Environmental Influences on the Partitioning and Diffusion of

Feb 14, 2008 - (Agilent) HPLC system that was coupled to a Quattro LC triple quadrupole mass ... 40% B at 0 min, held for 1 min, increased to 95% B at...
0 downloads 0 Views 646KB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 1990–1996

Environmental Influences on the Partitioning and Diffusion of Hydrophobic Organic Contaminants in Microbial Biofilms DANIEL WICKE,† UTA BÖCKELMANN,‡ AND T H O R S T E N R E E M T S M A * ,†,§ Department of Water Quality Control, Technical University of Berlin, Sekr KF 4, Strasse des 17 Juni 135, 10623 Berlin, Germany, Department of Environmental Microbiology, Technical University of Berlin, Sekr FR1-2, Franklinstrasse 29, 10587 Berlin, Germany, Federal Institute for Risk Assessment, FGr67, POB 330013, 14191 Berlin, Germany

Received September 10, 2007. Revised manuscript received December 19, 2007. Accepted January 2, 2008.

A biofilm reactor was used to investigate kinetic and thermodynamic aspects of the sorption of polycyclic aromatic hydrocarbons (PAH) as model compounds for hydrophobic organic contaminants (HOC) to intact microbial biofilms. Effective diffusion coefficients are in the range of 10-10 cm2 · s-1, resulting in equilibration times of more than 3 days for a biofilm of 100 µm thickness. Diffusion in the biofilm was strongly temperature-dependent and increased by a factor of 3 (phenanthrene) to 6 (fluoranthene, pyrene) between 5 and 35 °C. Drying and rewetting of the biofilm as well as the inclusion of Ca2+ ions and of humic acids all strengthened the biofilm rigidity and slowed down the diffusion of PAH. The later two factors also influenced the thermodynamics of the process as they supported the partitioning of PAH into the biofilm. Humic acid inclusion from solution into the biofilm illustrates that a microbial biofilm can act as a primer allowing for the buildup of a particulate organic phase from dissolved organic matter. PAH metabolites (3-hydroxy-phenanthrene and 1-hydroxy2-naphthoic acid) showed lower partition coefficients as compared to their parent compounds and 3-hydroxy-phenanthrene also showed a higher diffusion constant, indicating that these transformation products would be easily released into the water phase upon formation during PAH biodegradation in a biofilm. These results allow the quantification of the influence of environmental conditions on a biofilm’s function as a sink or as a diffusion barrier for PAH from aqueous solution, and they indicate the importance of kinetic aspects of this partitioning process.

Introduction Biofilms, consisting of layers of microorganisms embedded in a matrix of extracellular polymeric substances (EPS), are ubiquitous in heterogeneous aqueous systems. They occur * Corresponding author phone: +49-30-8412 3607; fax: +49-3084123685; e-mail: [email protected]. † Department of Water Quality Control, Technical University of Berlin. ‡ Department of Environmental Microbiology, Technical University of Berlin. § Federal Institute for Risk Assessment. 1990

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008

in natural systems such as soils, sediments (1), and aquifers (2), as well as in technical installations like sewer systems (3), wastewater treatment plants, water supply systems, on sand or activated carbon filters, or membrane reactors (4). In these systems biofilms form the boundary layer between a particulate and the dissolved phase, and they can influence exchange processes between these two phases. Membrane biofouling is one critical aspect that is presently studied intensively because of its adverse effects on system performance (4). Besides affecting the performance of technical installations, biofilms may also significantly influence the fate of hydrophobic organic compounds (HOC) in heterogeneous systems consisting of an aqueous and a particulate phase. But contrary to the interaction of biofilms with dissolved metals, their interaction with hydrophobic organic contaminants has rarely been studied. When the solid phase provides only poor sorption properties (e.g., water supply systems, sand filters, aquifer material), the evolution of a microbial biofilm on its surface may provide an important sink and reservoir for hydrophobic contaminants delivered by the aqueous phase. On the other hand, when the solid phase itself has good sorption properties (soil, activated carbon), a biofilm coverage may act as a barrier that slows down the mass transfer of a hydrophobic compound from the dissolved to the particulate phase. Recently we have presented a biofilm reactor system together with a modeling approach that allows the study of thermodynamic and kinetic aspects of the partitioning of HOC into microbial biofilms (5). Distribution coefficients (KOC) and diffusion constants could be determined for PAH for the first time, and it could be shown that biofilms can not only act as sorbents for hydrophobic contaminants but also as a diffusion barrier for their transport. This system allows for a more systematic investigation of the interaction of HOC in general and PAH in particular with microbial biofilms. In most natural and technical systems, the environmental or operating conditions are not constant but may vary strongly over time. For example, the water availability and temperature in surface soil is subjected to strong changes (6) due to weather conditions, and the chemical composition of water may vary in surface waters and technical systems for water treatment. Changes in the cationic composition or the content of dissolved organic matter (DOM) may affect biofilm quality and properties (7, 8), thereby influencing the partitioning of HOC to microbial biofilms as well as their diffusion in such biofilms. Knowledge on how such changes of environmental conditions affect mass transport processes in heterogeneous systems is also of importance to transfer quantitative results from one system to another. Biofilms are not only critical boundary layers for phase transition, but also important environmental compartments of high biological activity hosting a diverse population of microorganisms. While a biofilm may protect its organisms against unfavorable environmental conditions (9) it also allows bacteria to cooperate in the degradation of contaminants (e.g. ref 10). Thus, the fate of PAH in heterogeneous systems is not only influenced by partitioning into biofilms and a putative hydrophobic phase underneath, but it may also be affected by microbial degradation in such biofilms. Microbial transformation of PAH has widely been shown for 3- and 4-ring PAH, mainly resulting in more hydrophilic metabolites such as hydroxylated PAH and ring opening products of lower molecular weight (e.g. refs 11, 12). Once formed, such metabolites may behave quite different from their parent compound. 10.1021/es702267s CCC: $40.75

