Environ. Sci. Technol. 2008, 42, 3989–3995
Absorption or Adsorption? Insights from Molecular Probes n-Alkanes and Cycloalkanes into Modes of Sorption by Environmental Solid Matrices S A T O S H I E N D O , * ,† P E T E R G R A T H W O H L , † AND TORSTEN C. SCHMIDT‡ Center for Applied Geoscience (ZAG), Eberhard-Karls-University of Tübingen, Sigwartstrasse 10, D-72076 Tübingen, Germany, and Chair of Instrumental Analytical Chemistry, University of Duisburg-Essen, Lotharstrasse 1, D-47048 Duisburg, Germany
Received October 1, 2007. Revised manuscript received March 4, 2008. Accepted March 11, 2008.
Environmental solid matrices such as soils and aerosols contain a variety of absorbents (e.g., organic matter) and adsorbents (rigid carbonaceous geosorbents, minerals) but the contribution of both modes of sorption to the overall sorption behavior is often uncertain. Absorption of a cycloalkane from air to bulk phases is generally stronger than that of the n-alkane of the same number of carbon atoms, while adsorption onto surfaces does not differ between these two compounds, or rather favorsthe n-alkane.Thepresentstudyexploresthischaracteristic sorption behavior of alkanes and eventually claims that determination of n-alkane-to-cycloalkane sorption coefficient ratios (Kn/Kc) helps elucidate the mode of sorption by complex mixtures in the environment. Differences in sorption coefficients from air (K) between n- and cycloalkanes were explained based on the linear free energy relationship (LFER) models in the form log K ) - a V + b MR + constant, where V and MR are the molar volume and the molar refraction, respectively. The LFER models predict Kn/Kc< 1 for absorption and Kn/Kc ∼ 1 for adsorption. An extensive number of experimental K values of C5-C8 alkanes for known ab- and adsorbents were evaluated. The data matched the model expectations and indeed exhibited a distinct difference in Kn/Kc between aband adsorption. Steric factors due to sorbate and sorbent geometries generally favored adsorption of n-alkanes over that of cycloalkanes (Kn/Kc > 1), clarifying the contrast between the two sorption modes even more. The application to environmental solid matrices is demonstrated using sorption data for diesel soot, aerosols and snow. The results are in excellent agreement with previous discussions on the modes of sorption in these materials.
Introduction Whether a given material absorbs organic compounds in its bulk phase or adsorbs them onto its surface is a critical issue for the assessment of the environmental fate and behavior * Corresponding author phone: +49-7071-29-74693; fax: +49-70715059. † Eberhard-Karls-University of Tübingen. ‡ University of Duisburg-Essen. 10.1021/es702470g CCC: $40.75
Published on Web 05/01/2008
2008 American Chemical Society
of contaminants. Elucidating the prevailing mode of sorption by environmental phases is important as it indicates which phase component dominates the overall sorption behavior of sorbate compounds. In environmental solid matrices such as soils, sediments, and aerosols, organic contaminants absorb into organic matter (OM), while minerals and rigid carbonaceous geosorbents (CGs), e.g., coal, charcoal, soot, and coke, adsorb compounds. Knowing the primary mode of sorption as well as the material governing the overall sorption is essential in practical fields, since this information determines how concentrations should be expressed (e.g., per organic carbon content or per surface area) and which sorption models should be used for estimating sorption coefficients of various organic contaminants. Regarding nonpolar to weakly polar organic compounds, particular attention has been paid to relative importance of absorption to OM and adsorption to CGs. Until the early 1980s, sorption of these compounds by soils was considered principally as an absorption process in OM (1, 2). Later, frequent observations of nonlinear, extensive, and competitive sorption, which cannot be explained by the absorption model, and with the recognition of widespread occurrence of CGs in the environment led to the hypothesis that a combination from both absorption in OM and adsorption to CGs accounts for the overall sorption. On this issue, extensive review articles for soils and sediments are available (3, 4). Currently, it is generally accepted that adsorption to CGs present in environmental solid matrices can significantly contribute to the total sorption and can be a main cause of nonlinear and competitive sorption. However, for individual samples containing both OM and CGs, it is not easy to determine whether adsorption to CGs indeed affect the sorption. The reasons for this include (i) quantification of CGs are not unambiguous (ref 4 and references therein), (ii) CGs comprise a range of materials with varying sorption properties (5, 6), (iii) separation procedures of CGs from environmental samples can change properties of CGs (4, 7), (iv) OM also exhibits nonlinear and competitive sorption apart from that attributable to CGs (8), (v) in environmental media, coexisting compounds (including OM) may coat CG surfaces and thus change their adsorptive properties (9, 10). Hence, approaches that can provide independent evidence for adsorption to CGs without the need for extensive