Environ. Sci. Technol. 2005, 39, 7585-7591
Assessing Sequestration of Selected Polycyclic Aromatic Hydrocarbons by Use of Adsorption Modeling and Temperature-Programmed Desorption ABDUL ABU* AND STEVE SMITH Department of Life Sciences, King’s College London, London SE1 9NN, United Kingdom
Sequestration of phenanthrene and pyrene was investigated in two soilssa sandy soil designated SBS and a siltloam designated LHSsby combining long-term batch sorption studies with thermal desorption and pyrolysis of amended soil samples. The Polanyi-based adsorption volume and the adsorbed solute mass increased with aging for both soils, thus demonstrating the mechanism for observed sequestration. Despite rigorous thermal analysis, 30-62% (SBS sand) and 8-30% (LHS silt-loam) of phenanthrene could not be recovered after 30-270 days of sorption, with the increase in desorption resistance showing greater significance in SBS sand. For both soils, these values were 20-65% of adsorbed phenanthrene mass. Activation energies estimated from the temperature-programmed desorption (TPD) of sorbed phenanthrene at e375 °C were 51-53 kJ/mol, consistent with values derived for desorption of organic compounds from humic materials. The activated first-order model fitting of observed TPD data supports the conclusion that the desorption-resistant fraction of phenanthrene has become sequestered onto condensed organic domains and requires temperatures exceeding 600 °C to be released. The work demonstrates the use of thermal analysis in complementing the Polanyi-based adsorption modeling approach for assessing the mechanistic basis for sequestration of organic contaminants in soils.
Introduction Sequestration of hydrophobic organic contaminants (HOCs) in soil is a complex process (1-3). Current research has demonstrated the utility of several dual-mode models that incorporate linear partitioning in an amorphous or rubbery organic domain and nonlinear adsorption onto hardened or condensed organic carbon fractions in describing observed sorption phenomena (4-11). Sequestration or aging and slow desorption of HOCs are presumed to be largely influenced by retention within condensed microporous nonpolar domain of soils, which have been reported to comprise several altered organic materials such as kerogens, black carbon, and coal particles (9, 12-15). For polycyclic aromatic hydrocarbons (PAHs), distinguishing between the fraction partitioned into an amorphous organic matrix from that * Corresponding author phone: (44) 207-8484253; fax: (44) 2078484500; e-mail:
[email protected]. 10.1021/es050669b CCC: $30.25 Published on Web 08/23/2005
2005 American Chemical Society
adsorbed or sequestered onto condensed organic domains over long periods is of critical importance and has implications for fate and transport modeling, bioavailability, site remediation, and the risk assessment of these compounds. Many previous works have been useful in demonstrating the role of hardened organic carbon in the sorption of HOCs onto natural solids and in several observations of isotherm nonlinearity at low aqueous concentrations (14-16). However, most of these works have been limited to short-term equilibrium studies. If the risks to human health and the environment of persistent organic pollutants like PAHs are to be fully assessed, then the mechanisms that underlie their sequestration and/or aging in soils must be examined, as well as the importance of these phenomena in enhancing natural attenuation. In an effort to address some of these issues, sorption studies involving 14C-labeled phenanthrene and pyrene were combined with thermal desorption and pyrolysis gas chromatography-flame ionization detection (GCFID) and gas chromatography-mass spectrometry (GCMS) in order to evaluate sequestration of these PAHs in two soils, including one from a former coal gasification (MGP) facility. Several thermal techniques have been employed to analyze PAHs associated with soils. These include thermal desorption or pyrolysis and temperature-programmed desorption (TPD) (17-20). The thermal techniques used in this study were originally developed for characterizing petroleum fractions in geologic materials (21). Previous work has shown that the thermal desorption and pyrolysis methods can provide a means of characterizing PAH compounds in airborne particulate matter (19). The system, in tandem with GCFID or GCMS, can be applied to thermally desorb and/or pyrolyze soil samples or as TPD to determine the rate of evolution and hence the activation energy of sorbed components of soils. The aim of this study was to demonstrate the mechanistic basis of sequestration by applying a dualmode modeling approach with a Polanyi potential theorybased model (5) for describing the nonlinear adsorption component of PAH sorption over long periods. The hypothesis that thermal analytical methods such as TPD can be used to complement the dual-mode approach to HOC interactions with geosorbent matrices was also tested by evaluating the desorption profile and thermodynamics of phenanthrene sequestered onto soils. It is expected that these investigations would lead to a better understanding of the natural attenuation processes of persistent organic pollutants such as PAHs and the design of cost-effective remediation strategies.