 2008 American Chemical Society

Published on Web 02/14/2008

These few examples may have outlined that biofilms, and factors that influence their properties, may thus become a third player in the game. It will not override the other two, the solute and the sorbent properties, but it may modulate their interaction to an extent that is significant for the functioning of (technical) systems as well as for the fate of the compound under consideration. Here we report how environmental conditions, like temperature, the cation or organic matter composition of the water phase, and the water status of the biofilm influence the partitioning process of PAH to microbial biofilms.

Materials and Methods System Setup. Details regarding biofilm production and sorption experiments are described in detail elsewhere (5). Briefly, a biofilm is established in a small glass reactor (length 12 cm, inner diameter 3.4 cm, volume 80 mL) that is filled with glass beads (3 mm diameter, total surface area is approximately 800 cm2). All experiments are performed with strain 30A as a member of Rhizobiaceae (Sinorhizobium sp.) isolated from a slow sand filtration unit. Each sorption experiment consists of two phases: a growth phase of 6 days to establish the biofilm in the reactor, and a sorption phase, in which the sorption of PAH (phenanthrene, PHE; fluoranthene, FLU; pyrene, PYR) and some of its metabolites (3-hydroxyphenanthrene, 3-OH-PHE; 1-hydroxypyrene, 1-OHPYR; 1-hydroxy-2-naphthoic acid, 1H2NA) onto the biofilm was studied. Standard Sorption Experiments. After biofilm establishment, reactors are immediately connected to the sorption experiment setup that is operated as a differential column batch reactor with a recycle flow. This concept allows us to model the system as a completely mixed batch reactor. The adsorption resistant piston pump Buechi B-681 (Buechi, Flawil, Switzerland) is set to a flow rate of 100 mL · h-1. Experiments are performed at room temperature (T ) 21 °C ( 2 °C, except experiments investigating temperature effects, see below). In the first step, each reactor is flushed with 1 L of tap water to remove nonattached biomass and the growth medium. Then, the supply bottle, which is continuously mixed by a magnetic stirrer, is filled with 0.5 L of tap water and spiked with the target compounds by adding 10 µL of a stock solution (62.5 mg · L-1 of each compound in propanol; 1H2NA 625 mg · L-1) to obtain the start concentration of 1.25 µg · L-1 (12.5 µg · L-1 for 1H2NA). After starting the pump, the sorption experiment was allowed to run for 2–3 h. Tap water was used as aqueous phase to suit the living conditions of the microorganisms (e.g., regarding ionic strength). More details have been provided elsewhere (5). Modified Sorption Experiments. Several effects that might influence sorption and diffusion processes were investigated: temperature, drying and rewetting of the biofilm, and incorporation of soil humic acids into the biofilm prior to sorption, and the addition of polyvalent cations into the dissolved phase. In addition, sorption of the metabolites 3-OH-PHE, 1H2NA, and 1-OH-PYR was investigated in comparison to phenanthrene and pyrene. Modification of experimental standard setup for investigation of effects influencing sorption processes: Temperature. Sorption of PHE, FLU, and PYR was performed at five different temperatures: 5, 10, 15, 25, and 35 °C in a climatic chamber. Drying and Rewetting. Three reactors were grown in parallel. One reactor was immediately used for a sorption experiment at reference conditions (DR0). The other two reactors were drained and dried with sterile air supplied by a diaphragm pump through a 0.45 µm filter for three days until the weight stabilized. Then, one reactor (DR1) was rewetted (refilled with water) 30 min before starting the sorption experiment. The other reactor (DR2) was dried for