characterization of the sorbent sample may be useful to evaluate whether or not CGs play a significant role in the sorption of organic compounds. Previously, a multiparameter linear free energy relationship (LFER) approach using tens of probe compounds was demonstrated to suggest the mode of sorption (11, 12). Further, the use of perfluorinated compounds to probe for adsorption was proposed (13). These molecular probe approaches seem promising as they need neither separation nor quantification of CGs and can directly investigate original, untreated environmental materials. Adding to these, we propose an alternative probe approach that uses fewer and simpler compounds, namely a pair of nand cycloalkanes. Alkanes are often used to probe for sorption properties and surface or pore geometries of both natural and synthetic materials (14–19). Surface or pore geometries of sorbents are assessed by comparing sorption behavior of alkanes of various sizes and shapes. Data interpretation is relatively simple since alkanes, unlike aromatic or functionally substituted compounds, undergo only nonspecific van der Waals interactions with sorbing phases, and thus the influence of specific interactions can be completely excluded. VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3989
Among others, normal and cyclic alkanes are frequently compared with respect to their sorption behavior in a given system. It is known that a cycloalkane absorbs from air to a bulk phase to a larger extent than does the n-alkane of the same number of carbon atoms, while surface adsorption often rather favors n-alkanes (20, 21). Thus, it may be possible to differentiate the dominant mode of sorption by comparing the sorption behavior of n- and cycloalkanes. The objective of this paper is to test and apply this concept. First, LFER models are established for ab- and adsorption of n-alkanes and cycloalkanes from air. Using the LFER models, differences (or no differences) in sorption coefficients between n- and cycloalkanes are discussed in terms of cavity formation, molecular interactions and steric factors. Second, experimental sorption coefficients of the alkanes from air to single defined absorbents (liquids, rubbery polymers) and adsorbents (liquid, inorganic, and graphitized/activated carbon surfaces) are evaluated to examine whether the two sorption modes are indeed experimentally separable. Finally, implications obtainable from alkane sorption data for environmental solid matrices are discussed. It should be noted here that the current approach (i) requires careful interpretation of the results if applied to complex environmental mixtures, (ii) is best suited for estimating the sorption mode of relatively small, nonpolar compounds, (iii) cannot identify CG adsorption if OM also adsorbs, and (iv) will not be examined for adsorption to solid-water interfaces (or wet surfaces) in the present study. These points are detailed in the end of the paper.
Theory The equilibrium absorption coefficient Kabsorption of a solute between a bulk phase and air, and the equilibrium adsorption coefficient Kadsorption of an adsorbate between a surface and air are defined, respectively, as, Kabsorption ) Cbulk phase/Cair
(1)
Kabsorption ) Csurface/Cair
(2)
where Cbulk phase and Cair are volume-based equilibrium concentrations (e.g., mol/m3) of the solute/adsorbate in the bulk condensed phase and in air, respectively. Csurface is the adsorbed concentration per unit surface area. Hence, Kabsorption and Kadsorption have units of, e.g., m3-air/m3-bulk phase and m3-air/m2-suface, respectively. The sorption coefficients have the following relationships with the free energy changes of the respective processes, ∆Gabsorption ) - RTln Kabsorption + constant
(3)
∆Gabsorption ) - RTln Kabsorption + constant
(4)
where R is the gas constant and T is the absolute temperature (K). The constants in eqs 3 and 4 depend on the units of concentrations, the choice of the standard state, and the solvent molar volume (for eq 3), but not on the solute/ adsorbate-specific parameters at infinite dilution. In both sorption modes, the free energy changes are solely determined by molecular interactions occurring in/on the condensed phase, since molecular interactions in air are considered negligible. Following the cavity model used by others (22, 23), ∆Gabsorption can be described as, ∆Gabsorption ) ∆Gcavity + ∆Ginteraction
(5)
where ∆Gcavity is the cavity formation energy and ∆Ginteraction is the interaction energy. ∆Gcavity is the energy required to disrupt attractive interactions between phase-constituting molecules, while ∆Ginteraction arises from interactions between the solute molecule and the surrounding molecules. ∆Gcavity 3990
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
TABLE 1. Molar Refraction and Molar Volume of n- and Cycloalkanes of Five to Eight Carbon Atoms compound
molar refraction (MR)a [cm3/mol]
molar volume (V)b [cm3/mol]
n-pentane cyclopentane n-hexane cyclohexane n-heptane cycloheptane n-octane cyclooctane
17.77 17.33 21.74 21.76 25.75 26.18 29.74 30.79
81.31 70.45 95.40 84.54 109.49 98.63 123.58 112.70
a MR is back calculated by eq 13 in ref 25 using excess molar refraction data in refs 26, 27. b McGowan’s characteristic volume (24).