Experimental Section Sorbents and Chemicals. Two different soils were used for these studies. A sandy soil designated SBS which was obtained from an uncontaminated area of a former MGP facility in the north of England, which has been the subject of previous research on bioremediation (22). The other soil was a siltloam designated LHS which was obtained from a site on the south coast of England that had no history of contamination with HOCs. This soil has also been well characterized and used as sorbent in previous biodegradation studies (23). Bulk SBS sand and LHS silt-loam were stored in sealed polythene bags at 4 °C. For all experiments, samples were ground, airdried, and passed through a 2 mm sieve. In addition to the soil properties listed in Table 1, pyrolysis GCMS was used to determine the range of organic compounds present in unspiked soil samples. Unlabeled phenanthrene and pyrene (>96% purity) and the radiolabeled chemicals [9-14C]VOL. 39, NO. 19, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7585
TABLE 1. Properties of Soils SBS sandy soil
LHS silty loam
0.24 2.67 1.33 12.25
1.50 12.40 10.28 10.49
TOC (%) CEC (mequiv/100 g) BET-N2 SSA (m2/g) C-N ratio
TABLE 2. Physicochemical Properties of Selected PAHs MW (g/mol) solubility (mg/L) log Kow subcooled liquid solubility Sc (mg/L) at 22 °C density (g/cm3) molar volume (cm3/mol) a Obtained from ref 37. physicochemical data.
b
phenanthrene
pyrene
178.2 1.29a 4.57 6.1c
202.3 0.132b 5.13 3.48c
1.02 174.7c
1.271 158.93c
Obtained from ref 29. c Estimated from
phenanthrene (12.4 mCi/mmol) and [4,5,9-14C3]pyrene (58.7 mCi/mmol), both with >98% purity, were obtained from Sigma-Aldrich Co. (Dorset, U.K.). Some physicochemical properties of the selected PAHs are presented in Table 2. Batch Sorption Experiments. Sorption experiments were conducted in glass Pyrex tubes fitted with silicone septa, which were also lined with poly(tetrafluoroethylene) (PTFE) tape, and cleaned according to standard procedures. Stock solutions of solutes were prepared in methanol according to procedures described by others (4). Radiolabeled phenanthrene and pyrene were added to stock solutions in predetermined proportions to the unlabeled analogue to enable analysis by liquid scintillation counting (LSC) of the amounts sorbed by soils. One gram of soil sample was mixed with 16 mL of PAH stock solution in glass tubes, which typically left a headspace of ∼0.5 mL. Stock solutions comprised distilled deionized water and CaCl2 (0.01 M) with HgCl2 (200 mg/L) added to inhibit biological activity. Mercuric chloride has been described as a highly effective chemosterilant for longterm studies involving HOCs (24). Samples were placed vertically on an incubator shaker operating in darkness at 24 °C and 120 rpm for 30, 60 or 270 days. Preliminary experiments with the soil samples showed that apparent sorption equilibrium was achieved in 20-30 days. Tubes containing stock solutions but no soil were included in all experiments to account for any losses to glass and/or cap assemblies. After various periods, tubes were removed from the incubator and centrifuged at 2000 rpm for 30 min, and the aqueous supernatant was decanted for analysis of solution-phase PAH concentration by LSC. Tubes were then refilled with 16 mL of solute-free background solution, shaken for 60 s on a vortex mixer, and placed again on the incubator shaker set to the same conditions as in the previous sorption phase for a 21-day desorption study. Additional information on apparent equilibrium times, determination of solid-phase PAH concentrations, and desorption kinetic studies have been provided in the Supporting Information as accompanying notes to Figure S1 and Tables S1-S3. Thermal Analysis. For thermal analysis experiments, radiolabeled phenanthrene and the unlabeled compound were added to 500 µL of dichloromethane (DCM) and spiked onto 10 g of soil in precleaned amber screw-capped bottles by use of a Hamilton syringe to give 100 µg of phenanthrene/g of soil. Bottles that contained amended soil samples were then vigorously shaken on a vortex mixer for 60 s every 15 min for 90 min (25) to facilitate distribution of phenanthrene within soil and evaporation of DCM. This was followed by 7586
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 19, 2005