another three weeks and reconnected to the sterile setup for another growth phase of 6 days to test the possibility of biomass regrowth from the dried biofilm. The growth potential of the bacteria in the dried and rewetted biofilm was tested by streaking a sample from the reactor on an agar plate. The regrown reactor was then also used for sorption experiment. Soil Humic Acid. Two reactors were grown in parallel: one reactor (H0) was used for comparison, whereas soil humic acid (Elliot soil humic acid standard, purchased from the International Humic Substance Society) was added to the other reactor (H1). For this purpose, 75 mg humic acid (HA) standard was dissolved in 100 µL of NaOH (0.1 M) and 2 mL of ultrapure water and made up to 500 mL with tap water. While the formation of HA microflocs started, the solution was pumped through the biofilm reactor at a flow rate of 50 mL · min-1 for 30 min. During this time, dissolved and precipitating humic acid was incorporated into the biofilm matrix. Afterward, the reactor was flushed for 2 min with tap water at a flow rate of 100 mL · min-1, and the sorption experiment was started. Using this procedure, 19 mg of soil HA was incorporated into 74 mg of attached biofilm (25% of biofilm dry mass). Calcium. While at standard conditions the sorption experiments were performed with tap water containing 150 mg · L-1 of Ca, one series of experiments was performed with a calcium free solution. This solution was prepared from ultrapure water with 50 mg · L-1 MgCl2 · 6H2O, 70 mg · L-1 NaCl, 185 mg · L-1 Na2SO4, and 300 mg · L-1 NaHCO3. This composition corresponds to that of the tap water with Ca2+-ions exchanged by Na+-ions and with similar conductivity (723 µS · cm-1 instead of 725µS · cm-1). Metabolites. One stock solution containing 3-OH-PHE (Ehrenstorfer, Augsburg, Germany), 1-OH-PYR (Aldrich, Steinheim, Germany), PHE, PYR (62.5 mg · L-1 each) and 1H2NA (Aldrich, Steinheim, Germany; 625 mg · L-1) was prepared in propanol, and 10 µL were added to each supply bottle (500 mL) of two parallel grown biofilm reactors. However, 1-OH-PYR turned out to be unstable in aqueous solution: as visible by HPLC analysis the compound vanished within 1-2 hours from such solutions. 1-OH-PYR had to be excluded from further studies. Sampling and Analysis. Samples were taken in intervals of 1-20 min from the column outlet, and 800 µL were filled in vials containing 200 µL of acetonitrile to prevent sorption of the PAH to the glass walls. All compounds were measured using HPLC-FLD, except 1H2NA (measured with LC-MS). HPLC-FLD-Analysis. Determination of the PAH and their metabolites was performed by HPLC-FLD according to Wicke et al. (5), with the following modifications to include the PAH metabolites. Gradient elution started with 50% AcCN and increased to 95% AcCN at 18 min, back to 50% AcCN at 19 min, and equilibration until 24 min. Time programmed fluorescence detection with λEX ) 260 nm and λEM ) 360 nm for 3-OH-PHE (RT 3.8 min) and with λEX ) 260 nm and λEM ) 385 nm for 1-OH-PYR (RT 5.7 min). LC-MS-Analysis. 1H2NA was analyzed by an HP 1100 (Agilent) HPLC system that was coupled to a Quattro LC triple quadrupole mass spectrometer (Micromass, Manchester, UK). To each sample 2H3NA was added as internal standard at a concentration of 10 µg · L-1. Eluents were H2O (A) and MeOH (B), each with 0.1% (v/v) formic acid. The separation was performed using a 3 × 150 mm Betasil PhenylHexyl column with 5 µm particle size (Thermo, Waltham, MA) at 25 °C. A sample volume of 50 µL was injected. Separation was achieved by gradient elution, starting with 40% B at 0 min, held for 1 min, increased to 95% B at 7.25 min, and decreased to 40% B at 8 min. After 15 min, the system was ready for the next injection. The flow was set to 0.3 mL · min-1. The mass spectrometer used electrospray VOL. 42, NO. 6, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1991

ionization (ESI) in the negative ion mode with a source temperature of 120 °C and a desolvation temperature of 100 °C. Detection was performed by multiple reaction monitoring (MRM) with the following conditions: parent ion: m/z 187, product ion: m/z 143, capillary voltage: 3.5 kV, cone energy: 25 V, collision energy: 17 eV. Additional Measurements. Biofilm dry mass was determined by washing the biofilm off the glass beads, filtration over 0.45 µm membrane filters, and weighing of the filters after dryness. Water content (89%), biofilm volume, biofilm thickness, and carbon content were determined as described in Wicke et al. (5). The partition coefficient KOC is the ratio between the equilibrium concentration in the biofilm (q∞ on a dry weight basis) and the corresponding equilibrium concentration in solution (c∞) normalized to the organic carbon fraction (foc ) 0.48 in dry biofilm): KOC )

q∞ 1 · c∞ fOC

(1)