is proportional to the size of the cavity, and thus to the molecular size of the solute to be introduced into the phase. ∆Ginteraction is in our case equal to the nonspecific interaction energy, since alkanes cannot experience other interactions such as constant dipole–dipole interactions. ∆Ginteraction for nonspecific interactions is best approximated by the molar refraction (MR) of the solute (13). Together with eqs 3 and 5, one obtains an LFER equation, log Kabsorption ) - a V + b MR + constant
(6)
where McGowan’s molar volume V [cm3/mol] is used for the molecular size of the solute (24). Proportionality coefficients a and b include the scaling factors for the terms as well as quantitative descriptions of the corresponding properties of the bulk phase. That is, coefficient a describes the cohesive forces acting between absorbent molecules and b measures the capability of the bulk phase interacting with the solute through nonspecific interactions. Equations analogous to 5 and 6 can be set up for adsorption processes from air, but in this case the cavity term drops; thus, ∆Gadsorption ) ∆G interaction
(7)
log Kadsorption ) b MR + constant
(8)
To facilitate the comparison between n- and cycloalkanes, we compute the ratio of the sorption coefficient of n-alkane to that of cycloalkane of the same number of carbon atoms, Kn/Kc ) Kabsorption, n-alkane/Kabsorption, cycloalkane or Kadsorption, n-alkane/Kadsorption, cycloalkane (9) Note that this ratio is dimensionless irrespective of the unit of K and thus allows for direct comparison between different systems. Table 1 lists the MR and V values for n- and cycloalkanes containing five to eight carbon atoms. These data lead to two general expectations. First, the values of MR for n- and cycloalkanes of the same number of carbon atoms are only slightly different. Therefore, by eq 8, log Kadsorption values of n- and cycloalkane should not differ much (i.e., Kn/Kc ∼ 1). Second, molar volumes of cycloalkanes are always smaller than those of n-alkanes. Hence, eq 6 expects that log Kabsorption is larger for cycloalkanes than for n-alkanes (Kn/Kc < 1). Note that eqs 7 and 8 do not include molecular shape terms. Various steric factors exist which potentially cause deviations from simple LFER models, as previously pointed out by Goss (20). First, contact between adsorbate and adsorbent is known to significantly influence adsorption energies. A typical example is the adsorption of alkanes to a flat surface. It is sometimes stated that all carbon atoms of rod-shaped n-hexane can interact with a flat surface, unlike
nonplanar cyclohexane, which can only do so with three or four of its carbon atoms (17, 21). This leads to Kn/Kc > 1 for adsorption. Microporous materials (e.g., zeolites, activated carbons) can strongly interact with compounds whose molecular shapes fit exactly into the pores of the respective materials. For example, some zeolites favor slender n-alkanes over bulkier branched alkanes (28–30), whereas others prefer branched or cyclic alkanes to n-alkanes (14, 31, 32) because of their better contact within the pores. In addition, microporous materials may exhibit a size exclusion effect. Sorbates having dimensions larger than a certain cutoff size of material cannot enter the pores, which renders adsorption considerably weak (33, 34). In our case, dimensions of cycloalkanes are larger than those of n-alkanes, as exemplified in critical diameters of n-hexane (4.9 Å) and cyclohexane (6.9 Å) (35). Therefore, size exclusion generally leads to Kn/Kc > 1. In conclusion, according to the LFER models, absorption generally leads to Kn/Kc < 1. Adsorption should exhibit Kn/Kc ∼ 1 if steric differences are negligible. Significant steric factors cause Kn/Kc to be >1, except for those porous materials whose pores fit better to cycloalkanes. Note that the discussion above only applies to the sorption of compounds from air onto surface-air interfaces or into bulk phases. Sorption from water to surface-water interfaces should be considered as another process (36).
Data Evaluation Experimental Kabsorption and Kadsorption data of n-alkanes and cycloalkanes were collected from the literature, e.g., refs 16, 17, 21, 27–29, 32-34, 37–58. The complete data set and references are listed in the Supporting Information. The absorptive materials considered were liquid solvents and amorphous organic polymers. The adsorbents included liquid surfaces, inorganic materials (e.g., minerals, zeolites) and model CGs (graphitized and activated carbons). The considered Kabsorption and Kadsorption data were obtained under the infinite dilution condition, or equivalently, with the zerosurface coverage for surface adsorption. Data not considered are associated with irregular increases in log K along with CH2 increments of n-alkanes, possibly absorption-adsorption combined cases, or too small net retention times. Details for these exclusions are explained in the SI. Because of the large data availability, C6-alkanes will be mainly considered in the following discussions.