the addition of 50 mL of background solution made up of CaCl2 (0.01 M) and HgCl2 (200 mg/L). The bottles were subsequently placed on an incubator shaker for 2, 60, or 270 days. Because of the larger volume of soil used, inhibition of microbial activity in all samples was confirmed by the absence of bacterial growth on nutrient agar inoculated with soil slurries following centrifugation and analysis of aqueous supernatants by LSC as described previously. Soil pellet remaining in bottles was then allowed to air-dry for 24 h at laboratory temperatures prior to analysis. GCFID thermal analysis of 100-mg dry samples of soils similarly treated after 0 days (2 h) of sorption resulted in 92-110% recovery of added phenanthrene. The analytical oven was set at 75 °C and then ramped to 300 or 600 °C in 20 s. The GC oven temperature was set at 40 °C for 13 min during desorption or pyrolysis, ramped to 300 °C, and held for 25 min. For TPD analysis, a 100-mg amended dry soil sample was weighed into a glass-lined stainless steel liner and introduced into the analytical oven, which was set at 75 °C, then ramped at 5 °C/min for 20 min to achieve a temperature of 175 °C and held for 10 min. Analytical oven temperature was then ramped to 375 °C at a rate of 10 °C/min and held for 20 min. The mass spectrometer was a Fisons MD800 run in electron impact (EI) mode and scanned from m/z 40 to 520 at 2 scans/ s. Desorbed phenanthrene was detected and monitored by the extraction of the specific ion chromatogram (m/z ) 178), which was matched with library scan data. A detailed description of the GCFID thermal analysis and adaptation of GCMS system for TPD analysis have been provided in the Supporting Information on thermal analytical methods. Analysis of Data. All batch sorption and desorption data were initially fitted with the Freundlich model (Supporting Information). The adsorption-partitioning model proposed by Xia and Ball (5, 6) was also used to fit all sorption and desorption data of the selected PAHs, with the subcooled liquid solubility (Sc) values used for calculating the adsorption potential of solutes. The relationship between the adsorption potential and the adsorbed solute volume is given in eq 1 as explained in ref 5 and references therein:
log (qe′) ) log (Qo′) + a′(sw/Vs)b′
(1)
where qe′ is the adsorbed solute volume per unit mass of sorbent (cubic centimeters per kilogram), Qo′ is the adsorption volume capacity at saturation (cubic centimeters per kilogram), sw is the effective adsorption potential (joules per mole) of the hydrophobic solute from aqueous solution, Vs is the bulk molar volume (cubic centimeters per mole) of the adsorbate at the temperature of adsorption, a′ is a fitted parameter, and the exponent b′ in our application was set to 2, which corresponds to a log-normal sorption energy distribution for organic molecules on heterogeneous surfaces as described in the work of Condon (26, 27). The partition coefficient (Kp) was estimated from the Koc-Kow relationship provided by Karickhoff (28). Here, Kp ) Kocfoc:
log Koc ) 0.99(log Kow) - 0.35
(2)
The effect of temperature on the thermodynamics of phenanthrene desorption from soils was approximated to an activated first-order rate process of the form described by the Arrhenius equation:
k ) Ae-Ea/RT
(3)
where A is described as the preexponential factor, Ea is the activation energy, R is the molar gas constant (joules per mole), and T is temperature (kelvins). By use of four replicates, the desorption rate constants and the preexponential factors were estimated from the normalized release rate of phenan-
FIGURE 1. Combined adsorption-partitioning isotherm with Polanyi-based model isotherm for the adsorption component of sorption of phenanthrene and pyrene: (b) fitted total sorption for LHS silt-loam; (O) fitted adsorption component; (2) fitted total sorption for SBS sand; (4) fitted adsorption component.
TABLE 3. Fitted Adsorption (Polanyi-Based) Model and Partitioning Parameter Values for Sorption of PAHs onto Soils soil
Qo′ (cm3/kg)
SBS sand (30 d) SBS sand (60 d) SBS sand (270 d) LHS loam (30 d) LHS loam (60 d) LHS loam (270 d) SBS sand (30 d) SBS sand (270 d) LHS loam (30 d) LHS loam (270 d)
Qoa (mg/g)
MWSEb
Kd (L/kg)
Kp (L/kg)
parking efficiencyc
0.033 ( 0.039 0.087 ( 0.028 0.295 ( 0.037 0.083 ( 0.032 0.123 ( 0.057 0.165 ( 0.046
Phenanthrene 0.034 ( 0.040 0.0939 0.089 ( 0.037 0.0445 0.301 ( 0.038 0.0398 0.085 ( 0.033 0.0549 0.125 ( 0.058 0.0797 0.168 ( 0.047 0.0589
94.09 265.44 703.77 393.28 592.93 799.27
35.85
0.234 0.617 2.092 0.589 0.872 1.170
0.055 ( 0.046 0.068 ( 0.045 0.035 ( 0.019 0.061 ( 0.051
0.060 ( 0.058 0.086 ( 0.057 0.044 ( 0.044 0.078 ( 0.065
224.10
Pyrene 0.1180 0.0768 0.0541 0.0898
279.62 691.50 1109.5 1416.5
128.50 803.14
0.390 0.482 0.248 0.433
a Adsorbed solute mass as distinguished from the adsorbed solute volume (Q ′) (5). b Mean weighted square error as defined in ref 5. c Parking o efficiency (ηp), which is the ratio of adsorbed volume of the solid compound to that of a reference liquid compound (5).