Modeling. As we have shown before (5), the intrabiofilm diffusion coefficients can be determined by using the analytical solution of the diffusion equation ∂2c ∂c )D 2 ∂t ∂x

(2)

for the problem of diffusion from a stirred solution of limited volume into a plane sheet (13). Although the biofilm was grown on spherical glass beads a spherical model would have been inappropriate because it would have assumed the full sphere radius as the diffusion distance. The solution of eq 2 in a form expressing the amount of solute in the sheet at time t as a fraction of the corresponding amount at equilibrium (q/q∞) is shown in eq 3: ∞

q 2R(1 + R) )1exp(-Dλn2 t ⁄ l 2) 2 q∞ n)1 1 + R + R λ

∑

(3)

n

where the λn’s are the nonzero positive roots of tanλn ) -Rλn ,

(4)

· s ), l is the biofilm thickness (cm), and R a variable depending on the final fractional uptake of solute by the sheet at equilibrium (eq 5).

D the diffusion coefficient (cm2

1-

-1

c∞ 1 ) c0 1 + R

(5)

Visual fitting of the modeled curve to the experimentally determined time course of solute concentration was performed by selecting an appropriate value for the diffusion coefficient D in eq 3. Due to the fact that a biofilm is not homogeneous, the experimentally derived diffusion coefficients of this study are effective diffusion coefficients averaging spatial variations within the biofilm. More details on the modeling approach and its application to the biofilm reactor system can be found elsewhere (5).

Results and Discussion To study the partitioning process of PAH to intact biofilms, sorption experiments were performed in differential column batch reactors filled with glass beads that were covered with biofilm, and the concentration decrease of the solutes from the aqueous phase was followed over time. The time course of the solute concentration in each sorption experiment (concentration decrease at the reactor outlet) is displayed as normalized concentration (c/c0) versus time (Figure 1a). Provided that certain experimental conditions are fulfilled (see Materials and Methods section), these diagrams provide 1992

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008

FIGURE 1. Effect of temperature on sorption of PAH to biofilm. (a) Concentration decrease of fluoranthene (FLU) in the differential column batch reactor at two different temperatures (5 and 25 °C). (b) Arrhenius plot for phenanthrene, fluoranthene, and pyrene derived from the effective diffusion coefficients at five different temperatures between 5 and 35 °C. information on both (a) the thermodynamics of the sorption/ partitioning process as visible from the final dissolved concentration at equilibrium and (b) on the kinetics of the process, indicated by the slope of the curves. In a previous study we have shown that these experimental results can be successfully modeled to derive effective diffusion coefficients for PAH in microbial biofilms (5). Under standard conditions the effective diffusion coefficients of three- and four-ring PAH in biofilms were in the range of 10-10 cm2 · s-1, which is between those in aqueous solutions (10-6 cm2 · s-1) and those found for condensed organic matter (10-16 cm2 · s-1) (5). Partitioning of HOC into biofilms and their diffusion in the biofilm may depend upon the biofilm quality, which is influenced by the composition of the biofilm and external effects such as the presence of ions (14). Results of sorption experiments with changing experimental conditions and its influences on thermodynamics and kinetics are described in detail below. Temperature Effects. The temperature effect on the sorption of PAH to a microbial biofilm is shown in Figure 1a, where the steeper gradient of the curve at 25 °C illustrates the faster sorption process at higher temperatures. The effective diffusion coefficients derived from these experiments by using the above model at temperatures between 5 and 35 °C are summarized in Table 1 and show a pronounced temperature dependence. By increasing the temperature from 5 to 35 °C diffusion is accelerated by a factor of 3.4 (PHE) to 5.9 (PYR).

TABLE 1. Effective Diffusion Coefficients (D) for PAH in Biofilm at Different Temperatures and Resulting Activation Energies effective diffusion coefficient [10-9 cm2 · s-1] 5 °C phenanthrene 0.17 fluoranthene 0.09 pyrene 0.085

15 °C 25 °C 35 °C 0.30 0.16 0.15

0.35 0.25 0.25

0.58 0.53 0.50

activation energies [kJ · mol-1] 24.9 39.2 40.1

This effect of temperature on diffusivity can be described by eq 6, which is an application of the Arrhenius equation to diffusion controlled reactions: D)

kB T - EA e RT 6πη∞ rA

(6)

