluene 2-hexanone cyclohexane carbon tetrachloride 2-nitrotoluene halothane 4-ethyl-3-hexanol decane ethyl tert-butyl ether
3.51 0.90 2.46 1.12
3.41 0.99 2.30 0.97 3.00a 3.76 2.75 1.21 4.01 3.22 2.51 2.40 2.07a 6.58 1.43
3.47 0.97 2.19 1.03
3.54 1.06 2.27 1.06 3.01 3.54 2.63 1.31 4.06 3.19 2.45 2.26 1.93 6.25 1.35
3.48 0.94 2.32 0.93
3.48 ± 0.05 0.97 ± 0.06 2.31 ± 0.10 1.02 ± 0.08 3.01 3.65 ± 0.11 2.67 ± 0.05 1.19 ± 0.10 4.05 ± 0.04 3.18 ± 0.03 2.40 ± 0.13 2.28 ± 0.07 1.92 ± 0.07 6.39 ± 0.15 1.37 ± 0.04
3.77 2.68 1.04 4.01 3.14 2.26 6.52 1.31
3.60 2.65 1.25 4.08 3.20 2.43 2.25 1.91 6.28 1.36
3.58 2.63 1.16 4.08 3.17 2.21 2.23 1.93 6.33 1.39
a
Milk fat data were not included in the average because of possible impact of milk proteins on the values (main text and Table SI-2 of the Supporting Information for details).
Technologies, and Rtx-VMS (30 m × 0.25 mm i.d., 1.4 μm film thickness) from Restek (Bellefonte, PA). Experimental Work. Equilibrium partition coefficients were determined at 37 ± 1 °C with batch sorption experiments using headspace measurements. Two sets of 20 mL vials were prepared; one filled with 1 mL oil and the other with 5 mL water. Both the oil vial and the water vial were spiked with 10 μL of methanolic stock solutions. After equilibration for 4 h in a temperature controlled horizontal shaker (GLS 400, Grant Instruments, UK), the headspace was analyzed with GC. The storage lipid/water partition coefficient (Kstorage lipid/water) was determined from the comparison of peak areas between oil vial and water vial. For more details on the experimental setup and the data evaluation, we refer to ref 8. Experiments at 7 °C were carried out using fish oil with the same headspace method as described above but equilibration and analysis took place at 7 °C and the equilibration time was 18 h. Fish oil was liquid at 7 °C. Model Calibration. The logarithm of a partition coefficient is a free energy related property and can be described with a solvation parameter model. The partitioning of neutral organic chemicals between two condensed phases can be described by each of the following two equations:
used to derive a model that predicts the temperature dependence of partitioning into storage lipids.
■
MATERIALS AND METHODS Materials. The linseed oil (Kunella) and the goose fat (LARU) were standard products from the local grocery store. The fish oil (cod liver oil, Caelo) was obtained from the pharmacy. These storage lipid samples were all liquid at 37 °C. Chemicals measured in this study for calibration were carefully selected so that wide ranges of descriptor values were covered and cross correlations of the descriptors within the calibration data set were minimized. The chemicals used cover many different compound classes like halogenated aliphatic and aromatic compounds, ketones, alcohols, esters, ethers, aldehydes, alkanes, alkyl-aromatic compounds, and N-containing compounds. They were purchased from Sigma-Aldrich, Alfa-Aesar or Dr. Ehrenstorfer GmbH. Methanol was purchased from Merck (of LiChrosolv quality). All solutes were first dissolved in methanol. Pure water produced by a Milli-Q Gradient A10 system (Millipore) was used. Instruments. The following equipment was used: A Hewlett-Packard GC System HP 6890 series with a flame ionization detector and an electron capture detector linked to a HP 7694 Headspace Sampler; a 7890A GC System with a 5975C inert MSD with Triple Axis Detector (Agilent Technologies, Inc., Wilmington, DE) coupled to a Multi Purpose Sampler (MPS 2XL; Gerstel, Mülheim a.d. Ruhr, Germany). The following analytical columns were used: HP-1 (30 m × 0.32 mm i.d., 4 μm film thickness), HP-5 MS (30 m × 0.25 mm i.d., 0.25 μm film thickness), and DB-1701 (30 m × 0.25 mm i.d., 0.25 μm film thickness) from Agilent
log K i1/2 = e1/2E i + s1/2Si + a1/2A i + b1/2Bi + v1/2Vi + c1/2
(1)
log K i1/2 = l1/2L i + s1/2Si + a1/2A i + b1/2Bi + v1/2Vi + c1/2
(2)
The dependent variable in eqs 1 and 2, log Ki 1/2, is the partition coefficient of substance i between two phases 1 and 2. The solute descriptors in capital letters (Ei, Li, Si, Ai, Bi, Vi) describe the extent to which a given solute undergoes molecular 9520
dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
Environmental Science & Technology
Article
