Environ. Sci. Technol. 2003, 37, 116-122
Sorption Kinetics and Microbial Biodegradation Activity of Hydrophobic Chemicals in Sewage Sludge: Model and Measurements Based on Free Concentrations E L S A A R T O L A - G A R I C A N O , * ,† IRIS BORKENT,† KAY DAMEN,† TJALLING JAGER,‡ AND WOUTER H. J. VAES§ Institute for Risk Assessment Sciences, Toxicology Division, Utrecht University, P.O. Box 80.176, NL-3508 TD Utrecht, The Netherlands, Laboratory for Ecotoxicology, National Institute of Public Health and the Environment (RIVM), P.O. Box 1, NL-3720 BA Bilthoven, The Netherlands, and TNO Nutrition and Food Research, P.O. Box 360, NL-3700 AJ Zeist, The Netherlands
In the current study, a new method is introduced with which the rate-limiting factor of biodegradation processes of hydrophobic chemicals in organic and aqueous systems can be determined. The novelty of this approach lies in the combination of a free concentration-based kinetic model with measurements of both free and total concentrations in time. This model includes microbial biodegradation activity of the chemical in the aqueous phase and chemical sorption kinetics with respect to organic carbon and aqueous phases. The time dependency of free and total concentrations of 7-acetyl-1,1,3,4,4,6hexamethyltetrahydronaphthalene and 7-acetyl-1,1,3,4,4,6hexahydro-4,6,6,7,8,8-hexamethylcyclopenta(g)-2-benzopyrane in activated sludge was experimentally determined in vitro. Evaporation losses from the test system were also determined. Least-squares regression to optimize the model parameters resulted in a model that is in accordance with the experimental data. Additionally, the model shows that a comparison between the decrease of free and total chemical concentrations in time, in combination with an independent measurement of the organic carbon/ water partition coefficient provides information about the ratelimiting step of the degradation process. This information can be used by sewage treatment plant managers to decide whether the microbial biodegradation activity itself or the desorption from organic carbon to the aqueous phase should be improved.
Introduction Many domestically and commercially used chemicals are partially removed in sewage treatment plants (STP) before the effluent is emitted into the environment. For most organic chemicals, sewage treatment decreases the concentration to a level that they are not harmful to the environment. * Corresponding author telephone: +31-30-2535018; fax: +3130-2535077; e-mail:
[email protected]. † Utrecht University. ‡ National Institute of Public Health and the Environment (RIVM). § TNO Nutrition and Food Research. 116
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003
Generally, STPs consist of two purification systems. The mechanical purification, which removes solids depending on their size and density, consists of a pretreatment unit and primary and secondary sedimentation tanks. The biological purification takes place in an aeration tank, where aerobic bacteria degrade chemicals usually via oxidation reactions. Here, the predominant means of decreasing the concentration of many organic chemicals is biodegradation. However, the quantification of this biodegradation is confounded for several reasons. First of all, loss processes such as evaporation, binding to organic matter, degradation, and growth of bacteria lead to a decrease in concentration but are not always separately quantified. Second, in many standard biodegradation tests (1-4), the concentrations of the test chemicals are far above realistic environmental concentrations. At these high concentrations, chemicals will be degraded as primary substrate, which is often not the case in STPs. Normally, the contribution of an individual chemical to the total energy flow in a STP system is almost negligible (5). Third, possible discrepancies between the use of artificial sewage and natural sewage to determine biodegradation rate constants have been described previously (6, 7). When quantifying biodegradation, it is often assumed that the total amount of chemical present is available for microbial biodegradation activity. Recent biodegradation studies have shown that the presence of a matrix such as slurry-particles, soil, or biomass decreases the rate of biodegradation. The conclusion of these studies (8-11) is that sorbed organic substrates are not available for the microbial biodegradation activity; therefore, the net biodegradation rate can decrease as a result of sorption of the chemical to a matrix. This effect is more pronounced when the desorption of a substance from organic carbon to the aqueous phase is slower than its microbial biodegradation activity (12, 13). Biodegradation of a chemical in sludge can be limited either by desorption from the matrix to the water or by microbial biodegradation activity. Thus, biodegradation