Results and Discussion Kn/Kc of Absorption and Adsorption Processes. Distributions of Kn/Kc values exhibit several clear trends (Figure 1, Table 2). First, absorption media (bulk liquids and organic polymers) considered here resulted in Kn/Kc values within a relatively narrow range, 0.13 (water) to 0.64 (poly(perfluoroalkyl ether)). In contrast, Kn/Kc values for surface adsorption scatter broadly, with an accumulation of data points between 1 and 5. Some zeolites and activated carbons showed extremely high Kn/Kc, likely because of size exclusion. Most notable is that, despite the large data set from many research groups covering a variety of materials, there is no overlapping of Kn/Kc values between absorption and adsorption, with just one exception (0.3 for zeolite 5A (32)). This evidence strongly supports the theoretical expectations that it is possible to distinguish between absorptive and absorptive mechanisms based on the Kn/Kc ratio. The experimental data in Figure 1 agree with the theoretical predictions stated earlier. Thus, C6-Kn/Kc values in absorption media were indeed less than 1, which follows eq 6. For adsorption, fairly dense data accumulation does exist around Kn/Kc equal to 1, as is expected by eq 8. In particular, all of the six liquid surfaces exhibited Kn/Kc values
FIGURE 1. Kn/Kc of C6-alkanes in/on various materials. Considered is absorption to 24 liquids (organic solvents, surfactants, ionic liquids, water) and 19 organic polymers as well as adsorption to six liquid–air interfaces, ca. 20 inorganic materials and ca. 20 graphitized/activated carbons. The numbers of materials depend on the way of counting, as data for a single material pretreated in different conditions or measured at multiple temperatures are all included (see Tables S1 to S4 in the Supporting Information). of nearly 1. The reason may be that molecules composing a liquid surface are mobile and can interact well with both n-hexane and cyclohexane, rendering the shape difference irrelevant. In contrast, many solid surfaces showed Kn/Kc considerably higher than 1. For these materials, steric factors favoring n-hexane over cyclohexane are evident. The only adsorbents to give Kn/Kc ratios below 1 were synthetic materials, namely zeolite 5A (32), Fe-exchanged zeolite NaX (29) and copper-hexacyanoferrate (53), which may have special surface or pore shapes that favor cyclohexane, and are rather irrelevant as such materials are not typically found in the environment. Furthermore, no carbon-containing adsorbent showed Kn/Kc lower than 1, indicating that separation between CG and OM is possible. Kn/Kc ratios for C7 and C8 alkanes tended to be slightly smaller than those for C5 and C6 alkanes in/on individual materials (Table 2). This trend is not readily explainable by eqs 6 and 8, possibly because MR is not a fully accurate descriptor for the interaction term. The differences in Kn/Kc, however, are small and do not lead to contradiction with the previous discussions that use MR. Nevertheless, this trend suggests that care should be taken when Kn/Kc values of alkanes with different numbers of carbon atoms are compared. Influence of Solvent Cohesiveness on Kn/Kc. Generally, Kn/Kc in polar solvents, especially in self-associated (bipolar) liquids such as glycols and water, was smaller than in nonpolar solvents (Table 2). This is explained by the fact that more cohesive liquids require more energy for creation of the cavity of a given size (i.e., a in eq 6 is larger), placing more importance on the differences in the molecular size for determining the value of Kabsorption. Additionally, it is consistent with this line of argumentation that poly(perfluoroalkyl ether) showed the highest Kn/Kc of the absorption materials considered. Perfluorinated polymers (and solvents) exhibit less mutual interactions than the other absorbents considered (13, 59), and thus the ∆Gcavity required for absorption is quite low. The size of the compound, then, is not as important for absorption as it is for the other solvents, and thus n- and cycloalkanes absorb to a more similar extent and the Kn/Kc ratio approaches 1. Sorption by Glassy Polymers. The organic polymers considered were all amorphous. The cited data were all derived from inverse gas chromatography (IGC) measurements at above glass transition temperatures (Tg) of polymers, i.e., polymers were in the “rubbery” state, except for poly(vinyltrimethylsilane) (PVTMS) (46). A particular interest VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3991
TABLE 2. Some Experimental Kn/Kc Values for Absorption to Bulk Liquids and Organic Polymers and for Adsorption to Liquid, Inorganic and Graphitized/Activated Carbon Surfacesa Kn/Kc temp [C°] n-hexadecane n-octacosane chloroform olive oil methanol 1-octanol ethylene glycol diethylene glycol water
25 100 25 37 25 25 25 25 25
poly(ethylethylene) polypropylene poly(methyl acrylate) poly(perfluoroalkyl ether) PVTMSb polyvinyl chloride
70 70 70 25 150 90
C5
C6
bulk liquids 0.55 0.56 0.46 0.55 0.39 0.16
0.58 0.57 0.41 0.49 0.42 0.37 0.27 0.30 0.13
organic polymers 0.48 0.49 0.49 0.50 0.34 0.37 0.64 0.55 0.44
C7
C8
0.43 0.44
0.37
0.15 0.22
0.23
0.049
0.043
0.36 0.38 0.25
0.30 0.32 0.19
ref 37 38 39 27 40 41 42 43 44 45 45 45 48 46 47
liquid surfaces ethylene glycol diethylene glycol water
25 25 25
1.13 1.10 1.09
0.85
0.86
42 43 49
inorganic surfaces alumina silica silica goethite zirconia chromia zeolite 13X zeolite 5A
250 60 104 104 104 130 250 250
1.14 1.27 0.97 1.62 1.34 5.91 1.57 73.1
graphite GTCBc activated carbon activated carbond activated carbone Saran carbonf
60 80 250 260 260 225
graphitized/activated carbons 3.82 3.26 4.52 1.55 12.2 2.72 401
0.97 0.87 1.51 1.28 1.15 44.5
4.41
0.83 1.60 1.36
28 50 16 16 16 21 28 28 54 55 28 58 58 57
a The complete lists are provided in the Supporting Information. b poly(vinyltrimethylsilane). c graphitized thermal carbon black. d olive stone, 10% burnoff. e olive stone, 22% burnoff. f carbonized at 1573 K.