where D is the diffusion coefficient, kB is the Boltzmann constant, η∞ is a material (biofilm) dependent coefficient, rA is the hydrodynamic radius of the diffusing molecule, T is the absolute temperature, R is the gas constant, and EA is the activation energy of diffusion. By plotting (lnD - lnT) against 1/T, values for the activation energy can be extracted from the slope of the resulting linear plot (Arrhenius plot). For all three PAH, the Arrhenius plot results in straight lines as shown in Figure 1b, and the corresponding activation energies EA are summarized in Table 1. The obtained values range between 24.9 kJ · mol-1 (PHE) and 40.1 kJ · mol-1 (PYR), and increase from three-ring to four-ring compounds. Activation energies of diffusion processes appear to depend upon the viscosity of the medium in which diffusion takes place. For organic contaminants in polymers an average activation energy of 60 kJ · mol-1 has been reported (15) and values between 42 and 60 kJ · mol-1 have been found for toluene in humic acid disks (16). In contrast, diffusion through aqueous solutions is typically associated with activation energies around 20 kJ · mol-1 (15, 17). It is, therefore, reasonable if the activation energies for PAH in biofilms range between these values. While the temperature changes clearly affected the speed of diffusion in the microbial biofilm, partitioning of the PAH between dissolved phase and the biofilm was not significantly influenced within the investigated temperature range (Table 2). Only a small influence of temperature in that range on partitioning of PAH to organic carbon has already been shown in several studies, such as a 10% increase between 4 and 26 °C (18). Although thermodynamics may suggest a negligible influence of temperature at environmental conditions, our data show that temperature can be of relevance when a system is not at equilibrium and transport processes become important. Effect of Calcium. Calcium and other polyvalent cations from solution can be included into a biofilm (19) where they may exert a strengthening effect on the biofilm matrix, presumably by cation cross-linking of EPS (7, 8). Potential consequences of Ca inclusion into a biofilm were investigated by comparing the time course of PAH partitioning into biofilm from aqueous solutions with and without calcium (150 mg · L-1; Figure 2). While the three-ring compound PHE was not significantly affected, diffusion coefficients for the larger four-ring compounds FLU and PYR are about 50% higher in the absence of calcium. This indicates a stabilizing effect of calcium on the biofilm, resulting in lower diffusivities with increasing size of the diffusing molecule. These first data on the influence of polyvalent cations on the diffusion in microbial biofilms show one important consequence of the previously detected increase of the cohesiveness of microbial biofilms by calcium (7, 8).

Contrary to temperature, the inclusion of Ca into the biofilm also affected the thermodynamics of partitioning: in the absence of calcium partitioning of PAH was about 50% lower (log KOC smaller by 0.3) for all three PAH than in its presence (Table 2). One could assume that the reduction of the net charge of the EPS molecules allows for a more dense arrangement of the biofilm, with a lower content of polar water molecules. This would render the biofilm more hydrophobic, i.e. favorable for PAH sorption. Effect of Soil Humic Acid. The influence of natural DOM on the partitioning of organic contaminants between a dissolved and a particulate phase has been intensively studied, and it has been found that enhanced DOM concentrations may reduce sorption and can also influence the kinetics by facilitating the transport of HOC (20). In the case of microbial biofilms DOM, if incorporated in that biofilm, may change biofilm quality and properties. This can be of special relevance in systems with high concentrations of dissolved organic carbon such as soils or membrane systems for wastewater treatment. Humic acids are major contributors to DOM in natural systems, and for this reason it was studied whether humic acid inclusion into a microbial biofilm could affect the partitioning of PAH. A biofilm was amended with 25% dry weight of soil humic acid (IHSS Eliott HA) and the effective diffusion coefficients for the three PAH were about 20% lower compared to the unchanged reactor HA0 (Figure 2). Obviously, the inclusion of HA makes the biofilm more rigid and slows down the diffusion as compared to a pure microbial biofilm. If one considers the 20% organic carbon fraction of the humic acids in that amended biofilm, a 20% reduction in the diffusion coefficient is a remarkable effect. It implies that diffusion in pure humic acid is about 1 order of magnitude slower than in pure biofilm. For toluene, the diffusion coefficient in humic acid (0.7 × 10-8 cm2 · s-1, ref 16) is indeed about 1 order of magnitude lower compared to diffusion through a biofilm (1.3 × 10-7 cm2 · s-1; ref 21). The inclusion of humic acid also influenced the thermodynamics of the partitioning process. The sorbed amounts of PHE, FLU, and PYR increased by 25–26% compared to a pure biofilm due to the additional sorbent mass provided by the humic acid. Thus, inclusion of DOM into the biofilm clearly increased its sorption capacity and decreased the solute concentration at equilibrium. The partition coefficients, if normalized to the total organic carbon content of the sorbent (biofilm plus humic acid) however, remained unchanged (Table 2). Although humic acid is much more rigid than the biofilm, it is only slightly more hydrophobic (log Koc values of 4.2 and 5.0 for PHE and PYR; ref 22) than the biofilm organic matter (Table 2). Effect of Drying. Biofilms have been considered as hydrogels with the EPS as extremely hydrated polymer gel (23). Drying of hydrogels leads to shrinking and increasing density. During the drying processes, physically bound water evaporates from the biofilm but condensation reactions may also take place that lead to enhanced cross-linking of the EPS molecules. Namely the second process may not be completely reversible when a dried biofilm is rewetted. Then, a more condensed structure would remain that could lead to decreased diffusivity in the dried and rewetted biofilm compared to a fresh one. In our experiments, drying and rewetting of the biofilm (reactor DR1) results in 25–50% lower effective diffusion coefficients for the three investigated PAH compared to a fresh biofilm. This effect was stronger for the 4-ring compounds FLU and PYR (Figure 2). The same effect was found for the biofilm in reactor DR2, with a longer drying period and subsequent second microbial growth period of 6 days. These results suggest that drying is, to a certain extent, irreversible so that the rewetted biofilm retains a higher VOL. 42, NO. 6, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1993