is dependent on the balance between these two processes (12). Although interest in measuring the free concentration in environmental studies has increased substantially in recent years, biodegradation rate constants are generally still being determined under the assumption that the total amount of chemical present is available for degradation. The current study describes a mathematical model that combines microbial biodegradation activity of the freely dissolved chemical with kinetics of the chemical exchange between the organic carbon and the aqueous phase. Theoretical examples are shown that emphasize the information that can be obtained about rate-limiting processes when using the model in combination with free concentration measurements. The strength of the approach described in this paper lies in the combination of a free concentrationbased model with free concentration measurements. From these, conclusions can be drawn about the rate limitation of biodegradation. To demonstrate the validity of this approach, sludge was collected from the aeration tank of a STP and used without any additional spiking of analytes to study the in vitro biodegradation of two polycyclic musks. These polycyclic musks, AHTN (7-acetyl-1,1,3,4,4,6-hexamethyltetrahydronaphthalene) and HHCB (7-acetyl-1,1,3,4,4,6-hexahydro-4,6,6,7,8,8hexamethylcyclopenta(g)-2-benzopyrane), are common fragrances found in household products (14, 15). Physicalchemical parameters for AHTN and HHCB are given in Table 1. These chemicals are discharged to the sewer and are 10.1021/es020115y CCC: $25.00
2003 American Chemical Society Published on Web 11/26/2002
where k () Vmax/KM) is the degradation rate constant and Vmax is the maximum rate of the metabolic process. As depicted in Figure 1, changes in the concentration of the chemical in the organic phase and in the free concentration can be described by the following equations (see also Appendix):
TABLE 1. Physical-Chemical Properties of AHTN and HHCB (28) log Kow (L/L) aqueous solubility (mg/L) H (Pa‚m3/mol) MW (g/mol)
AHTN
HHCB
5.7 1.25 12.5 258.4
5.9 1.75 11.3 258.4
dCoc ) k1Cfree - k2Coc dt
(3)
dCfree Voc Voc ) k2Coc - k1Cfree - (kbio,true + kev)Cfree dt Vw Vw
FIGURE 1. Two-compartment model of the fate of organic chemicals in STPs assuming first-order kinetics for all processes, where k1 and k2 are the rate constants of mass transfer of the chemical from the aqueous to the organic phase and vice versa, respectively. Note that the rate constants for both microbial biodegradation activity (kbio,true) and evaporation (kev) are defined as dependent on Cfree. generally found in the micrograms per liter range in effluents of STPs (16-20). Both free and total concentrations of AHTN and HHCB were determined in the sludge at different time intervals. Free concentrations were determined using negligible depletion solid-phase microextraction (nd-SPME) as described by Vaes et al. (21). Total concentrations were determined using conventional liquid-liquid extraction.
Theory The free concentration-based model is used to determine the true (i.e., sorption-independent) biodegradation rate constant. The terminology “true” biodegradation rate constant is used in this paper to emphasize that sorptiondependent rate constants are not really biodegradation constants, but they are the product of a (true) biodegradation constant and the sorption coefficient. The model is schematically represented in Figure 1 and is based on the assumption that only the freely dissolved fraction is available for microbial biodegradation activity. The model distinguishes between an organic and an aqueous compartment with mass transfer of the chemical between these two compartments. Evaporation and microbial biodegradation activity are the only loss processes included in the model. A set of mass balance equations for the two compartments forms the basis of the model:
dAtotal dAoc dAfree ) + dt dt dt
(1)
where Atotal represents the total amount of the chemical, and Aoc and Afree represent the amount of the chemical in the organic and aqueous phases, respectively, at a certain point in time. The total amount of chemical in the model system changes only as a result of microbial biodegradation activity and evaporation from the aqueous phase. Biodegradation of chemicals by bacteria can be described using the Michaelis-Menten equation (22). When the free concentration of the chemical (C) is much smaller than the Michaelis-Menten constant KM (C , KM), the MichaelisMenten equation reduces to first-order kinetics:
dC ) -kC dt
(2)
(4)
