Environ. Sci. Technol. 2010, 44, 1204–1210
Particle-Scale Measurement of PAH Aqueous Equilibrium Partitioning in Impacted Sediments U P A L G H O S H * ,† A N D STEVEN B. HAWTHORNE‡ Department of Civil and Environmental Engineering, University of Maryland, Baltimore County, Baltimore, Maryland and Energy and Environmental Research Center, University of North Dakota, Grand Forks, North Dakota
Received July 23, 2009. Revised manuscript received January 6, 2010. Accepted January 11, 2010.
This research investigated the particle-scale processes that control aqueous equilibrium partitioning of PAHs in manufactured gas plant (MGP) site sediments. Dominant particle types in impacted sediments (sand, wood, coal/coke, and pitch) were physically separated under a microscope for equilibrium assessments. Solid-phase microextraction (SPME) combined with selected ion monitoring GC/MS and perdeuterated PAH internal standards were used to determine freely dissolved PAH concentrations in small (0.1-1 mL) water samples at concentrations as low as µg/L (for lower molecular weight PAHs) to ng/L (for higher molecular weight PAHs). For every particle class the initial release of PAHs into the aqueous phase was rapid, and an apparent equilibrium was reached in a matter of days. The average ratio of aqueous total PAH concentration for pitch vs coal/coke particles for eight sediment samples was 20. Thus, sediments that had aged in the field for many decades were not at equilibrium and were still going through a slow process of contaminant mass transfer between the different particle types. A possible consequence of this slow aging process is further lowering of the activity of the chemical as mass transfer is achieved to new sorption sites with time. This study also found that the presence of black carbon even at the level of 1/3 of sediment organic carbon does not necessarily imply a BC-dominated sorption behavior, rather source pitch particles if present may dominate PAH partitioning. To our knowledge this is the first report of equilibrium partitioning assessment conducted at the sediment particle scale.
Introduction Impacted sediments from industrial and harbor sites often contain high levels of anthropogenic carbon such as coke, fossil fuel-derived soot, and coal tar pitch (1-5). Some of these anthropogenic carbon sources like pitch may also serve as the carriers of contaminants such as polycyclic aromatic hydrocarbons (PAHs) to the sediments. Even nonindustrial, but urban area sediments are continually impacted by PAHs associated with anthropogenic carbon from road runoff. Recent work by DiBlasi et al. (6) demonstrated that much of * Corresponding author phone: 410-455-8665; e-mail: ughosh@ umbc.edu. † University of Maryland. ‡ University of North Dakota. 1204
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 4, 2010
the organic carbon present in stormwater suspended solids may be comprised of traffic soot laden with combustionderived PAHs. Sediment risk assessment for PAHs needs to take into account the appropriate organic carbon phase that may be relevant for the particular impacted sediment. Before the extensive availability of natural gas, manufactured gas from coal, coke, and oil served as the major gaseous fuel for urban heating, cooking, and lighting in the U.S. The manufactured gas process from coal resulted in coal tar and pitch as the primary byproducts that contain high levels of PAHs. In addition to producing gas, tar, and pitch, the process of transforming coal also resulted in the release of a wide range of black carbon products such as coke and soot into the environment. Often the PAHs released from the coal or oil-based manufactured gas operations are associated with the black carbon byproducts. Hong et al. (7) characterized several lampblack-impacted soil samples from California and found that when PAHs are associated with lampblack carbon, aqueous availability of the PAHs is greatly reduced. For coal coking impacted sediments from Milwaukee Harbor, Ghosh et al. (3) found that the majority of the PAHs in sediments were strongly bound to coal and coalderived coke particles in sediments, and their bioavailability was strongly reduced (8). Exposure of sediment-bound contaminants to higher trophic level organisms through the water column or through diet can be affected by the nature of organic carbon in sediments. Organic geochemists have developed several intensive extraction techniques to isolate and quantify black carbon in soils and sediments. Reviews of these black carbon isolation techniques are provided in Nguyen et al. (9) and Hammes et al. (10). The commonly used method to isolate black carbon is the destruction of all other forms of carbon. Inorganic carbon is removed by the addition of HCl and natural organic matter (nonblack carbon) is removed by oxidation at 375 °C in the presence of air for 24 h (2). These treatments leave an inorganic sediment residue that has black carbon as the only carbon form. Some researchers have removed the inorganic sediment components by the treatment of the sediment with hydrofluoric acid followed by other wet extraction procedures to isolate black carbon (11). The residue can then be analyzed to measure the fraction of black carbon present. Sorption experiments can also be carried out with the residual sediment matrix