interactions in condensed phases. The corresponding phase descriptors lower case letters (e1/2, l1/2, s1/2, a1/2, b1/2, v1/2) describe the difference in capacity between the two phases with regard to intermolecular interactions. The five descriptor pairs quantify the intermolecular interactions that govern the partition process: polar interactions (s1/2 Si), van der Waals interactions and cavity formation (v1/2 Vi, e1/2 Ei, or l1/2 Li), and H-bond interactions (a1/2 Ai, b1/2 Bi). These phase descriptors are calibrated with experimental log Ki 1/2 by multiple linear regression analysis. Eq 1 was established by Abraham and coworkers,12 whereas a modified version of this equation (eq 2) was proposed by Goss.13 These two equations generally give the same quality of fit. Here, we refer to eqs 1 and 2 as polyparameter linear free energy relationships (pp-LFERs).
Therefore, for phenols and anilines we performed additional experiments with skim milk (0.3 log units).20,24 Thus, storage lipids are certainly among the phases that are most accurately predictable in the environment. A comparison of eq 4 with the respective equation for olive oil/water partitioning from the literature18 reveals some minor differences in the polar interaction parameters (part A of Table SI-4 of the Supporting Information). Interestingly, a direct comparison of the predictions from both equations for all of 9521
dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
Environmental Science & Technology
Article
descriptors that are used in the pp-LFER equations.26−28 The respective fitting based on the descriptor combination Li, Si, Ai, Bi, and Vi is given in eq 6. ΔHi = (10.51 ± 7.74)L i − (49.29 ± 6.26)Si − (16.36 ± 8.71)A i + (70.39 ± 6.27)Bi − (66.19 ± 12.51)Vi + 38.95 ± 4.73 (SD: 5.96, n = 46, R2 = 0.83)
(6)
Figure 2 shows the goodness of fit for this equation. All experimental data are presented in the Supporting Information,
Figure 1. Experimental storage lipid/water partition coefficients for all compounds and all lipids versus storage lipid/water partition coefficient calculated with eq 4, 517 data points.
our 247 chemicals shows a better performance of our equation (eq 4) for highly fluorinated compounds (part B of Table SI-4 and Figure SI-4 of the Supporting Information). Contrary to expectations, improvement in estimations for H-bond donor compounds and PAHs was relatively small (by up to 0.1 log unit). Endo et al.1 noted a difference in the partitioning properties of membrane lipids and storage lipids based on the olive oil partition model from ref 18. In line with this finding, we observe substantial differences between the experimental storage lipid/water data reported here and predicted (based on a pp-LFER from ref 1) membrane/water partitioning for two groups of chemicals: H-bond donor compounds like alcohols and undissociated acids have partition coefficients that are 0.5 to 1.5 log units higher for membrane lipids as compared to storage lipids. In contrast, alkanes, cycloalkanes, and alkenes show partition coefficients that are 0.5 to 1.5 log units smaller for membrane lipids as compared to storage lipids (also, part B of Table SI-4 and Figure SI-4 of the Supporting Information). The calibrated pp-LFER equation for storage lipids also deviates from the pp-LFER equation for octanol (part A of Figure SI-4 of the Supporting Information) This suggests that octanol might not in all cases be a good model for storage lipids. We will further investigate this in an upcoming article. As a validation of the calibrated pp-LFER model, experimental data for the partitioning to human and rat adipose tissue at 37 °C collected in ref 25 were compared with predictions using eq 4. The agreement is good, with an rmse of 0.19 (a figure and a table are shown in SI-5 of the Supporting Information). Temperature Dependence. Experimental fish oil/water partition coefficients at 7 °C (Ki oil/water[7 °C]; data shown in SI-6 of the Supporting Information) were used in combination with the corresponding data at 37 °C (Ki oil/water[37 °C]) to calculate the enthalpies of partitioning (ΔHi) based on the van't Hoff equation. ΔHi = −
(log K i oil/water[7°C] − log K i oil/water[37°C])
( 280.131 K − 310.131 K )
Figure 2. Experimental enthalpy versus fitted enthalpy for fish oil/ water partitioning, 46 data points.