where Cfree, Coc, and Ctotal represent time-dependent concentrations of the chemical in the aqueous phase, in the organic carbon, and the total concentration, respectively. Additionally, k2 and k1 are the rate constants of mass transfer of the chemical from the organic to the aqueous phase and from the aqueous to the organic phase, respectively. The rate for microbial biodegradation activity (kbio,trueCfree) and evaporation (kevCfree) are thus defined as being dependent on Cfree. Note that k1 and k2 are referenced to the volume of the organic phase (Voc), whereas kbio,true and kev are referenced to the volume of the aqueous phase (Vw). The sum of these two volumes is Vtotal. Combining eqs 1, 3, and 4 gives
dCtotal Vw ) -(kev + kbio,true)Cfree dt Vtotal
(5)
When free and total concentrations are determined experimentally, the model described by eqs 1 and 3-5 can be fitted to these data to obtain mass transfer, true biodegradation, and evaporation rate constants. Moreover, the model is used in this study to determine the rate-limiting step in the biodegradation of the system. Such information is helpful in deciding which step of the biodegradation process should be improved: microbial biodegradation activity or desorption from organic carbon to the aqueous phase. When considering the balance between two interdependent processes for which different units are used (concentration per volume of water and concentration per mass of organic carbon), it is more convenient to consider fluxes (i.e., the change in amount of chemical per unit of time). Whether the flux of the microbial biodegradation activity or the desorption flux is rate-limiting depends on the physicalchemical properties of the chemical (e.g., hydrophobicity, aqueous solubility) and the characteristics of sludge (e.g., bacterial species, organic carbon content, particle size). In Figure 2, model predictions are shown for these two different theoretical rate-limiting situations. When microbial biodegradation activity is rate-limiting (kbio,trueCfreeVw < k2CocVoc - k1CfreeVoc), net desorption is fast enough to virtually maintain partitioning equilibrium, resulting in parallel curves for the logarithms of both free and total concentration versus time. When desorption from the organic carbon into the aqueous phase is rate-limiting (kbio,trueCfreeVw < k2CocVoc - k1CfreeVoc), two different situations should be considered. First, under the assumption that the system is close to partitioning equilibrium when biodegradation starts, the decrease of log Cfree is faster than the decrease of log Ctotal, because the desorption flux is not fast enough to maintain this partitioning equilibrium. The flux of the microbial biodegradation activity decreases as a result of a decreasing free concentration, but only until both fluxes are equal (kbio,trueCfreeVw ≈ k2CocVoc - k1CfreeVoc). The second situation arises, beyond this point in time, when both slopes become parallel again. Nevertheless, there is then no partitioning equilibrium between the aqueous and the VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
117
FIGURE 2. Model predictions (A) when microbial biodegradation activity is rate-limiting and (B) when desorption is rate-limiting for total (- -) and free (s) concentrations. organic phases, thus the apparent Koc, which is calculated from k1/k2, will appear to be higher than the true value. In summary, when the curves of the logarithms of free and total concentrations in time have different slopes, desorption is rate-limiting. On the other hand, if both slopes are parallel, a comparison between true and apparent Koc values can be used to determine which process is ratelimiting.
Materials and Methods Chemicals. 7-Acetyl-1,1,3,4,4,6-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta(g)-2-benzopyrane (HHCB) was obtained from International Flavors and Fragrances (IFF) (Hilversum, The Netherlands), and 7-acetyl-1,1,3,4,4,6-hexamethyltetrahydronaphthalene (AHTN) was from PFW Aroma Chemical B.V., Hercules Incorporated (Barneveld, The Netherlands). Tenax TA (60-80 mesh; 177-250 µm) was obtained from Chrompack (Bergen op Zoom, The Netherlands). NaN3 was purchased from Merck (Darmstadt, Germany) and was used for inactivating degradation in sewage samples. Solvents were obtained from J. T. Baker (Deventer, The Netherlands). Materials. SPME fibers of 1-cm length with a 100-µm poly(dimethylsiloxane) coating were purchased from Supelco (Bellefonte, CA). In all cases the fibers were cut to 1-mm length to obtain small enough coating volumes for negligible depletion in the samples. New fibers were conditioned for 1 h at 250 °C in a GC split injector to desorb all impurities. Clean up of sewage extracts was carried out with 6 mL of 500-mg silica SPE columns (J&W, Folsom, CA) on a SPE-21 vacuum device (J. T. Baker). AHTN and HHCB Biodegradation. Biodegradation of AHTN and HHCB in a STP was mimicked according to a 118
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003
