Environ. Sci. Technol. 2002, 36, 3725-3734
Sorption of Polycyclic Aromatic Hydrocarbons and Polychlorinated Biphenyls to Soot and Soot-like Materials in the Aqueous Environment: Mechanistic Considerations MICHIEL T. O. JONKER* AND ALBERT A. KOELMANS Aquatic Ecology and Water Quality Management Group, Department of Environmental Sciences, Wageningen University, P.O. Box 8080, 6700 DD Wageningen, The Netherlands
Recent studies have shown that sorption of polycyclic aromatic hydrocarbons (PAHs) in soot-water systems is exceptionally strong. As a consequence, soot may fully control the actual fate of PAHs in the aquatic environment. However, sorption has only been characterized for a limited number of PAHs to diesel soot, and the mechanism is poorly understood. In this paper, we present an extensive data set of sorbent-water distribution coefficients (KS, n ) 236) for a series of PAHs (both native and added) and polychlorinated biphenyls (PCBs) to five different types of soot and five soot-like materials. Both KS values and physicochemical properties of the sorbents show large variation. In general, sorption is very strong, with KS values up to 1010, showing the highest distribution coefficients on a mass basis ever reported. Sorption of in particular PAHs is often over 1000 times as strong as sorption to amorphous sedimentary organic carbon. The variation in KS values cannot be explained by “soot carbon fractions” or specific surface areas of the sorbents. Instead, values for native PAHs are mostly determined by the sorbates’ molar volume, and values for added PAHs and PCBs are determined by the sorbents’ average pore diameter. From differences in KS values between native and added PAH analogues, it can be deduced that generally more than 50% (with values up to 97%) of the native PAH concentration in soot is not available for distribution to the aqueous phase. We conclude that this is caused by physical entrapment of the chemicals within the solid matrix. Furthermore, most sorbents appear to preferentially sorb PCBs with planar configurations, a phenomenon most likely driven by sorption in molecularsized pores. Pore sorption is also concluded to be the most important sorption mechanism for added PAHs together with π-π interaction processes with flat aromatic sorbent surfaces. Frequently observed, slowly desorbing, resistant contaminant fractions in sediments may very well be explained on the basis of these results. * Corresponding author e-mail:
[email protected]; phone: +31 317 485485; fax: +31 317 484411. 10.1021/es020019x CCC: $22.00 Published on Web 07/25/2002
2002 American Chemical Society
Introduction Sorption to sediment is a key process in determining the actual fate and risk of hydrophobic organic chemicals (HOCs) in aquatic environments. It lowers aqueous concentrations and therefore reduces mobility, bioavailability, and chemical and biological degradation processes. Because of their hydrophobic nature, HOCs predominantly sorb to the hydrophobic regions that are present in sediments (1, 2). The volumetrically most important sedimentary hydrophobic domain is natural organic carbon, the degradation product of dead biomass. Therefore, sorption is commonly described as being a function of the organic carbon content in sediments. It has been assumed that HOCs show simple partitioning between the sediment organic carbon and the surrounding aqueous phase, which can be expressed in terms of an equilibrium partitioning coefficient (KOC). However, numerous studies have revealed that the application of this so-called organic carbon equilibrium partitioning model often leads to an overestimation of aqueous concentrations in field situations (e.g., refs 3-6) and, therefore, an overestimation of bioaccumulation (7) and toxic effects to aquatic organisms (7, 8). In other words, HOCs often show much stronger sorption to sediments than can be expected on the basis of hydrophobic interactions only. A general hypothesis that was brought up to explain this phenomenon is that contaminants slowly diffuse into mineral micropores or nanovoids located in “hard glassy” or “condensed” organic matter (9, 10). Here, the chemicals are being “sequestered”, which makes them more resistant to the desorption process (11). However, since a lot of resistant sorption data concern polycyclic aromatic hydrocarbons (PAHs) (3), also another more specific hypothesis was introduced. PAHs are mostly formed during incomplete combustion processes together with a condensed particulate carbon