TABLE 2. Partition Constants (log KOC) of PAH between an Aqueous and a Biofilm Phase at Different Environmental Conditions drying and rewetting

temperature

phenanthrene fluoranthene pyrene a

ref 5

5 °C

15 °C

25 °C

35 °C

4.1 ( 0.1 4.5 ( 0.2 4.6 ( 0.2

4.2 4.6 4.7

4.2 4.6 4.7

4.3 4.6 4.7

4.3 4.5 4.6

4.0 4.4 4.6

addition of humic acid

4.2a 4.6a 4.8a

4.1 4.5 4.6

calcium deficient

3.8 4.1 4.3

log Koc values related to the organic carbon of the biofilm only.

FIGURE 2. Influence of calcium, humic acid, and drying/ rewetting on the effective diffusion coefficients in biofilms (normalized values, reference state is tap water (with Ca), without humic acid and without drying of the biofilm). density as compared to the fresh biofilm. An influence on the thermodynamics of the partitioning was, however, not discernible in these experiments (Table 2). Remarkably, the bacteria embedded into the biofilm were still able to grow after drying for about three weeks at room temperature, as demonstrated by isolating bacteria of strain 30A after the second growth phase. This emphasizes an important ecological function of biofilms, their ability to protect bacteria from unfavorable environmental conditions such as desiccation (24). Metabolites. Biofilms as compartment of high microbiological density and activity are capable of transforming organic contaminants including PAH. Metabolites of hydrophobic compounds are usually less hydrophobic due to the inclusion of hydroxyl- or carboxyl moieties and are often of lower molecular mass. Metabolite behavior in heterogeneous systems with biofilms should, hence, differ from that of their parent compounds. Therefore, sorption experiments with PHE and PYR together with some of their metabolites, 1-hydroxy-2naphthoic acid (1H2NA), 3-hydroxyphenanthrene (3-OHPHE), and 1-hydroxypyrene (1-OH-PYR) were performed. 1H2NA has been shown to be one of the main metabolites in the microbial degradation of PHE (25, 26). 3-OH-PHE and 1-OH-PYR have been found as metabolites of PHE and PYR, respectively, in the presence of rooted poplar cuttings (27). Time curves of the concentration decrease of PHE, PYR, 3-OH-PHE, and 1H2NA are shown in Figure 3a (1-OH-PYR turned out to be unstable during the duration of the experiment). From the slope of these curves effective diffusion coefficients were calculated (Table 3). For the metabolite 3-OH-PHE it is almost twice as high as for its parent compound PHE. This sequence of diffusion coefficients is in line with our previous finding that compounds of higher polarity show faster diffusion in biofilms (5). Figure 3a also shows that the dissolved concentrations of the metabolites 3-OH-PHE and 1H2NA at equilibrium are higher compared to the parent compound PHE and PYR. No significant concentration decrease was found for 1H2NA 1994

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008

FIGURE 3. Sorption of PAH and selected metabolites to biofilm. (a) Sorption curves of the PAH-metabolites 1H2NA and 3-OH PHE and of PHE and PYR. (b) Water-organic carbon partition coefficients (log Koc) for biofilm organic matter versus octanol– water partition coefficients (log KOW, values from SciFinder, American Chemical Society). The KOC of 1H2NA is a literature value.