containing black carbon. However, these intensive extraction procedures result in the loss of native PAHs and sorption studies can only be performed with freshly spiked PAHs. Also, the residual black carbon obtained can comprise several carbon forms including soot, charcoal, and coke, some of which can be greatly altered in the extraction procedure. Ghosh et al. (3) and Rockne et al. (12) used physical separation techniques to isolate the lighter density organic particulate fractions from the heavier density mineral fractions in sediments. Although the density separation technique itself does not allow isolation of the individual black carbon fraction in sediments, it provides a method to isolate carbon particulate fractions across several size/density ranges and makes it easier to perform visual analysis and quantitative assessment of the composition of organic carbon. As will be shown subsequently in the paper, the technique also allows assessment of whether PAHs may be associated with submicrometer sized soot particles or much larger particles of coal, coke, wood, charcoal, and pitch. Goals of the present study include determining the first directly measured PAH partitioning coefficients from several types of sediment particles, investigating whether such 10.1021/es902215p
2010 American Chemical Society
Published on Web 01/25/2010
particles are at equilibrium in the bulk sediment, and determining which particle types may dominate PAH partitioning in the bulk sediment. Physical separation of particle types was carried out by sieving, density separation, and individual particle identification and separation under an optical microscope. A solid phase microextraction (SPME) method capable of achieving low ng/L sensitivites with small (0.1-1 mL) water samples was used to determine freely dissolved PAH concentrations and probe equilibrium partitioning behavior of isolated particle types in MGP-impacted sediments. The equilibrium partitioning values measured for the different particle types were then used to provide a better understanding of PAH sorption behavior observed for a large number of sediment samples obtained from MGP impacted sediment sites.
Materials and Methods Sediment Sites and Collection. Sediments used in this study were part of a large set of 114 sediment samples collected in 2003 from the vicinity of former manufactured gas plant or aluminum smelting operations. Further description of the 114 samples and previous work with these samples are presented elsewhere (13). Seven of the sediment samples used in this study covered a large range of PAH concentrations and sediment TOC values and came from three locations along the Hudson River: two samples from Troy, NY (TR-12 and TR-15), three samples from Hudson, NY (HD-3, HD-5, and HD-6), and two samples from a city located on the lower Hudson River (NY-5 and NY-18). One additional sample was collected in 1999 from a freshwater harbor site located near a former manufactured gas plant in Harbor Point, Utica, NY (HP-2) and is described in more detail elsewhere (14). All sediment samples were mixed, stored in glass bottles with Teflon-lined caps, and kept refrigerated at 4 °C in the laboratory. A description of PAH concentrations, total organic carbon (TOC), black carbon (BC), and other geochemical characteristics of the sediment samples is presented in Supporting Information (SI) Table S1. Size and Density Separation. Wet sieving was performed to separate the sediment samples into four size fractions: 1000 µm. These particle sizes were further separated into heavy and light density particles using a cesium chloride solution of specific gravity 1.8. The size and density separation technique is described in Ghosh et al. (14). Total Organic Carbon (TOC) and BC Analysis. Sediment TOC analysis was performed using a Shimadzu TOC analyzer with a solid sample module (TOC-5000A and SSM-5000A). BC measurement is based on the 375 °C oxidation of non-BC materials followed by TOC measurement (2). The 375 °C oxidation method is widely used for BC determination, but has also been demonstrated to depend on BC type (9). Particle-Scale Measurement of PAH Partitioning. To investigate differences in PAH abundances and aqueous partitioning for the different particle types in the sediment samples, a quantity of lighter density particles in the 250-1000 µm size fraction was manually separated under the microscope into four particle types: wood, coal/coke, sand, coaltar pitch. As observed earlier (3), the sand particles typically contain surface deposits of natural organic matter and other fine grained minerals. Several particles (0.2 mg for pitch to 20 mg for sand) of each class were placed in 2 mL silanized glass vials with Teflon-lined caps and shipped to the University of North Dakota for PAH concentration and particle-scale equilibrium measurements. Initial kinetic studies were carried out using archived sediments (HP-2). Milligram quantities of separated particles were contacted with water in 2 mL autosampler silanized glass vials as illustrated in Figure 1. For the kinetics experiment, 1.8 mL of water was mixed with the particles in a 2 mL silanized