short-term batch experiment method developed by Nyholm et al. and Berg and Nyholm (6, 7). Experiments were conducted in duplicate with activated sludge obtained from the oxidation tank of the STP in De Bilt (The Netherlands) on November 22, 2000. Sludge was collected and aerated to maintain bacterial activity during transportation to the laboratory, where the experiment was started immediately upon arrival. The experimental setup consisted of a separatory funnel with 1.5 L of fresh sludge. A bottom airflow of about 300 mL/min was applied to aerate the system. All air leaving the system was led through tubes containing 0.4 g of Tenax TA. Breakthrough of AHTN and HHCB in these tubes was verified by placing a second tube in series for 24 h and accounted for less than 1% of the total volatilization for both chemicals. Sludge samples were taken from the separator funnel at different time periods over the course of 2 days to determine (i) the free and (ii) total concentrations in the sludge, (iii) the losses due to evaporation, and (iv) the organic carbon content of the sludge. Determination of Free Concentrations. Extractions were carried out using negligible depletion solid-phase microextraction (nd-SPME) according to Vaes et al. (21). This technique uses the fact that, by extraction of a minor (negligible) fraction of the freely dissolved chemical, the partitioning equilibrium of the chemical under study between the aqueous phase and the organic phase will not be perturbed. Therefore, the extracted fraction is related to the truly freely dissolved concentration. nd-SPME is derived from solid-phase microextraction (SPME), as introduced by Arthur and Pawliszyn (23), and has been used before to determine bioavailable concentrations (24). In particular, SPME methods have been applied successfully by Heberer et al. to the
analysis of musks in surface waters and STP effluents (25). At each time interval, two 10-mL aliquots of the sludge solution were transferred to 10-mL vials. Subsequently, NaN3 was added (final concentration 1 mM) to inhibit bacterial degradation activity. Analyses were carried out using a Varian Star 3600 CX GC equipped with a 8200 CX SPME autosampler with agitation device (Varian, Palo Alto, CA), a split/splitless injector, a 30 m × 0.32 mm fused silica DB 5.625 column with a 0.25-µm film thickness (J&W Scientific, Folson, CA), and a Saturn 2000 ion trap mass spectrometer. The injector temperature was maintained at 270 °C. The fiber was exposed to the sample for 10 min and immediately after exposure was desorbed in the injector for another 10 min. Depletion of the sample during absorption was 1.7%, which is considered a negligible amount. During desorption, the column was maintained at 50 °C. Subsequently, the injector was switched to the split mode, the fiber was removed from the injector, and the column temperature was raised at a rate of 40 °C/min to 150 °C, followed by 4 °C/min to 210 °C. Calibration was performed using external standards dissolved in water. Determination of Total Concentrations. Total concentrations were determined using liquid/liquid extraction followed by SPE. Extractions were carried out in duplicate on 10-mL samples of sewage sludge to which NaN3 was added directly after sampling (final concentration 1 mM). Extractions were carried out by addition of 6 mL of cyclohexane (J. T. Baker) followed by shaking for 2 h. After centrifugation of the samples at 3000 rpm for 10 min, 4 mL of the solvent was collected and concentrated to 1 mL under a gentle stream of nitrogen. Subsequently, the extract was cleaned using cyclohexane-conditioned 500-mg silica SPE columns, cyclohexane washing, and 6 mL of cyclohexane:ethyl acetate (98:2) as eluens. Recoveries of the procedure were between 85 and 110% for both chemicals and were determined using both spiking recoveries and consecutive extractions on nonspiked sludge. Cleaned extracts were analyzed on a Carlo Erba 5300 GC (Milan, Italy) equipped with a split/splitless injector, a 30 m × 0.25 mm fused silica DB-5MS column (J&W Scientific, Folson, CA) with a 0.25-µm film thickness, and a QDM 1000 mass spectrometer (Carlo Erba Instruments, Milan, Italy). Analyses were carried out by splitless injection of 1 µL at 225 °C. The column temperature was maintained at 90 °C for 1 min and raised by 30 °C/min to 150 °C, followed by 4 °C/min to 210 °C. The MS was scanned, using selected ion monitoring (SIM), for m/z 243. Determination of Losses Due to Evaporation. Losses due to evaporation were determined using adsorption to tubes filled with Tenax (0.4 g). After each sampling occasion, the old tubes were replaced. The Tenax was extracted with 4 mL of cyclohexane, and GC/MS analyses of these extracts were carried out as described above. Determination of Organic Carbon Content. Samples were taken periodically from the system to determine the organic carbon content of the sludge. Samples were freezedried, and part of the dried sample was transferred to a crucible. The crucible was placed in an oven, and the sample was burned at 550 °C for 2 h. Once the samples had returned to ambient temperature, the organic carbon content of the sample was determined gravimetrically. Application of the Model. Equations 1, 3, 4, and 5 were fitted to experimental data obtained in this study for AHTN and HHCB using Berkeley Madonna, version 8.0.1 (Berkeley, CA). Initial free and total concentrations were fixed at 1.15 and 5.25 µg/l for AHTN and 1.58 and 10.33 µg/L for HHCB, respectively. These data were experimentally determined at time zero. Free and total concentrations in time and losses due to evaporation were entered as experimental data. The