TABLE 3. Partition Constants (log KOC) and Effective Diffusion Coefficients (D) for Partitioning of 3-OH-PHE, PHE and PYR to Biofilm of Strain 30A (Rhizobiaceae), Mean of Two Parallel Experiments

3-OH-phenanthrene phenanthrene pyrene

log KOC

D [10-9 cm2 · s-1]

3.7 ( 0.1 4.0 ( 0.1 4.5 ( 0.2

0.75 ( 0.04 0.39 ( 0.06 0.26 ( 0.06

under the investigated conditions. This is in agreement with the reported Koc value of 1.3 for this compound that corresponds to a concentration decrease of only 0.3% of the initial concentration. Corresponding to the higher equilibrium concentration the carbon-normalized partition coefficients (log KOC) of the metabolite 3-OH-PHE is lower than that of the parent compound PHE (Table 3). Plotting log KOC versus the

octanol–water partition coefficients (log KOW) results in a good correlation for the four compounds (Figure 3b), which also validates the experimental approach. The thermodynamic and kinetic data correspondingly demonstrate the higher mobility of PAH-metabolites in heterogeneous systems with biofilms as compared to their parent compounds. Implications for Biofilm Diffusion. The experimental results outlined above may be combined in one model of diffusion/partitioning of nonionic hydrophobic compounds into biofilms: A microbial biofilm is a heterogeneous medium with more dense regions with a higher organic matter content and regions with a higher water content. Those regions with a higher water content are more readily accessible for more hydrophilic compounds, whereas less hydrophilic molecules can penetrate only into more hydrophobic regions of the biofilm. Correspondingly more hydrophilic compounds exhibit higher diffusion coefficients and are transported faster than more hydrophobic compounds. But our results have also shown that activation energies increase with the hydrophobicity of the solute. In accordance with literature data from other media this can be consistently explained by the fact that the more hydrophobic compounds move through the polymeric matrix with higher activation energies, whereas more hydrophilic compounds prefer the more aqueous regions with lower activation energies. Consequently, the diffusion of hydrophilic compounds will be less affected than their more hydrophobic counterparts by temperature and “environmental” factors that change the rigidity of the biofilm, such as drying and rewetting, the inclusion of polyvalent cations or humic acids. These kinetic effects will be even more pronounced for compounds with octanol–water partition coefficients above the range considered here (log Kow > 5). In these experiments equilibrium was reached within hours and, hence, such kinetic effects may seem to be relevant only for processes on a short time scale. However, this strongly depends upon the thickness of the biofilm (see eq 3). If it increases from 10 or 20 µm, as in these experiments, to 100 µm the time scale for equilibration extends to more than 3 days (80 h). Changes in biofilm quality, obviously, may also affect the thermodynamics of the partitioning process. We found that Ca-inclusion renders the biofilm more favorable for sorption, possibly by allowing a more condensed shape, as did the extra mass of organic carbon provided by humic acids. Implications for Contaminant Fate. These experimental data have shown that the fate of nonionic organic contaminants in heterogeneous systems may well be affected by biofilms, as their sorption tendency is comparable to that of other particulate organic matter. Thus, biofilms can form a significant sink for HOC in systems with low organic carbon content and high surface areas. In such compartments a microbial biofilm may also act as a primer that attracts humic acid and other DOM, and increases the mass of the sorbent and its sorption capacity further. In many other cases microbial biofilms may form a diffusion barrier that slows down the mass transfer between a solution and a particulate phase underneath. The kinetic data have shown that the importance of biofilms in such systems will increase with increasing hydrophobicity of the compound of interest and with decreasing temperature. Finally a biofilm can also function as a bioreactor from which microbial metabolites are more readily/easily released than the parent HOC. Because a biofilm responds to changing environmental conditions, these conditions have to be considered to adequately assess the role of a biofilm as a sorbent or as a diffusion barrier for HOC.

Acknowledgments This work was financially supported by the German Research Council (DFG, Bonn) through “INTERURBAN, TP MOBIL” (RE 1290/5-3, 5–4). We thank Katharina Knobel and Elke Profft for valuable support in the laboratory and Volker Niehaus (Department of Environmental Chemistry, TU Berlin) for support with the HPLC-FLD system.