regarding polymer sorption may be to elucidate differences between rubbery and glassy polymers, because soil OM has been proven to possess a glassy thermal property (60) and soil/sediment OM sorption is argued to mimic that of glassy polymers (8). Sorption by glassy polymers at low concentrations is said to be dominated by a “nanohole-filling” mechanism, which is considered as an adsorption process (8). In our data set, PVTMS was studied 20 °C below its Tg, and polyvinyl chloride just 10 °C above its Tg. However, their Kn/Kc values were just similar to those of rubbery polymers (Tables 2 and S2) and apparently suggested absorption. Kabsorption data for glassy polymers are scarce, since diffusion in glassy polymers is generally slow compared to usual gas flow velocities in IGC columns, which makes the determination of Kabsorption difficult. Further research on equilibrium sorption by glassy polymers may be thus warranted. Temperature Dependency of Kn/Kc. Temperature dependency of Kn/Kc has to be discussed since our collected Kn/Kc values were measured at various temperatures, many of which were above ambient temperatures. Kn/Kc for absorption is generally temperature-independent or slightly increases with increasing temperature (Figure 2 and Supporting Information Tables S1 and S2). In the eight polymers studied by Tian and Munk (45), the Kn/Kc ratios varied by no more than 0.03 from 60 to 110 °C. Therefore, Kn/Kc ratios for 3992
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
absorption can be assumed practically constant over the investigated temperature range. On the other hand, judging from the available data, Kn/Kc for adsorption appears not to change much or to decrease with increasing temperature. This may be common to most adsorbents, as it has been noted (61) that, especially for adsorptive processes, there is an empirical linear relationship between the logarithm of K and the corresponding enthalpy change (∆H) (also see SI for an additional discussion for this relationship). Thus, if Kn/Kc is approximately equal to 1, then ∆Hn-alkane is approximately equal to ∆Hcycloalkane, i.e., the temperature dependencies of Kn-alkane and Kcycloalkane are similar, and therefore, Kn/Kc is approximately 1 at all temperatures. If Kn/Kc is >1 at a temperature, then the temperature dependency for n-alkane is stronger than for cycloalkane. This means that, because K for adsorption from air to surfaces generally increases with decreasing temperature, Kn-alkane increases more than Kcycloalkane does when the temperature decreases, and thus the Kn/Kc is >1 at a lower temperature, too. Therefore, Kn/Kc >1 at above ambient temperatures can be taken as an indirect evidence for Kn/Kc >1 at ambient temperatures. Applications to Environmental Solid Matrices. Sorption of small alkanes by environmental matrices has been far less studied compared to the sorption of compounds of environmental concern. To our best knowledge, only the series
FIGURE 2. Temperature dependency of Kn/Kc values of C6-alkanes in/on various materials. Data sources are ref 58 (activated carbons 1 and 2), refs 54–56 (graphitized carbons), ref 52 (mercury sulfide), refs 50, 51 (silica), and ref 45 (poly(ethylethylene) and polyepichlorohydrin). Activated carbons 1 and 2 were generated from olive stones with 10 and 22% burnoff, respectively (58). of IGC studies by Roth et al. (11, 12, 62) allows us to obtain Kn/Kc for environmentally important phases. Roth et al. (11, 12) investigated sorption by diesel particulate matter (NIST SRM 2975) and three aerosol samples. From their data, Kn/Kc of C8-alkanes for the diesel particulate matter was calculated to 1.15, while that of the aerosols was 0.26–0.38 (Table S5 in SI). Comparing these values with the values for C8-alkanes in Table 2 indicates that a significant contribution of adsorption is evident to the total sorption by the diesel soot, and that absorption can completely explain the aerosol sorption. Roth et al. also discussed the mode of sorption by using a multiparameter LFER approach and concluded that adsorption by elemental carbon is dominant in the diesel particulate matter, and that absorption to OM is the prevailing process in two out of the three aerosols. For the other aerosol, the authors could not specify the governing mode. Since their approach is completely independent from ours, the agreement of the conclusions puts more confidence in both methods. Strictly, the Kn/Kc value of 1.15 for diesel particles does not rule out the possibility of a partial contribution of absorption, since Kn/ Kc on elemental carbon occurring in the diesel soot could be much higher than 1 due to steric effects. What is proven from this Kn/Kc ratio alone is that significant adsorption is present, since otherwise Kn/Kc > 1 cannot be realized. In general, experimental Kn/Kc ratios for complex mixtures need careful interpretation (see SI for further discussions). The mode of sorption by snow is also a subject of current research. Sorption of C8-alkanes by snow (62) exhibited Kn/ Kc of 0.45, which is too small for adsorption only, and is near the upper boundary of the literature values for absorption. This result supports the hypothesis that absorption into a quasi-liquid layer on ice is an important mechanism along with adsorption to the ice surface. Finally, some cautions and limitations for the Kn/Kc approach are summarized. First, care should be taken when sorption by a complex mixture is of concern as mentioned above. Second, the mode of sorption estimated by Kn/Kc is not directly applicable to other compounds, especially polar ones, because specific interactions may cause differences from alkanes. For instance, in the case where a mixture contains a polar surface and a nonpolar bulk phase, polar compounds may be adsorbed even if Kn/Kc indicates absorption. Yet, if Kn/Kc signals adsorption in this example, a majority of nonpolar and polar compounds are expectedly adsorbed as well. Perfluorinated compounds may also behave differently compared to alkanes because they generally prefer adsorbents to absorbents more than the other compounds do (13). Similarly, large molecules such as polycyclic com-
pounds could also show different behavior. Third, because of the diversity of environmental OM and CGs, there is the possibility that some OM contains adsorptive domains and some CGs absorptive domains. If this is the case, the Kn/Kc ratio cannot distinguish between OM and CGs. Forth, this study neither theoretically nor empirically assessed Kn/Kc for adsorption to solid-water interfaces. Surfaces of a solid in water are completely covered with water molecules which may compete at adsorption sites and/or interact with adsorbing compounds. Water molecules are also present on the surfaces exposed to air at nonzero humidities. However, water molecules could be unimportant if nonpolar surfaces are of concern, as nonpolar surfaces generally bind water molecules just weakly. This is probably the case for adsorption to CGs. Kn/Kc ratios for adsorption to CG-water interfaces are currently being investigated in our group. For absorption to water-containing phases (or even pure water), the proposed theory can be applied.