rate constants kbio,true, k1, and kev were optimized by the program using least-squares regression (while k2 was set to unity). Apparent Koc was derived from the ratio of k1 and k2. Determination of Koc. True Koc values were determined by an independent measurement of both AHTN and HHCB using the method described by Urrestarazu-Ramos et al. (24). Sludge was collected from the oxidation tank of the STP in De Bilt (The Netherlands) on January 15, 2001. NaN3 was added (1 mM final concentration), and the sludge was stirred overnight. Samples with different sludge concentrations were prepared by diluting with phosphate buffer (pH 7.0) and were shaken for 24 h at room temperature to achieve equilibrium. Both free and total concentrations were determined in each sample as described above. Resulting concentrations were fitted to the following equation, which is derived from the equilibrium partition coefficient and the mass-balance equation between two phases, by optimizing Ksludge:
f)
1 1 + KsludgeCsludge
(6)
where f is the free fraction (ratio between free and total concentrations), Ksludge is the sludge/water partition coefficient of the chemical, and Csludge is the concentration of sludge in the aqueous phase. Curve-fitting was performed using GraphPad Prism, version 2.01 (GraphPad Software, San Diego, CA). Koc was obtained by normalization of Ksludge to the organic carbon content of the sludge.
Results and Discussion Model Outcome and Experimental Results. The organic carbon content was stable during the entire experiment (data not shown). Previously, Gobas et al. (26, 27) described that a decrease in the organic carbon content in time would result in a decrease of the fugacity capacity and, consequently, in an increase of the free concentration. Since the organic carbon content of the system did not change during the study, this effect was not included in the model. In this study, data are collected from an aerated system. Biodegradation in this system was quantified indirectly from free and total concentrations at each point in time, while assuming a mass-balance. Clearly, the confidence in the model and its derived rate constants depends heavily on the quality to which this mass-balance assumption is applicable to the used test system. Sorption phenomena like binding to the glass wall can be virtually excluded because of the high concentration of competitive sorptive sludge particles. Sequestration of the chemicals was minimized by using sludge without additionally spiking the chemicals of interest. Therefore, most sequestration would already have occurred at the start of the experiment. Nevertheless, aeration of the system could be an important factor to influence the quality of the data set that is presented in the current study. Therefore, the study was designed to minimize aeration losses by including tubes filled with Tenax at the air exit. Breakthrough of the Tenax tubes was less than 1%, and the amount that was trapped on these Tenax tubes was included in the model as volatilization losses. Therefore, abiotic controls were not included in this study. The model, as described by eqs 1 and 3-5, was regressed to the experimental data, which were obtained in this study. Model curves, shown in Figure 3, are in agreement with the measured values, and it can be concluded that the model describes the data rather well. The fit in the initial part of the evaporation curve is less accurate but can be related to relatively short sampling times and, consequently, to concentrations that are close to the limit of detection of the method. VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
119
FIGURE 3. Comparison between model results (s) and experimentally determined values (9, Cfree; b, Ctotal; 2, cumulative amount evaporated) for AHTN (A) and HHCB (B).
TABLE 2. Optimized Values (( SEM) Obtained from Regression Analysis (h-1)
kbio,true kev (h-1) log Koc (k1/k2)a (L/kg)
AHTN
HHCB
0.023 ( 0.010 0.0085 ( 0.0002 3.1 ( 0.2
0.071 ( 0.030 0.021 ( 0.001 3.2 ( 0.2
a k and k are not given separately. The overall biodegradation rate 1 2 was limited by microbial biodegradation activity; therefore, only the ratio of k1 and k2 can be derived accurately.