threne from soils over a given range of time and temperature of desorption following the approach described in refs 18 and 29. And by fitting these values into the integral form of eq 3, activation energies associated with phenanthrene desorption from the soils were estimated. The initial sorption data and GC thermal desorption/pyrolysis results were expressed as the mean ( standard deviation (SD) and compared between the two soils by use of a two-way t-test or subjected to analysis of variance (ANOVA) of the general linear model at the 95% confidence interval.
Results and Discussion Adsorption-Partitioning Modeling of Sorption Data. The combined sorption isotherms with the Polanyi-based adsorption isotherm for the adsorption component are illustrated in Figure 1. Fitted parameters for adsorption and partitioning of phenanthrene and pyrene are also summarized in Table 3. After 30 days of equilibrium, the adsorption capacity of soils ranged from 0.03 to 0.08 cm3/kg for phenanthrene and pyrene. Values ranging from 0.02 to 0.04 cm3/kg have been reported for 2-4-ring PAHs on an aquitard material (5) with TOC content similar to LHS siltloam. However, while adsorption capacity and the adsorbed
solute mass increased significantly for the sorption of phenanthrene onto SBS sand after 60 and 270 days, the observed increases were not significant in LHS silt-loam. Similar increases in pyrene adsorption parameters also did not exhibit any significance for both soils. The observation of increased adsorption capacity with aging of samples is qualitatively in agreement with results obtained for sorption of solid-phase adsorbates onto carbonaceous samples exhibiting a range of coalification (8). For this class of solid compounds (which includes PAHs), adsorption from aqueous solution has been observed to be less efficient compared to liquid-phase adsorbates, presumably because of structural incompatibility with the adsorbent pore geometry (5, 30). The parking efficiency (ηp) defined by Xia and Ball (5) was calculated for the adsorption of phenanthrene and pyrene, by use of the adsorption capacity of 0.141 cm3/kg reported for liquid chemicals by these authors. After 30 days of sorption, ηp values for phenanthrene were 0.23-0.59 in the two soils (Table 3). The parking efficiency of phenanthrene in both soils increased to 0.62-0.87 after 60 days, consistent with the average value of 0.77 reported for adsorption of solid compounds onto activated carbon (30). However, after 270 days of aging, ηp values of 1.20 and 2.09 were observed VOL. 39, NO. 19, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7587
TABLE 4. Estimated Rate Constants Associated with the Desorption of PAHs from Soils and Thermal Desorption Activation Energies aqueous desorption 30 d phenanthrene ka (min-1) % desorption % pyrolysis Ead (kJ/mol) pyrene ka (min-1) phenanthrene ak (min-1) % desorption % pyrolysis Ead (kJ/mol) pyrene ka (min-1)
60 d
thermal desorption 270 d
6.05 × 10-5 b
SBS Sand 1.98 × 10-5 5.70 × 10-6
2.93 × 10-5
nd
2.37 × 10-5 b
LHS Silt-Loam 1.08 × 10-5 5.70 × 10-6
4.37 × 10-6
nd
8.82 × 10-6
4.33 × 10-6
2d
60 d
270 d
1.00 × 10-5 65.54 ( 5.98 4.90 ( 1.78 51 nd
ndc 46.8 ( 6.04 2.43 ( 0.20 nd nd
1.00 × 10-5 35.57 ( 3.1 2.04 ( 0.30 51 nd
2.00 × 10-5 73.88 ( 4.44 9.13 ( 0.71 53 nd
nd 81.32 ( 8.8 10.94 ( 0.9 nd nd
2.00 × 10-5 57.93 ( 6.3 12.12 ( 0.6 53 nd
Desorption rate constant. Rate constant for fraction of fast-desorbing pool (f) in aqueous solutions (f > 0.98). Some notes on aqueous desorption are provided in the Supporting Information. c Not determined. d Activation energy corresponding to peak temperature of phenanthrene release from soil. a
b
for LHS silt-loam and SBS sand, respectively. Despite small increases in the parking efficiency of pyrene after 270 days, the ηp values of 0.25-0.48 were still much lower than the average reported value. After 270 days of aging, the adsorbed masses of phenanthrene in both soils were observed to be higher than those for pyrene, confirming the greater parking efficiency of this PAH in the soils studied. Lower adsorbed mass of larger HOC molecules compared to smaller ones for the same sorbent has been attributed to limited accessibility to microporous domains of organic matrices (4, 10, 31). Plausible explanations of these results are structural rearrangement of the soil organic matrix in addition to slow diffusion into microporous domains, mechanisms that were observed to be more pronounced with the adsorption of phenanthrene after sample aging. In earlier studies with charcoal particles, Braida et al. (32) proposed these as the likely mechanisms underlying benzene adsorption. The fractional contribution of adsorption was estimated by subtracting the partitioned fraction from the total sorption (Kd). Adsorption of phenanthrene increased from 62% to 95% in SBS sand and from 43% to 72% in LHS silt-loam over the 30-270-day period. Adsorption of pyrene during the same period