Acknowledgments We acknowledge fruitful discussions with Georg Jansen, University of Duisburg-Essen. Kai-Uwe Goss, Yasufumi Miyake, and Stefan B. Haderlein are thanked for valuable comments on our manuscript. The comments by three anonymous reviewers significantly improved our manuscript. This work was partly supported by the European Union FP6 Integrated Project Aqua Terra (Project no. GOCE 505428).
Supporting Information Available Supplemental descriptions for data evaluation, Tables S1–S5 for experimental Kn/Kc values, relationships between log Kadsorption and ∆H, and additional discussions on interpretation of Kn/Kc for mixtures. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Chiou, C. T.; Peters, L. J.; Freed, V. H. A physical concept of soil-water equilibriums for nonionic organic compounds. Science 1979, 206, 831–832. (2) Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Sorption of hydrophobic pollutants on natural sediments. Water Res. 1979, 13, 241–248. (3) Allen-King, R. M.; Grathwohl, P.; Ball, W. P. New modeling paradigms for the sorption of hydrophobic organic chemicals to heterogeneous carbonaceous matter in soils, sediments, and rocks. Adv. Water Resour. 2002, 25, 985–1016. (4) Cornelissen, G.; Gustafsson, O.; Bucheli, T. D.; Jonker, M. T. O.; Koelmans, A. A.; van Noort, P. C. M. Extensive sorption of organic compounds to black carbon, coal, and kerogen in sediments and soils: Mechanisms and consequences for distribution, VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3993
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
3994
bioaccumulation, and biodegradation. Environ. Sci. Technol. 2005, 39, 6881–6895. Grathwohl, P. Influence of organic matter from soils and sediments from various origins on the sorption of some chlorinated aliphatic hydrocarbons: Implications on Koc correlations. Environ. Sci. Technol. 1990, 24, 1687–1693. Kleineidam, S.; Schueth, C.; Grathwohl, P. Solubility-normalized combined adsorption-partitioning sorption isotherms for organic pollutants. Environ. Sci. Technol. 2002, 36, 4689–4697. Jeong, S.; Werth, C. J. Evaluation of methods to obtain geosorbent fractions enriched in carbonaceous materials that affect hydrophobic organic chemical sorption. Environ. Sci. Technol. 2005, 39, 3279–3288. Pignatello, J. J.; Lu, Y.; LeBoeuf, E. J.; Huang, W.; Song, J.; Xing, B. Nonlinear and competitive sorption of apolar compounds in black carbon-free natural organic materials. J. Environ. Qual. 2006, 35, 1049–1059. Pignatello, J. J.; Kwon, S.; Lu, Y. Effect of natural organic substances on the surface and adsorptive properties of environmental black carbon (char): Attenuation of surface activity by humic and fulvic acids. Environ. Sci. Technol. 2006, 40, 7757– 7763. Cornelissen, G.; Gustafsson, O. Effects of added PAHs and precipitated humic acid coatings on phenanthrene sorption to environmental black carbon. Environ. Pollut. 2006, 141, 526– 531. Roth, C. M.; Goss, K. U.; Schwarzenbach, R. P. Sorption of a diverse set of organic vapors to diesel soot and road tunnel aerosols. Environ. Sci. Technol. 2005, 39, 6632–6637. Roth, C. M.; Goss, K. U.; Schwarzenbach, R. P. Sorption of a diverse set of organic vapors to urban aerosols. Environ. Sci. Technol. 2005, 39, 6638–6643. Goss, K.-U.; Bronner, G. What is so special about the sorption behavior of highly fluorinated compounds. J. Phys. Chem. A 2006, 110, 9518–9522. Denayer, J. F. M.; Ocakoglu, A. R.; Martens, J. A.; Baron, G. V. Investigation of inverse shape selectivity in alkane adsorption on SAPO-5 zeolite using the tracer chromatography technique. J. Catal. 2004, 226, 240–244. Denayer, J. F. M.; Ocakoglu, R. A.; Thybaut, J.; Marin, G.; Jacobs, P.; Martens, J.; Baron, G. V. n- and isoalkane adsorption mechanisms on zeolite MCM-22. J. Phys. Chem. B 2006, 110, 8551–8558. Brendle, E.; Papirer, E. A new topological index for molecular probes used in inverse gas chromatography for the surface nanorugosity evaluation. 1. Method of evaluation. J. Colloid Interface Sci. 1997, 194, 207–216. Smiciklas, I. D.; Milonjic, S. K.; Zec, S. An inverse gas chromatographic study of the adsorption of alkanes on hydroxyapatite. J. Mater. Sci. 2000, 35, 2825–2828. Yampolskii, Y. P.; Soloviev, S. A.; Gringolts, M. L. Thermodynamics of sorption in and free volume of poly(5,6-bis(trimethylsilyl)norbornene). Polymer 2004, 45, 6945–6952. Dorris, G. M.; Gray, D. G. Adsorption of n-alkanes at zero surface coverage on cellulose paper and wood fibers. J. Colloid Interface Sci. 1980, 77, 353–362. Goss, K.