Figure 3 shows that the curves of the logarithms of total and free concentrations in time are parallel for both AHTN and HHCB. As mentioned earlier, there are two situations in which this can occur: (i) the rate of the degradation process is limited by the flux of the microbial biodegradation activity or (ii) there exists a steady-state situation where microbial biodegradation activity and desorption fluxes are exactly balanced. A comparison between the apparent and true Koc would be conclusive concerning the rate limitation of the process. Table 2 gives values for the model parameters after model optimization. The optimized and thus apparent log Koc values, defined as log(k1/k2), are 3.1 (( 0.2) and 3.2 (( 0.2) for AHTN and HHCB, respectively. The corresponding true log Koc values, which were determined in an independent experiment during this study with inactivated sludge, were 3.8 (( 0.2) and 3.8 (( 0.2) L/kg. Although these values are higher than the log Koc values obtained from the model, it should be noted that for these experiments sludge was used that was collected at different points in time. Additionally, the data 120
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003
obtained from the model are also lower than previously reported data (28, 29). Variation in characteristics of the organic phase might explain such a difference between different locations at different times (30). Nevertheless, as pointed out earlier, Koc values obtained from the model are definitely not higher than the independently measured data either from this or earlier studies. This leads to the conclusion that, in the described experimental setup, microbial biodegradation activity is the rate-limiting step in the biodegradation of these two chemicals. Rate Constants Based on Free and Total Concentrations. Several studies have been published that suggest that only the freely dissolved chemical is available for microbial biodegradation activity (31-33); nevertheless, no actual free concentration measurements have been performed to support these suggestions. Clearly, analytical measurements of free concentrations continue to be a challenge. Therefore, eq 2 is normally used to determine biodegradation rate constants, but the concentration term in this equation normally refers to the total concentration:
dCtotal ) -kbio,apparentCtotal dt
(7)
The biodegradation rate constants kbio,apparent and kbio,true that can be obtained using total or free concentrations, respectively, are related through the ratio of these concentrations. Of course this ratio will only be constant in time under steadystate conditions. If at steady state the system is also close to partitioning equilibrium, free and total concentrations are related through the partition coefficient (Koc), the total
volume, and the volumes of the organic and aqueous phases. Therefore, only when microbial biodegradation activity is rate-limiting can eqs 2 and 7 be rewritten to
dCtotal Ctotal ) -kbio,true dt (1 + KocVoc/Vw)
(8)
For two systems with comparable biodegradation capacity but different sorption characteristics, the kbio,true will be the same in both systems, while kbio,apparent will be different and very dependent on the organic carbon contents. More generally, the use of kbio,true enables comparisons of biodegradation rates in completely different systems, without the need to correct for partition coefficients, organic carbon contents, or rate-limiting steps. Nevertheless, for most situations rate-limiting steps are not known; therefore, eqs 1 and 3-5 should be considered. Langworthy et al. (34) reported a half-life of 21 h for HHCB in a STP. In this particular study, the decrease of the total concentration in time was monitored. This resulted in a biodegradation rate constant (kbio,apparent) of 0.03 h-1. When kbio,true given in Table 1 is transformed into kbio,apparent, a value of 0.01 h-1 is obtained. Because these values are similar, it can be concluded that the ratio of microbial biodegradation activity to organic carbon content of both STPs was similar. This study demonstrates that the presented approach works well for chemicals like AHTN and HHCB, i.e., moderately hydrophobic chemicals for which the free concentration is sufficiently high to allow the detection of biodegradation. For very hydrophobic chemicals for which the microbial biodegradation activity is low and which are present at low concentrations (e.g., PCBs), the approach might not be applicable because of analytical detection limitations or long experimentation times. Thus, the current approach needs to be tested for other compound groups before a more general applicability can be claimed. In previous studies, desorption and biodegradation rates have been compared for several systems. Bosma et al. (10) described a mathematical concept for bioavailability, taking mass transfer and degradation into account. Carmichael et al. (35) determined desorption and mineralization of PAHs separately. Similar work was carried out earlier by Rijnaarts et al. (36) investigating the fate of R-hexachlorocyclohexane. Mihelcic and Luthy (37) also compared these two processes. Detailed calculations were performed by Ramaswami et al. (38). Their models included intraparticle dissolution and diffusion terms, which is predominantly of concern when desorption is ratelimiting. For these situations, they describe experimental systems for PAHs and their biotransformation in coal-tar slurry systems (39). These methods might well be useful for sewage treatment plants in general. All these studies investigated situations where either microbial biodegradation activity or desorption could be ratelimiting. In the current study, a new method is introduced by which the rate-limiting factor of the process can be determined by combining a simple free concentration-based kinetic model with measurements of both free and total concentrations in time. The measurement of these free concentrations by nd-SPME can be very straightforward for individual chemicals. Nevertheless, this hydrophobicitydependent extraction technique, which demands that only a negligible portion of the free concentration can be extracted, is rather difficult to apply to multi-compound analyses for chemicals with highly varying hydrophobicities. This is in contrast to total concentration measurements. For example, in this study, compounds with a higher hydrophobicity than the polycyclic musks would not comply to the negligible depletion criterion, while on the other hand, compounds with lower hydrophobicities can be measured with less sensitivity. These arguments show that only chemicals with
very similar hydrophobicities, like AHTN and HHCB, can be studied together. Implications. Sludge originating from STPs is sometimes used as a source of nutrients in agricultural land and amended soils (40, 41). Sludge-amended soil can however also be hazardous if it contains toxic and hydrophobic chemicals. For this reason, several countries consider incineration of sludge a more satisfactory method of sludge disposal. A STP produces two main products: effluent and sludge. Depending on the fate of these two products, priority can be given to clean either the effluent or the sludge. Thus, in countries where incineration of sludge is applied, the freely dissolved concentration in the effluent should be minimized by decreasing desorption and increasing microbial biodegradation activity. On the other hand, when sludge is used for amendment of soils, the concentration of pollutants in both the aqueous phase (the effluent) and the organic carbon (the sludge) should be minimized by increasing biodegradation of the total concentration by both microbial biodegradation activity and desorption. The current study has introduced tools to help with these optimizations.