accounted for 54-81% of total sorption by SBS sand and 28-43% by LHS silt-loam (Table 3). Adsorption of HOCs has also been reported to account for between 37% and 68% of total sorption at Ce/S ) 0.1-0.35 in an aquitard material (5) and in both peat and mineral soils (31). However, the partition contribution in the present study was expected to have increased at the elevated solute concentrations employed, and Kp values would likely have been underestimated. The results obtained from fitting the adsorption model to PAH desorption data (Table S4, Supporting Information) also show a similarly increasing trend in the adsorbed fraction when the 30- and 270-day samples are compared. However, the model parameter values for phenanthrene desorption decreased between 60 and 270 days. This observation could not be made for pyrene desorption since the 60-day study was not included. Nonetheless, the observations of others support the hypothesis that the observed decrease in adsorption capacity may be as a result of different desorption mechanisms related to the changing geometry of the soil organic matrix through expansion during sorption and/or collapse or constriction as desorption proceeds (32). Phenanthrene Recovery by Thermal Desorption/Pyrolysis. The proportion of added phenanthrene retained by both soils after the different periods of equilibration was >98%, representing ∼10 µg/100 mg of sample. Table 4 shows the percentage recoveries of sorbed phenanthrene by 7588
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 19, 2005
combined thermal desorption and pyrolysis. Total recoveries of sorbed phenanthrene from LHS silt-loam were significantly higher than from SBS sand. Both soils, however, showed a clear trend of decreasing phenanthrene recovery with aging. It should be noted that the contribution of pyrolysis at 600 °C to total phenanthrene recovery was 2-12%. Using several surface and subsurface materials with a range of organic matter contents, Schultz et al. (17) reported recoveries of 5-33% of spiked phenanthrene after a 30-s heating program at 300 °C and 26-111% after pyrolysis at 800 °C. The difference with the present work may be the 10-min heating program, which in comparison to the flash pyrolysis method resulted in release of most of the recoverable fraction of phenanthrene at 300 °C. Additional pyrolysis at 600 °C may not have provided a sufficiently high temperature to achieve recovery of the fraction of phenanthrene that had become sequestered onto soil components. It has been suggested that high pyrolysis temperature and long pyrolysis time result in recombination and rearrangement reactions that may increase the yields of PAH pyrolytic fragments from soils (17). To ensure that sorbed phenanthrene was not undergoing pyrolytic breakdown at temperatures g300 °C as a result of the long pyrolysis time, spiked soil samples were directly pyrolyzed at 600 °C for 10 min without first desorbing at 300 °C. The amounts of phenanthrene recovered were comparable to the amounts recovered when samples were thermally desorbed at 300 °C. Although a definitive molecular analysis or functional correlation could not be made during initial characterization of soil samples by pyrolysis GCMS, the organic matter of LHS silt-loam appears to be much more characterized by phenolic and aliphatic pyrolytic fragments of humic substances (Table S5, Supporting Information). Temperature-Programmed Desorption. To ascertain the effect of sequestration on the release of soil-sorbed PAH, the thermodynamics of phenanthrene desorption was investigated by approximating the rate of release to an activated first-order process. From this, Ea values corresponding to the temperature of desorption were estimated. Figure 2 illustrates the effect of temperature on phenanthrene release from both soils. The TPD-MS normalized rate of release confirms earlier results from thermal desorption/pyrolysis GCFID, where significantly more of the PAH was recovered from the silt-loam LHS than from the sandy SBS. Phenanthrene release was also found to be significantly higher in samples equilibrated for 2 days compared to those aged for 270 days. However, desorption rate constants estimated from the TPD of phenanthrene were found to be the same after 2 or 270 days of sorption (Figures S2 and S3, Supporting