also the fitted equation based on Ei, Si, Ai, Bi, and Vi is shown in SI-6 of the Supporting Information. It must be noted that the calibration set for ΔHi is far less diverse than that for log Kstorage lipid/water (eq 4), and the experimental uncertainty for ΔHi is larger than for log Kstorage lipid/water because ΔHi is based on two experimental log Kstorage lipid/water values at different temperatures. This explains why the scatter is larger in Figure 2 than in Figure 1. Fortunately, the temperature dependence for many compounds is small (ΔHi close to zero) and, in practice, temperature extrapolations are needed for only small temperature ranges under typical environmental conditions. Therefore, estimations based on eq 6 should be sufficient for most applications. By using the general equation for predicting storage lipid/ water partitioning at 37 °C (eq 4) and extrapolating these values to the actual temperature based on the enthalpy predicted with eq 6, Kstorage lipid/water at a desired temperature can be calculated. From our own measurements at 7 °C in fish oil, there are data for six compounds that were not used for calibration of eq 6 (because of no value at 37 °C) and that can be used for validation of this predictive approach. In addition, data from the literature6,7 for the partitioning of PAHs and PCBs at 21 °C in two oils, which had not been among the oils used here (i.e., rapeseed oil and seal oil), were also used to validate the approach. Figure 3 compares these validation data with calculated values that result from the combined predictions of eqs 4 and 6. The physicochemical properties and experimental values used for Figure 3 are presented in SI-7 of the Supporting Information. Calibration data for eq 6 are also presented in Figure 3 for comparison. The predicted values
R × 2.303
(5)
The resulting enthalpies are reported in Table SI-6 of the Supporting Information. It has been shown before that enthalpies of partitioning can often be fitted with the same 9522
dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
Environmental Science & Technology
■
Article
ASSOCIATED CONTENT
S Supporting Information *
Tables of physicochemical properties and experimental results for the chemicals in the calibration data set. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected], phone: +49 341 235 1411. Present Address †
Landesbetrieb Hessisches Landeslabor, Fachbereich III.4, Glarusstraße 6, D-65203 Wiesbaden, Germany.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors would like to thank Andrea Pfennigsdorff and Luise Oemisch, Helmholtz Centre for Environmental Research UFZ, for helpful support in the lab. This work was supported by the Helmholtz Impulse and Networking Fund through Helmholtz Interdisciplinary Graduate School for Environmental Research (HIGRADE).
Figure 3. Experimental and calculated log Kstorage lipid/water. Calibration data for temperature dependence and validation data from this study are presented. Data from PAHs are from ref 6. Data for PCBs are from ref 7.