Acknowledgments We thank Heather A. Leslie for the careful revisions of the manuscript and Joop L. M. Hermens for the fruitful discussions.
Appendix Analogous to Figure 1, the amount of chemical in the organic carbon compartment in time can be described by
dAoc ) k1CfreeVoc - k2CocVoc dt
(A1)
Under the assumption of mass balance (i.e., the amount of chemical that leaves the organic carbon compartment diffuses into the aqueous phase):
dAfree ) k2CocVoc - k1CfreeVoc - (kbio,true + kev)CfreeVw dt (A2) The total amount of the substance at any point in time is the sum of the amount in the organic carbon and in the aqueous phase:
dAtotal dAoc dAfree ) + ) -(kbio,true + kev)CfreeVw dt dt dt
(A3)
Describing these equations in concentration units:
dCoc ) k1Cfree - k2Coc dt dCfree Voc Voc - k1Cfree - (kbio,true + kev)Cfree ) k2Coc dt Vw Vw dCtotal Vw ) -(kev + kbio,true)Cfree dt Vtotal
(A4)
(A5)
(A6)
An explanation of terms is given in the body of the text.
Literature Cited (1) International Organization for Standardization (ISO). Water quality-evaluation of the elimination and biodegradability of organic compounds in an aqueous medium-Activated sludge simulation test; ISO/DIS 11733, ISO/TC 147/SC 5; ISO: Geneve, Switzerland, 1994. (2) ECC (Commission of the European Communities). Off. J. Eur. Communities 1988, L133, 123-127. VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
121
(3) ECC (Commission of the European Communities). Off. J. Eur. Community 1988, L133, 106-117. (4) OECD (Organization for Economic Cooperation and Development). Guideline for testing of chemicals No. 303 A, simulation test-Aerobic sewage treatment; Coupled Units Test; OECD: Paris. (5) Rittman, B. E. Water Sci. Technol. 1992, 25, 403-410. (6) Nyholm, N.; Ingerslev, F.; Berg, U. T.; Pedersen, J. P.; FrimerLarsen, H. Chemosphere 1996, 33, 851-864. (7) Berg, U. T.; Nyholm, N. Chemosphere 1996, 33, 711-735. (8) Alexander, M. Environ. Sci. Technol. 2000, 34, 4259-4265. (9) Pignatello, J. J.; Xing, B. Environ. Sci. Technol. 1995, 30, 1-11. (10) Bosma, T. N.; Middeldorp, P. J. M.; Schraa, G.; Zehnder, A. J. B. Environ. Sci. Technol. 1997, 31, 248-252. (11) Stringfellow, W. T.; Alvarez-Cohen, L. Water Res. 1999, 33, 25352544. (12) Wick, L. Y.; Colangelo, T.; Harms, H. Environ. Sci. Technol. 2001, 35, 354-361. (13) Grimberg, S. J.; Miller, C. T.; Aitken, D. Environ. Sci. Technol. 1996, 30, 2967-2974. (14) Balk, F.; Ford, R. A. Toxicol. Lett. 1999, 111, 57-79. (15) Rimkus, G. G. Toxicol. Lett. 1999, 111, 37-56. (16) Eschke, H. D.; Traud, J.; Dibowski, H. J. UWSF. Z. Umweltchem. O ¨ kotoxicol. 1994, 6, 183-189. (17) Eschke, H. D.; Dibowski, H. J.; Traud, J. UWSF. Z. Umweltchem. O ¨ kotoxicol. 