FIGURE 2. Average TPD-MS response for the release of phenanthrene from soils: (b) 2-d LHS s-loam; (O) 270-d LHS s-loam; (2) 2-d SBS sand; (4) 270-d SBS sand. Information). Activation energies associated with the release of phenanthrene from soils have also been presented in Table 4. The Ea values correspond to the peak temperature of phenanthrene evolution from soils, estimated to be 260 °C for SBS sand and 315 °C for LHS silt-loam. However, at any given temperature of desorption, the Ea value for SBS sand is ∼3 °C higher than the value for LHS silt-loam. It is apparent from the TPD profile and the extracted ion chromatogram (EIC) illustrated in Figure 3 that both soil TPD-MS responses exhibit different chromatographic shifts. This can be attributed to thermodynamic effects on the organic matrix as well as differences in diffusivity through intraparticle pores of associated soil minerals (1, 3, 18, 33, 35). No differences were observed in the TPD response curves, onset and peak desorption temperatures, and Ea values within each of the soils when the 2- and 270-day sorption results were compared. Ghosh et al. (18) observed that, considering different properties of the sorbent matrix, a possible explanation for higher temperatures of the TPD response is the
much slower increase in diffusivity as temperature increases. The results obtained from control GCMS analysis involving standard phenanthrene concentrations in DCM spiked onto nonadsorbent glass wool showed that the time and temperature of phenanthrene evolution were 8 min and 115 °C, respectively. This observation supports the conclusion that the differential TPD response of phenanthrene in the two soils studied is due to differences in kinetic effects and diffusive characteristics of the organic matrix. Similar observations have been made with both standard and organic matter-bound nitrotoluene compounds (20). Because additional pyrolysis of samples at 600 °C resulted in the release of a further 2-12% of sorbed phenanthrene, it was reasonable to conclude that no major evolution of the remaining sequestered PAH would occur between the peak desorption temperature of 260 or 315 °C and the pyrolysis temperature of 600 °C. Using measured rates of phenanthrene release from soils at temperatures below 375 °C, Ea values corresponding to temperatures of 600 and 1000 °C were predicted to be 79-84 and 115-122 kJ/mol, respectively. The Ea values estimated from observed TPD data (51-53 kJ/mol) are consistent with the Ea values derived for desorption of HOCs from amorphous organic matrices and aquifer materials by others (18, 34, 35). Generally Ea values greater than 60 kJ/mol have usually been associated with diffusion of HOCs out of condensed organic matter, coal materials, and glassy polymers (18, 33, 36). This would support the assumption that predicted Ea values at 600 °C are associated with the release of 2-12% of sorbed phenanthrene from condensed organic matrices of the soils studied. The Ea values predicted from the TPD data suggest that it would require temperatures exceeding 600 °C to achieve evolution of the remaining fraction of sorbed phenanthrene, presumably sequestered onto the condensed organic carbon domain of both soils.
FIGURE 3. GCMS chromatogram (or EIC) of representative TPD of phenanthrene (m/z ) 178) released from soils. (A) SBS sand; (B) LHS silt-loam. VOL. 39, NO. 19, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7589
Relating Sequestration to Thermal Desorption Profiles. The model fitting for describing the adsorption component shows that solute pore-filling within condensed nonpolar domains continued for long periods and that the concept of a maximum adsorption capacity for such heterogeneous sorbents may lack a mechanistic basis. This is especially true in the context of the relatively high concentrations of organic pollutants encountered at contaminated sites. Thus, many previous works where the dual-mode modeling approach has been applied to describe sorption data during shortterm equilibrium studies (7 days) have failed to demonstrate the true extent of the adsorption capacity of heterogeneous sorbents. The slow process of attaining adsorption saturation should be the mechanistic basis for observed sequestration or aging of the selected PAHs in the soils examined, as well as accounting for the observed limited desorption effects. Given the association of PAH sequestration with slow adsorption onto condensed organic matrices, thermal analysis offered a complementary means of distinguishing between the fraction sequestered onto condensed carbonaceous domains from the fraction partitioned within amorphous humic materials. The lower amounts of phenanthrene recovered from SBS sand suggest that sequestration of this PAH was much greater than in the silt-loam LHS, an observation that is consistent with higher adsorption contribution for this sandy soil. For both soils, the activated first-order model fitting of the TPD data supports the conclusion that the fraction of sorbed phenanthrene unrecovered by thermal analysis has become sequestered onto condensed organic matrices and would require temperatures exceeding 600 °C to be released. This fraction was estimated to be about 20-65% of the total mass of phenanthrene adsorbed during the 30-270 days of sorption. Last, TPD analyses of soil samples have been used for the first time to complement the dual-mode modeling. This approach should be of significant importance for the risk assessment of contaminated soils as well as the setting of appropriate remediation goals.