■
agree well with the experimental data validating the models presented here. Implications. The results of this work have important implications for the modeling and understanding of bioaccumulation of neutral organic chemicals. First, storage lipids from different origins do not differ in their accumulation properties for various polar and nonpolar organic chemicals. This is important for the prediction of bioaccumulation and for the normalization of measured bioaccumulation data. Interspecies differences in lipid-normalized biota-water partition coefficients as observed by van der Heijden and Jonker10 for PCBs cannot be explained by the different types of storage lipids. An alternative explanation may come from the membrane lipids, which also contribute significantly to the sorption capacity of organisms. Note, however, that partition coefficients of many hydrophobic compounds to storage lipids and to a model membrane (phosphatidylcholine liposome) have been shown to be similar as well.1 Thus, to elucidate the interspecies differences, further studies on partitioning into various biological membranes are suggested. Second, the presented pp-LFER model refines an existing model from the literature and allows more accurate predictions for neutral chemicals provided that the respective compound descriptors are available. This model was amended by the calibration of a predictive equation for the enthalpies of the partitioning process, which enables the prediction of temperature-dependent partition coefficients into storage lipids. The presented results will help to improve the accuracy of predictions for the partitioning in biological systems and account for temperature-dependent differences between mammals (homeotherm) and fish (poikilotherm). However, for high accuracy of the model predictions, a high quality of the experimental solute descriptors is a requirement. For many organic chemicals of environmental concern, such high quality experimental descriptors do not exist. In a future article, we will use the data presented here as well as additional experimental results to evaluate various models that predict partitioning to storage lipids based on the solutes’ molecular structure as the only model input.
REFERENCES
(1) Endo, S.; Escher, B. I.; Goss, K.-U. Capacities of membrane lipids to accumulate neutral organic chemicals. Environ. Sci. Technol. 2011, 45, 5912−5921. (2) Chiou, C. T. Partition-coefficients of organic-compounds in lipidwater systems and correlations with fish bioconcentration factors. Environ. Sci. Technol. 1985, 19 (1), 57−62. (3) Gargas, M. L.; Burgess, R. J.; Voisard, D. E.; Cason, G. H.; Andersen, M. E. Partition-coefficients of low-molecular-weight volatile chemicals in various liquids and tissues. Toxicol. Appl. Pharmacol. 1989, 98 (1), 87−99. (4) Pollien, P.; Roberts, D. Use of solid-phase microextraction for measuring oil-water partition coefficients and correlation with highperformance liquid chromatographic methods for lipophilicity. J. Chromatogr., A 1999, 864 (2), 183−189. (5) Van Ruth, S. M.; Grossmann, I.; Geary, M.; Delahunty, C. M. Interactions between artificial saliva and 20 aroma compounds in water and oil model systems. J. Agr. Food Chem. 2001, 49 (5), 2409−2413. (6) Mayer, P.; Torang, L.; Glaesner, N.; Jonsson, J. A. Silicone membrane equilibrator: Measuring chemical activity of nonpolar chemicals with poly(dimethylsiloxane) microtubes immersed directly in tissue and lipids. Anal. Chem. 2009, 81 (4), 1536−1542. (7) Jahnke, A.; McLachlan, M. S.; Mayer, P. Equilibrium sampling: Partitioning of organochlorine compounds from lipids into polydimethylsiloxane. Chemosphere 2008, 73 (10), 1575−1581. (8) Geisler, A.; Endo, M.; Goss, K.