FIGURE 1. Light microscopy images of the four most abundant and relevant particle types present in MGP sediments that were used for particle-scale equilibrium measurements. The isolated particles of each type were placed in 2 mL glass vials and equilibrated with deionized water containing sodium azide. glass vial and placed in a rotator. At the specified times (over 68 days), about 100 µL of water was removed, diluted, and analyzed (with d-PAH internal standards for each ring size) with SPME (15). The number of particles placed in each vial ranged from 10 to 30 (a few mg). In this method, some depletion of the “fast” fraction for low KD PAHs from lower concentration samples may take place during the equilibration. An analysis of whether this may have happened is provided in the Results and Discussion section. After the exposure to water was completed, the water was removed from the particles, and the particles were extracted with the aid of sonication for 18 h with 2 mL of 50:50 acetone/ methylene chloride containing a mix of d-PAH internal standards (16). Extracts were analyzed as previously described (16). The equilibrium studies with eight sediment particle isolates were contacted for a period of 14 days. PAH extraction from sediment samples, cleanup, and analysis procedures are described elsewhere (13). SPME Analysis. The SPME extraction and analysis process is described in detail in Hawthorne et al. (15) and summarized here. A commercially available 7 µm film thickness polydimethylsiloxane (PDMS)-coated fused silica fiber was used to extract water samples for 30 min. The target PAHs and d-PAH internal standards adsorb to the nonpolar PDMS phase at equivalent rates. The use of the d-PAHs to quantitate the target PAHs compensates for variations in equilibrium partitioning and kinetics. Following the sorption period, the SPME fiber was immediately desorbed to a GC/MS injection port in the splitless mode at 320 °C for 5 min. It should be noted that the SPME method used to measure dissolved PAH concentrations can determine “total-dissolved” concentrations (PAHs freely dissolved in water plus those associated with the dissolved organic matter, “DOM”) or “freelydissolved” concentrations (dissolved PAHs not associated with the DOM) (15). This distinction is important for calculating KD values, since more than 90% of the higher molecular weight PAH molecules can be associated with the DOM phase in sediment pore waters (15). In the present study, both total- and freely dissolved concentrations were determined for each sample, and were found to be the same within the reproducibility of the SPME method, most likely because the procedures used to separate the single particles effectively eliminated DOM from the samples.
Results and Discussion Particle-Scale PAHs. Eight sediments from MGP sites were used in the present study, although the results should extend to other sites (such as aluminum production) which use MGPproduced materials such as pitch in their processes. Results of the separation of particle types from the different sediment samples were presented in an earlier publication (17). The VOL. 44, NO. 4, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1205
TABLE 1. PAH Concentration in Separated Particle Classes from Harbor Point Sediment (µg/g)
naphthalene 2-methylnaphthalene 1-methylnaphthalene acenaphthylene acenaphthene fluorene phenanthrene anthracene fluoranthene pyrene benz[a]anthracene chrysene benzo[b+k]fluoranthene benzo[e]pyrene benzo[a] pyrene perylene indeno[1,2,3-cd]pyrene benzo[ghi]perylene total
sand
pitch
coal/coke
wood
0.06 0.04 0.02 0.06 0.03 0.90 0.55 0.05 0.49 0.48 0.21 0.47 0.35 0.16 0.09 0.03 0.05 0.08 4.12
130 75 137 379 280 270 1,870 371 1,306 1,840 1,007 1,319 1,184 622 969 172 403 551 12,887
74 68 63 29 66 52 392 100 415 459 331 385 583 233 370 77 178 209 4,084
3 5 10 38 36 33 382 71 230 337 116 154 105 56 64 10 28 32 1,709
majority of the mass (50-70%) of the light density sediment fraction was contributed by the wood particles. Although pitch particles comprise a smaller fraction of the light density particulate mass, they contributed the majority (80-90%) of the PAHs. Thus, pitch particles in these MGP sediments appear to be the primary carriers of the PAHs although significant levels of other types of carbon (e.g., coal and coke) are also present. Coal and coke show strong sorptive properties for PAHs, especially in experiments with ground coal and coke; however, the mass transfer rates into several hundred micrometer sized coal/coke particles may be very slow as demonstrated earlier by Ghosh et al. (16) and Ahn et al. (18). Sediment particles in the 250-1000 µm size class were chosen for particle-scale equilibrium analysis. In a majority of the sediment samples, this size class contributed to the maximum amount of PAHs averaging 40% of the total (17). Results of PAH analysis of separated particles are shown in Table 1. Sand has the smallest PAH concentration and coal tar pitch had the highest concentration (about 3000-fold higher than sand). It should be noted that the particle separation procedures require exposure to water, which raises the concern that PAH losses could occur based on partitioning to the water, especially for the more soluble lower molecular weight PAHs like naphthalene. However, the mass ratio of water to sediment used in the sieving and density separation steps is only 20:1. Since KD values range from 103 to 107 (as discussed later), the maximum loss of any PAH would be