Acknowledgments This work was part of a Ph.D. research at King’s College London supported by the World Bank (JJWBGSP). Many thanks to Dr. Dan Waterman and Billy Kang for their assistance with the GCMS/FID analyses.
Supporting Information Available Freundlich sorption-desorption isotherms, fitted adsorption/partitioning parameter values for PAH desorption, percentage peak areas of pyrolytic fragments of soil organic matter, and estimated thermal desorption rate constants. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) ) Pignatello, J. J.; Xing, B. Mechanisms of slow sorption of organic chemicals to natural particles. Environ. Sci. Technol. 1996, 30, 1-11. (2) Luthy, R. G.; Aiken, G. R.; Brusseau, M. L.; Cunningham, S. D.; Gschwend, P. M.; Pignatello, J. J.; Reinhard, M.; Traina, S. J.; Weber, W. J., Jr.; Westall, J. C. Sequestration of hydrophobic organic contaminants by geosorbents. Environ. Sci. Technol. 1997, 31, 3341-3347. (3) Weber, W. J., Jr.; Leboeuf, E. J.; Young, T. M.; Huang, W. Contaminant interactions with geosorbent organic matter: Insights drawn from polymer sciences. Water Res. 2001, 35, 853-868. (4) Huang, W.; Young, T. M.; Schalautman, M. A.; Yu, H.; Weber, W. J., Jr. A distributed reactivity model for sorption by soils and sediments. 9. General isotherm nonlinearity and applicability of the dual reactive domain model. Environ. Sci. Technol. 1997, 31, 1703-1710. 7590
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 19, 2005
(5) Xia, G.; Ball, W. P. Adsorption-partitioning uptake of nine lowpolarity organic chemicals on a natural sorbent. Environ. Sci. Technol. 1999, 33, 262-269. (6) Xia, G.; Ball, W. P. Polanyi-based models for the competitive sorption of low-polarity organic contaminants on a natural sorbent. Environ. Sci. Technol. 2000, 34, 1246-1253. (7) Accardi-Dey, A.; Gschwend, P. M. Assessing the combined roles of natural organic matter and black carbon as sorbents in sediments. Environ. Sci. Technol. 2002, 36, 21-29. (8) Kleineidam, S.; Schuth, C.; Grathwohl, P. Solubility-normalized combined adsorption-partitioning sorption isotherms of organic pollutants. Environ. Sci. Technol. 2002, 36, 4689-4697. (9) Allen-King, R.; 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 Res. 2002, 25, 985-1016. (10) Ran, Y.; Xing, B.; Rao, P. S. C.; Fu, J. Importance of adsorption (hole-filling) mechanism for hydrophobic organic contaminants on an aquifer kerogen isolate. Environ. Sci. Technol. 2004, 38, 4340-4348. (11) Lohmann, R.; Macfarlane, J. K.; Gschwend, P. M. Importance of black carbon to sorption of native PAHs, PCBs and PCDDs in Boston and New York harbor sediments. Environ. Sci. Technol. 2005, 39, 141-148. (12) Alexander, M. Aging, bioavailability, and overestimation of risk from environmental pollutants. Environ. Sci. Technol. 2000, 34, 4259-4265. (13) Northcott, G. L.; Jones, K. C. Partitioning, extractability, and formation of nonextractable PAH residues in soil. 1. Compound differences in aging and sequestration. Environ. Sci. Technol. 2001, 35, 1103-1110. (14) Chiou, C. T.; Kile, D. E. Deviations from sorption linearity on soils of polar and nonpolar organic compounds at low relative concentrations. Environ. Sci. Technol. 1998, 32, 338-343. (15) Chiou, C. T.; Kile, D. E.; Rutherford, D. W.; Sheng, G.; Boyd, S. A. Sorption of selected organic compounds from water to a peat soil and its humic-acid and humin fractions: Potential sources of the sorption nonlinearity. Environ. Sci. Technol. 2000, 34, 1254-1258. (16) Carmo, A. M.; Hundal, L. S.; Thompson, M. L. Sorption of hydrophobic organic compounds by soil materials: Application of unit equivalent Freundlich coefficients. Environ. Sci. Technol. 2000, 34, 4363-4369. (17) Schultz, L. F.; Young, T. M.; Higashi, R. M. Sorption-desorption behavior of phenanthrene elucidated by pyrolysis-gas chromatography-mass spectrometry studies of soil organic matter. Environ. Toxicol. Chem. 1999, 18, 1710-1719. (18) Ghosh, U.; Talley, J. W.; Luthy, R. G. Particle-scale investigation of PAH desorption kinetics and thermodynamics from sediment. Environ. Sci. Technol. 