-U. Partitioning of polar and nonpolar neutral organic chemicals into human and cow milk. Environ. Int. 2011, 37, 1253−1258. (9) Niimi, A. J. Solubility of organic-chemicals in octanol, triolein and cod liver oil and relationships between solubility and partitioncoefficients. Water Res. 1991, 25 (12), 1515−1521. (10) van der Heijden, S. A.; Jonker, M. T. O. Intra- and interspecies variation in bioconcentration potential of polychlorinated biphenyls: Are all lipids equal? Environ. Sci. Technol. 2011, 45 (24), 10408− 10414. (11) Jonker, M. T. O. What is causing biomagnification of persistent hydrophobic organic chemicals in the aquatic environment? Environ. Sci. Technol. 2012, 46, 110−111. (12) Abraham, M. H.; Andonian-Haftvan, J.; Whiting, G. S.; Leo, A.; Taft, R. S. Hydrogen bonding. Part 34. The factors that influence the solubility of gases and vapours in water at 298 K, and a new method for its determination. J. Chem. Soc., Perkin Trans. 2 1994, 7, 1777− 1791. 9523
dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
Environmental Science & Technology
Article
(13) Goss, K.-U. Predicting the equilibrium partitioning of organic compounds using just one linear solvation energy relationship (LSER). Fluid Phase Equilib. 2005, 233, 19−22. (14) Abraham, M. H.; Ibrahim, A. Gas to olive oil partition coefficients: A linear free energy analysis. J. Chem. Inf. Model. 2006, 46 (4), 1735−1741. (15) Belitz, H. D.; Grosch, W.; Schieberle, P. Lehrbuch der Lebensmittelchemie, 5th ed.; Springer Verlag: Berlin, 2001. (16) Töpel, A. Chemie und Physik der Milch., 1st ed.; B.Behŕs Verlag: Hamburg, 1994. (17) Helland, I. B.; Saarem, K.; Saugstad, O. D.; Drevon, C. A. Fatty acid composition in maternal milk and plasma during supplementation with cod liver oil. Eur. J. Clin. Nutr. 1998, 52, 839−845. (18) Sprunger, L. M.; Acree, W. E.; Abraham, M. H. Mathematical correlations for gas-to-olive oil, gas-to-saline solution, and saline solution-to-olive oil partition coefficients based on the Goss modified Abraham model. QSAR Comb. Sci. 2008, 27 (7), 890−900. (19) Oemisch, L.; Endo, S.; Goss, K.-U. Unpublished work. (20) Endo, M.; Goss, K.-U. Serum albumin binding of structurally diverse neutral organic compounds: Data and models. Chem. Res. Toxicol. 2011, 24 (12), 2293−2301. (21) Poole, C. F.; Poole, S. K. Separation characteristics of wallcoated open-tubular columns for gas chromatography. J. Chromatogr., A 2008, 1184 (1−2), 254−280. (22) Abraham, M. H., New solute descriptors for linear free energy relationships and quantitative structure-activity relationships. In Quantitative Treatments of Solute/solvent Interactions; Politzer, P., Murray, J. S., Eds. Elsevier: Amstderdam, 1994; Vol. 1, pp 83−134. (23) Acree, W. E.; Abraham, M. H. Solubility predictions for crystalline polycyclic aromatic hydrocarbons (PAHs) dissolved in organic solvents based upon the Abraham general solvation model. Fluid Phase Equilib. 2002, 201 (2), 245−258. (24) Bronner, G.; Goss, K.-U. Predicting sorption of pesticides and other multifunctional organic chemicals to soil organic carbon. Environ. Sci. Technol. 2011, 45, 1313−1319. (25) Abraham, M. H.; Ibrahim, A. Air to fat and blood to fat distribution of volatile organic compounds and drugs: Linear free energy analyses. Eur. J. Med. Chem. 2006, 41, 1430−1438. (26) Mintz, C.; Ladlie, T.; Burton, K.; Clark, M.; Acree, W. E.; Abraham, M. H. Enthalpy of solvation correlations for gaseous solutes dissolved in alcohol solvents based on the Abraham model. QSAR Comb. Sci. 2008, 27 (5), 627−635. (27) Mintz, C.; Gibbs, J.; Acree, W. E.; Abraham, M. H. Enthalpy of solvation correlations for organic solutes and gases dissolved in acetonitrile and acetone. Thermochim. Acta 2009, 484 (1−2), 65−69. (28) Mintz, C.; Burton, K.; Ladlie, T.; Clark, M.; Acree, W. E.; Abraham, M. H. Enthalpy of solvation correlations for gaseous solutes dissolved in dibutyl ether and ethyl acetate. Thermochim. Acta 2008, 470 (1−2), 67−76.
9524
dx.doi.org/10.1021/es301921w | Environ. Sci. Technol. 2012, 46, 9519−9524