-U. The air/surface adsorption equilibrium of organic compounds under ambient conditions. Crit. Rev. Environ. Sci. Technol. 2004, 34, 339–389. Onjia, A. E.; Milonjic, S. K.; Rajakovic, L. V. Inverse gas chromatography of chromia. Part I. Zero surface coverage. J. Serb. Chem. Soc. 2001, 66, 259–271. Abraham, M. H.; Chadha, H. S. Applications of a solvation equation to drug transport properties. In Lipophilicity in Drug Action and Toxicology; Pliska, V., Testa, B., van de Waterbeemd, H., Eds.; VCH Verlagsgesellschaft mbH: Weinheim, 1996; pp 311–337. Goss, K.-U.; Schwarzenbach, R. P. Linear free energy relationships used to evaluate equilibrium partitioning of organic compounds. Environ. Sci. Technol. 2001, 35, 1–9. Abraham, M. H.; McGowan, J. C. The use of characteristic volumes to measure cavity terms in reversed phase liquid chromatography. Chromatographia 1987, 23, 243–246. Abraham, M. H.; Ibrahim, A.; Zissimos, A. M. Determination of sets of solute descriptors from chromatographic measurements. J. Chromatogr. A 2004, 1037, 29–47. Abraham, M. H.; Chadha, H. S.; Whiting, G. S.; Mitchell, R. C. Hydrogen bonding. 32. An analysis of water-octanol and wateralkane partitioning and the DlogP parameter of seiler. J. Pharm. Sci. 1994, 83, 1085–1100.
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
(27) Abraham, M. H.; Ibrahim, A. Gas to olive oil partition coefficients: A linear free energy analysis. J. Chem. Inform. Model. 2006, 46, 1735–1741. (28) Diaz, E.; Ordonez, S.; Vega, A.; Coca, J. Adsorption characterisation of different volatile organic compounds over alumina, zeolites and activated carbon using inverse gas chromatography. J. Chromatogr., A 2004, 1049, 139–146. (29) Diaz, E.; Ordonez, S.; Vega, A.; Coca, J. Characterization of Co, Fe and Mn-exchanged zeolites by inverse gas chromatography. J. Chromatogr., A 2004, 1049, 161–169. (30) Denayer, J. F.; Souverijns, W.; Jacobs, P. A.; Martens, J. A.; Baron, G. V. High-temperature low-pressure adsorption of branched C5-C8 alkanes on zeolite beta, ZSM-5, ZSM-22, zeolite Y, and mordenite. J. Phys. Chem. B 1998, 102, 4588–4597. (31) Newalkar, B. L.; Jasra, R. V.; Kamath, V.; Bhat, S. G. T. Sorption of n-pentane, 2-methylbutane and cyclopentane in microporous AlPO4-5. Microporous Mesoporour Mater. 1997, 11, 195–205. (32) Bilgic, C.; Askin, A. Evaluation of the thermodynamic parameters for the adsorption of some hydrocarbons on alumina and molecular sieves 3A and 5A by inverse gas chromatography. J. Chromatogr., A 2003, 1006, 281–286. (33) Moreno-Castilla, C.; Fernandez-Morales, I.; Domingo-Garcia, M.; Lopez-Garzon, F. J. Carbon molecular sieves produced by the fixation of sulfur surface complexes. Chromatographia 1985, 20, 709–712. (34) Domingo-Garcia, M.; Fernandez-Morales, I.; Lopez-Garzon, F. J.; Moreno-Castilla, C.; Prados-Ramirez, M. J. The dynamic adsorption of several hydrocarbons on active carbons. J. Colloid Interface Sci. 1990, 136, 160–167. (35) Newalkar, B. L.; Jasra, R. V.; Kamath, V.; Bhat, S. G. T. Sorption of C6 alkanes in aluminophosphate molecular sieve, AlPO4-5. Adsorption 1999, 5, 345–357. (36) Goss, K.-U. Comment on ”Influence of soot carbon on the soilair partitioning of polycyclic aromatic hydrocarbon. Environ. Sci. Technol. 2004, 38, 1622–1623. (37) Weckwerth, J. D.; Carr, P. W.; Vitha, M. F.; Nasehzadeh, A. A comparison of gas-hexadecane and gas-apolane partition coefficients. Anal. Chem. 1998, 70, 3712–3716. (38) Du, Q.; Hattam, P.; Munk, P. Inverse gas chromatography. 7. Polymer-solvent interactions of hydrocarbon polymers. J. Chem. Eng. Data 1990, 35, 367–371. (39) Abraham, M. H.; Platts, J. A.; Hersey, A.; Leo, A. J.; Taft, R. W. Correlation and estimation of gas-chloroform and waterchloroform partition coefficients by a linear free energy relationship method. J. Pharm. Sci. 1999, 88, 670–679. (40) Abraham, M. H.; Whiting, G. S.; Carr, P. W.; Ouyang, H. Hydrogen bonding. Part 45. The solubility of gases and vapors in methanol at 298 K: An LFER analysis. J. Chem. Soc. Perkin Trans. 