phase, which is commonly referred to as “soot”. Pyrogenic PAHs are suggested to be partially occluded in the soot matrix during this coproduction process (12, 13), to be trapped in micropores, or to have extremely high affinities for the aromatic flat surfaces of soot (3). This would also result in reduced partitioning of sedimentary PAHs to the aqueous phase if the chemicals have a pyrogenic origin. Indeed, it was recently demonstrated that sorption of both added (14) and natively bound PAHs (15) to soot is exceptionally strong as compared to regular hydrophobic partitioning into natural organic carbon. Another recent study has shown that not only PAHs but also other organic chemicals may possibly exhibit enhanced sorption to soot (6). In addition to PAHs, chlorobenzenes and mono-ortho-polychlorinated biphenyls (PCBs) in two sediment layers from Lake Ketelmeer, The Netherlands, were found to have KOC values that were up to 2-3 orders of magnitude higher than could be anticipated based on the compounds’ hydrophobicities. Mono-ortho-PCBs as opposed to multiple-ortho-PCBs appeared to sorb preferentially to these sediments that contained soot-like material. It was suggested that organic contaminants with a planar molecular structure in general might have high affinities for soot (6). This hypothesis in fact would explain several very high in situ KOC values for PCBs, dioxins, furans, and DDT derivates (16). Because of its presumed high sorption and sequestration capacity, in some cases soot may also be responsible for very slowly desorbing fractions of (planar) HOCs in sediments (e.g., in Lake Ketelmeer; 17). Similar fractions were recently observed for PAHs in coal (2, 18), a material that from a physicochemical point of view may be defined as “soot-like”. Finally, according to some researchers, soot-like materials such as coal and charcoal particles may account for nonlinear VOL. 36, NO. 17, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3725
sorption isotherms that are frequently observed for HOCs in sediments and soils (18-20). Since soot-water distribution coefficients have only been measured for diesel soot and a few chemicals (PAHs) (14, 15), the mechanism of sorption to soot is poorly understood. Real insight into this mechanism can only be gained by studying different types of soot-like materials and chemicals with different molecular structures, both comprising a wide range of physical and/or chemical properties. Therefore, in the present study, distribution coefficients were determined for a series of natively sorbed PAHs, added deuterated PAHs, and added PCBs with different degrees of molecular planarity to five different types of soot and five soot-like materials. The coefficients were measured with the previously developed equilibrium solid-phase extraction partitioning method using polyoxymethylene (POM-SPE), which was specifically designed for this type of determinations (15).
Materials and Methods Chemicals. Deuterated PAHs (phenanthrene-d10, anthracened10, fluoranthene-d10, benz[a]anthracene-d12, benzo[k]fluoranthene-d12, and benzo[g,h,i]perylene-d12) were obtained from Cambridge Isotope Laboratories (CIL), Andover, MA, and all had a purity of g98%. 2-Methylchrysene (purity 99.2%) was supplied by The Community Bureau of Reference (BCR), Geel, Belgium. PCBs used (IUPAC Nos. 18, 28, 52, 72, 77, 101, 118, 126, 138, 156, 169, and 209; see Table 2 for full names) all had a declared purity of g98% and were obtained from Promochem (Wesel, Germany) except for PCB-72, which was purchased from Ultra Scientific, North Kingstown, RI. Other chemicals used were as follows: methanol (HPLC gradient grade; Mallinckrodt Baker, Deventer, The Netherlands), acetonitrile (HPLC grade; Lab-Scan, Dublin, Ireland), hexane and acetone (picograde; Promochem), isooctane (for pesticide analysis; Acros, Geel, Belgium), toluene and benzene (HPLC grade; Sigma-Aldrich, Steinheim, Germany), 1-propanol (HPLC grade; Riedel-de Hae¨n, Seelze, Germany), calcium chloride (analytical grade; Merck, Darmstadt, Germany), sodium azide (99%; Sigma-Aldrich), acetanilide (99.9+%; Sigma-Aldrich), calcium carbonate (analytical grade; Merck), aluminum oxide-Super I (ICN Biomedicals, Eschwege, Germany), and silica gel 60 (70-230 mesh; Merck). Prior to use, silica gel was activated at 180 °C during 16 h, and aluminum oxide was deactivated with 10% (w/w) Nanopure water (Barnstead, Dubuque, IA). Other chemicals were used as received. Sorbents. Pulverized charcoal (from combustion of bark, particles