1995, 7, 131-138. (18) Paxe´us, N. Water Res. 1996, 30, 1115-1122. (19) Simonich, S. L.; Begley, W. M.; Debaere, G.; Eckhoff, W. S. Environ. Sci. Technol. 2000, 34, 959-965. (20) Verbruggen, E. M. J.; van Loon, W. M. G. M.; Tonkes, M.; van Duijn, P.; Seinen, W.; Hermens, J. L. M. Environ. Sci. Technol. 1999, 33, 801-806. (21) Vaes, W. H. J.; Urrestarazu, E.; Verhaar, H. J. M.; Seinen, W.; Hermens, J. L. M. Anal. Chem. 1996, 68, 4463-4467. (22) Monod, J.; Changeux, J.-P.; Jacob, F. J. Mol. Biol. 1963, 6, 306329. (23) Arthur, C. L.; Pawliszyn, J. Anal. Chem. 1990, 62, 2145-2148. (24) Urrestarazu-Ramos, E.; Meijer, S. N.; Vaes, W. H. J.; Verhaar, H. J. M.; Hermens, J. L. M. Environ. Sci. Technol. 1998, 32, 34303435. (25) Heberer, Th.; Gramer, S.; Stan, H.-J. Acta Hydrochim. Hydrobiol. 1999, 27, 150-156.
122
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003
(26) Gobas, F. A. P. C.; Zhang, X.; Wells, R. Environ. Sci. Technol. 1993, 27, 2855-2863. (27) Gobas, F. A. P. C.; Wilcockson, J. B.; Russell, R. W.; Haffner, G. D. Environ. Sci. Technol. 1999, 33, 133-141. (28) van de Plassche, E. J.; Balk, F. Environmental risk assessment of the polycyclic musks AHTN and HHCB according to the EUTGD; RIVM Report 601503008; RIVM: Bilthoven, The Netherlands, 1997. (29) Winkler, M.; Kopf, G.; Hauptvogel, C.; Neu, T. Chemosphere 1998, 37, 1139-1156. (30) Chiou, C. T.; Kile, D. E.; Brinton, T. I.; Malcolm, R. L.; Leenheer, J. A. Environ. Sci. Technol. 1987, 21, 1231-1234. (31) Steinberg, S. M.; Pignatello, J. J.; Sawhney, B. L. Environ. Sci. Technol. 1987, 21, 1201-1208. (32) Hatzinger, P. B.; Alexander, M. Environ. Sci. Technol. 1995, 29, 537-545. (33) Hatzinger, P. B.; Alexander, M. Environ. Toxicol. Chem. 1997, 16, 2215-2221. (34) Langworthy, D. E.; Itrich, N. R.; Simonich, S. L.; Federle, T. W. Biotransformation of polycyclic musk, HHCB, in activated sludge and river water. Poster presented in the 10th Annual Meeting of SETAC Europe, Brighton, 2000. (35) Carmichael, L. M.; Christman, R. F.; Pfaender, F. K. Environ. Sci. Technol. 1997, 31, 126-132. (36) Rijnaarts, H. H. M.; Bachmann, A.; Jumelet, J. C.; Zehnder, A. J. B. Environ. Sci. Technol. 1990, 24, 1349-1354. (37) Mihelcic, J. R.; Luthy, R. G. Environ. Sci. Technol. 1991, 25, 169177. (38) Ramaswami, A.; Luthy, R. G. Environ. Sci. Technol. 1997, 31, 2260-2267. (39) Ramaswami, A.; Ghoshal, S.; Luthy, R. G. Environ. Sci. Technol. 1997, 31, 2268-2276. (40) Molina, L.; Diaz-Ferrero, J.; Coll, M.; Marti, R.; Broto-Puig, F.; Comellas, L.; Rodriguez-Larena, M. C. Chemosphere 2000, 40, 1173-1178. (41) Engwall, M.; Hjelm, K. Chemosphere 2000, 40, 1189-1195.
Received for review June 17, 2002. Revised manuscript received October 9, 2002. Accepted October 18, 2002. ES020115Y