2 1998, 1385–1390. (41) PhysProp, Syracuse Research Corporation. http://www.syrres. com/. (42) Arancibia, E. L.; Catoggio, J. A. Gas-chromatographic study of solution and adsorption of hydrocarbons on glycols. II. Ethylene glycol. J. Chromatogr. 1982, 238, 281–290. (43) Arancibia, E. L.; Catoggio, J. A. Gas chromatographic study of solution and adsorption of hydrocarbons on glycols. I. Diethylene glycol and triethylene glycol. J. Chromatogr. 1980, 197, 135–145. (44) Plyasunov, A. V.; Shock, E. L. Thermodynamic functions of hydration of hydrocarbons at 298.15 K and 0.1 MPa. Geochim. Cosmochim. Acta 2000, 64, 439–468. (45) Tian, M.; Munk, P. Characterization of polymer-solvent interactions and their temperature dependence using inverse gas chromatography. J. Chem. Eng. Data 1994, 39, 742–755. (46) Yampolskii, Y. P.; Kaliuzhnyi, N. E.; Durgaryan, S. G. Thermodynamics of sorption in glassy poly(vinyltrimethylsilane). Macromolecules 1986, 19, 846–850. (47) Demertzis, P. G.; Riganakos, K. A.; Akrida-Demertzi, K.; Kontominas, M. G. Gas chromatographic studies on the compatibility of PVC with a migration resistant plasticizer. Angew. Makromol. Chem. 1991, 192, 81–91. (48) Castells, R. C.; Romero, L. M.; Nardillo, A. M. Thermodynamic consideration of the retention mechanism in a poly(perfluoroalkyl ether) gas chromatographic stationary phase used in packed columns. J. Chromatogr., A 1995, 715, 299–308. (49) Hoff, J. T.; Mackay, D.; Gillham, R.; Shiu, W. Y. Partitioning of organic chemicals at the air-water interface in environmental systems. Environ. Sci. Technol. 1993, 27, 2174–2180. (50) Markovic, M. M.; Kopecni, M. M.; Milonjic, S. K.; Ceranic, T. S. Thermodynamics of adsorption of organics on a cobalt-modified solid obtained from colloidal silica. J. Chromatogr. 1989, 463, 281– 295.
(51) Milonjic, S. K.; Zhigunova, L. K.; Pavasovic, V. L. Adsorption of organic liquids on alkaline earth-metal modified silica. Chromatographia 1988, 26, 324–328. (52) Djordjevic, N. M.; Kopecni, M. M.; Milonjic, S. K. Chromatographic and adsorptive properties of mercury sulfide. Chromatographia 1980, 13, 226–230. (53) Onjia, A. E.; Milonjic, S. K.; Todorovic, M.; Loos-Neskovic, C.; Fedoroff, M.; Jones, D. J. An inverse gas chromatography study of the adsorption of organics on nickel- and copper-hexacyanoferrates at zero surface coverage. J. Colloid Interface Sci. 2002, 251, 10–17. (54) Lopez-Garzon, F. J.; Fernandez-Morales, I.; Domingo-Garcia, M. A gas-solid chromatographic study of the adsorption of hydrocarbons and alcohols on graphite. Chromatographia 1987, 23, 97– 101. (55) Kalashnikova, E. V.; Kiselev, A. V.; Petrova, R. S.; Shcherbakova, K. D. Gas-chromatographic investigation of the adsorption equilibrium on graphitized thermal carbon black. I. Henry constants and heats of adsorption of C1-C6 hydrocarbons at zero coverage. Chromatographia 1971, 4, 495–500. (56) Elkington, P. A.; Curthoys, G. Heats of adsorption on carbon black surfaces. J. Phys. Chem. 1969, 73, 2321–2326.
(57) Fernandez-Morales, I.; Guerrero-Ruiz, A.; Lopez-Garzon, F. J.; Rodriguez-Ramos, I.; Moreno-Castilla, C. Adsorption capacity of Saran carbons at high temperatures and under dynamic conditions. Carbon 1984, 22, 301–304. (58) Lopez-Garzon, F. J.; Moreno-Castilla, C.; Guerrero-Ruiz, A.; Rodriguez-Reinoso, F.; Lopez-Gonzalez, J.d.D. High-temperature adsorption of hydrocarbons by activated carbons prepared from olive stones. Adsorpt. Sci. Technol. 1984, 1, 103–109. (59) Alentiev, A. Y.; Shantarovich, V. P.; Merkel, T. C.; Bondar, V. I.; Freeman, B. D.; Yampolskii, Y. P. Gas and vapor sorption, permeation, and diffusion in glassy amorphous Teflon AF1600. Macromolecules 2002, 35, 9513–9522. (60) Hurrass, J.; Schaumann, G. E. Is glassiness a common characteristic of soil organic matter. Environ. Sci. Technol. 2005, 39, 9534–9540. (61) Goss, K. U.; Schwarzenbach, R. P. Empirical prediction of heats of vaporization and heats of adsorption of organic compounds. Environ. Sci. Technol. 1999, 33, 3390–3393. (62) Roth, C. M.; Goss, K.-U.; Schwarzenbach, R. P. Sorption of diverse organic vapors to snow. Environ. Sci. Technol. 2004, 38, 4078–4084.
ES702470G
VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3995