Literature Cited (1) Galle, T.; Van Lagen, B.; Kurtenbach, A.; Bierl, R. An FTIR-DRIFT study on river sediment particle structure: Implications for biofilm dynamics and pollutant binding. Environ. Sci. Technol. 2004, 38, 4496–4502. (2) Ross, N.; Deschenes, L.; Bureau, J.; Clement, B.; Comeau, Y.; Samson, R. Ecotoxicological assessment and effects of physicochemical factors on biofilm development in groundwater conditions. Environ. Sci. Technol. 1998, 32, 1105–1111. (3) Jahn, A.; Nielsen, P. H. Cell biomass and exopolymer composition in sewer biofilms. Water Sci. Technol. 1998, 37 (1), 17–24. (4) Miura, Y.; Watanbe, Y.; Okabe, S. Membrane biofouling in pilotscale membrane bioreactors (MBRs) treating municipal wastewater: Impact of biofilm formation. Environ. Sci. Technol. 2007, 41, 632–638. (5) Wicke, D.; Böckelmann, U.; Reemtsma, T. Experimental and modeling approach to study sorption of dissolved hydrophobic organic contaminants to microbial biofilms. Water Res. 2007, 41, 2202–2210. (6) Reemtsma, T.; Savric, I.; Jekel, M. A potential link between the turnover of soil organic matter and the release of aged organic contaminants. Environ. Toxicol. Chem. 2003, 22, 760–766. (7) Korstgens, V.; Flemming, H. C.; Wingender, J.; Borchard, W. Influence of calcium ions on the mechanical properties of a model biofilm of mucoid Pseudomonas aeruginosa. Water Sci. Technol. 2001, 43 (6), 49–57. (8) Ahimou, F.; Semmens, M. J.; Novak, P. J.; Haugstad, G. Biofilm cohesiveness measurement using a novel atomic force microscopy methology. Appl. Environ. Microbiol. 2007, 73, 2897– 2904. (9) Costerton, J. W.; Lewandowski, Z.; Caldwell, D. E.; Korber, D. R.; Lappinscott, H. M. Microbial biofilms. Annu. Rev. Microbiol. 1995, 49, 711–745. (10) Wolfaardt, G. M.; Lawrence, J. R.; Robarts, R. D.; Caldwell, S. J.; Caldwell, D. E. Multicellular organization in a degradative biofilm community. Appl. Environ. Microbiol. 1994, 60, 434– 446. (11) Cerniglia, C. E. Microbial metabolism of polycyclic aromatic hydrocarbons. Adv. Appl. Microbiol. 1984, 30, 31–71. (12) Johnsen, A. R.; Wick, L. Y.; Harms, H. Principles of microbial PAH-degradation in soil. Environ. Pollut. 2005, 133, 71–84. (13) Crank, J. the Mathematics of Diffusion, Chapter 4.35; Clarendon Press: Oxford, 1956. (14) Sutherland, I. W. Biofilm exopolysaccharides: A strong and sticky framework. Microbiology 2001, 147, 3–9. (15) tenHulscher, T. E. M.; Cornelissen, G. Effect of temperature on sorption equilibrium and sorption kinetics of organic micropollutantssA review. Chemosphere 1996, 32, 609–626. (16) Chang, M. L.; Wu, S. C.; Chen, C. Y. Diffusion of volatile organic compounds in pressed humic acid disks. Environ. Sci. Technol. 1997, 31, 2307–2312. (17) Haque, R.; Lindstrom, F. T.; Freed, V. H.; Sexton, R. Kinetic study of the sorption of 2,4-D on some clays. Environ. Sci. Technol. 1968, 2, 207–211. (18) Piatt, J. J.; Backhus, D. A.; Capel, P. D.; Eisenreich, S. J. Temperature-dependent sorption of naphthalene, phenanthrene and pyrene to low organic carbon aquifer sediments. Environ. Sci. Technol. 1996, 30, 751–760. (19) Huang, J.; Pinder, K. L. Effects of calcium on development of anaerobic acidogenic biofilms. Biotechnol. Bioeng. 1995, 45, 212–218. (20) Liu, H. M.; Amy, G. Modeling partitioning and transport interactions between natural organic matter and polynuclear aromatic hydrocarbons in groundwater. Environ. Sci. Technol. 1993, 27, 1553–1562. (21) Holden, P. A.; Hunt, J. R.; Firestone, M. K. Toluene diffusion and reaction in unsaturated Pseudomonas putida biofilms. Biotechnol. Bioeng. 1997, 56, 656–670. (22) Jonassen, K. E. N.; Nielsen, T.; Hansen, P. E. The application of high-performance liquid chromatography humic acid columns in determination of Koc of polycyclic aromatic compounds. Environ. Toxicol. Chem. 2003, 22, 741–745. VOL. 42, NO. 6, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1995

(23) Hall-Stoodley, L.; Costerton, J. W.; Stoodley, P. Bacterial biofilms: From the natural environment to infectious diseases. Nature Rev. Microbiol. 2004, 2, 95–108. (24) McAuliffe, L.; Ellis, R. J.; Miles, K.; Ayling, R. D.; Nicholas, R. A. J. Biofilm formation by mycoplasma species and its role in environmental persistence and survival. Microbiology 2006, 152, 913–922. (25) Smith, M. R. The biodegradation of aromatic hydrocarbons by bacteria. Biodegradation 1990, 1, 191–206.

1996

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008

(26) Guerin, W. F.; Jones, G. E. Two-stage mineralization of phenanthrene by estuarine enrichment cultures. Appl. Environ. Microbiol. 1988, 54, 929–936. (27) Kuhn, A.; Ballach, H. J.; Wittig, R. Studies in the biodegradation of 5 PAHs (phenanthrene, pyrene, fluoranthene, chrysene and benzo(a)pyrene) in the presence of rooted poplar cuttings. Environ. Sci. Pollut. Res. 2004, 11, 22–32.

ES702267S