2001, 35, 3468-3475. (19) Waterman, D.; Horsfield, B.; Leistner, F.; Hall, K.; Steve, S. Quantification of polycyclic aromatic hydrocarbons in the NIST standard reference material (SRM 1649A) urban dust using thermal desorption GCMS. Anal. Chem. 2000, 72, 3563-3567. (20) Weiss, J. M.; Mckay, A. J.; Derito, C.; Watanabe, C.; Thorn, K. A.; Madsen, E. L. Development and application of pyrolysis gas chromatography/mass spectrometry for the analysis of bound trinitrotoluene residues in soil. Environ. Sci. Technol. 2004, 38, 2167-2174. (21) Horsfield, B.; Disko, U.; Leistner, F. The microscale simulation of maturation: Outline of a new technique and its potential application. Geologis. Rundsch. 1989, 78, 361-374. (22) Jones, A. D. G.; Lorton, D. M.; Mason, J. R.; Smith, S.; Fenwick, C.; Glover, A.; Henkler, R. D.; Henkler, C. Assessing the potential for bioremediation. In In-situ and on-site bioremediation. Allenam, B. C., Lesson, A., Eds.; Battelle Press: Columbus, OH, 2001. (23) Shaw, L. J.; Beaton, Y.; Sousa, S.; Glover, L. A.; Killham, K.; Meharg, A. A. Mineralization of 2,4-dichlorophenol and glucose placed into the same or different hydrological domains as a bacterial inoculant. Soil Biol. Biochem. 2002, 34, 531-539. (24) Xing, B.; Pignatello, J. J. Time-dependent isotherm shape of organic compounds in soil organic matter: Implications for sorption mechanism. Environ. Toxicol. Chem. 1996, 15, 12821288. (25) Chung, N.; Alexander, M. Effect of concentration on sequestration and bioavailability of two polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 1999, 33, 3605-3608. (26) Condon, J. B. Equivalency of Dubinin-Polanyi equations and QM based sorption isotherm equation. A. Mathematical derivation. Microporous Mesoporous Mater. 2000, 38, 359-376.
(27) Condon, J. B. Equivalency of Dubinin-Polanyi equations and QM based sorption isotherm equation: B. Simulations of heterogeneous surfaces. Microporous Mesoporous Mater. 2000, 38, 377-383. (28) Karickhoff, S. W. Semiempirical estimation of sorption of hydrophobic pollutants on natural sediments and soils. Chemosphere 1981, 10, 833-846. (29) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental organic chemistry; John Wiley & Sons: New York, 1993. (30) Manes, M. Activated carbon adsorption fundamentals. In Encyclopedia of environmental analysis and remediation; Meyers, R. A., Ed.; John Wiley: New York, 1998. (31) Xing, B.; Pignatello, J. J. Dual-mode sorption of low-polarity compounds in glassy poly(vinyl chloride) and soil organic matter. Environ. Sci. Technol. 1997, 31, 792-799. (32) Braida, W. J.; Pignatello, J. J.; Lu, Y.; Ravikovith, P. I.; Neimark, A. V.; Xing, B. Sorption hysteresis of benzene in charcoal particles. Environ. Sci. Technol. 2003, 37, 409-417. (33) Werth, C. J.; McMillan, S. A.; Castilla, H. J. Structural evaluation of slow desorbing sites in model and natural solids using
(34)
(35)
(36)
(37)
temperature stepped desorption profiles. 1. Model development. Environ. Sci. Technol. 2000, 34, 2959-2965. Talley, J. W.; Ghosh, U.; Tucker, S. G.; Furey, J. S.; Luthy, R. G. Particle-scale understanding of the bioavailability of PAHs in sediment. Environ. Sci. Technol. 2002, 36, 477-483. Kleineidam, S.; Rugner, H.; Grathwohl, P. Desorption kinetics of phenanthrene in aquifer materials lacks hysteresis. Environ. Sci. Technol. 2004, 38, 4169-4175. Carroll, K. M.; Harkness, M. R.; Bracco, A. A.; Balcarcel, R. R. Application of a permeat/polymer diffusional model to the desorption of polychlorinated biphenyls from Hudson river sediments. Environ. Sci. Technol. 1994, 28, 253-258. Verschueren, K. Handbook of Environmental Data on Organic Chemicals, 3rd ed.; Van Nostrand Reinold Co. Inc.: New York, 1996.
Received for review April 7, 2005. Revised manuscript received July 14, 2005. Accepted July 18, 2005. ES050669B
VOL. 39, NO. 19, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7591
