Gaseous Hexane Biodegradation by Fusarium solani in

oil for enhancing the transport and subsequent biodeg- radation of hexane by the fungus Fusarium solani in various bioreactor configurations. Silicone...
13 downloads 0 Views 222KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 2390-2395

Gaseous Hexane Biodegradation by Fusarium solani in Two Liquid Phase Packed-Bed and Stirred-Tank Bioreactors SONIA ARRIAGA,† RAU Ä L MUN ˜ OZ,‡ S E R G I O H E R N AÄ N D E Z , † BENOIT GUIEYSSE,‡ AND S E R G I O R E V A H * ,† Departamento de Ingenierı´a de Procesos e Hidrau ´ lica, Universidad Auto´noma Metropolitana-Iztapalapa, Av. San Rafael Atlı´xco No. 186, Col. Vicentina, P.O. Box 55-534, 09340, Mexico City, Mexico, and Department of Biotechnology, Lund University, P. O. Box 124, 22100 Lund, Sweden

Biofiltration of hydrophobic volatile pollutants is intrinsically limited by poor transfer of the pollutants from the gaseous to the liquid biotic phase, where biodegradation occurs. This study was conducted to evaluate the potential of silicone oil for enhancing the transport and subsequent biodegradation of hexane by the fungus Fusarium solani in various bioreactor configurations. Silicone oil was first selected among various solvents for its biocompatibility, nonbiodegradability, and good partitioning properties toward hexane. In batch tests, the use of silicone oil improved hexane specific biodegradation by approximately 60%. Subsequent biodegradation experiments were conducted in stirred-tank (1.5 L) and packed-bed (2.5 L) bioreactors fed with a constant gaseous hexane load of 180 g‚m-3reactor‚h-1 and operated for 12 and 40 days, respectively. In the stirred reactors, the maximum hexane elimination capacity (EC) increased from 50 g‚m-3reactor‚h-1 (removal efficiency, RE of 28%) in the control not supplied with silicone oil to 120 g‚m-3reactor‚h-1 in the biphasic system (67% RE). In the packed-bed bioreactors, the maximum EC ranged from 110 (50% RE) to 180 g‚m-3reactor‚h-1 (>90% RE) in the control and two-liquid-phase systems, respectively. These results represent, to the best of our knowledge, the first reported case of fungi use in a two-liquid-phase bioreactor and the highest hexane removal capacities so far reported in biofilters.

Biological air treatment processes are based on the transfer of the contaminants into an aqueous phase prior to their biodegradation. Mass transfer is very relevant for the case of hydrophobic pollutants, such as hexane, where this step can limit bioavailability to the microorganisms (3). Hence, new methods must be found to increase the transfer of the pollutants to the microorganisms. In this perspective, it has recently been shown that fungi might improve the removal of hydrophobic compounds because their aerial mycelia, which are in direct contact with the gas, could take up these compounds faster than flat aqueous bacterial biofilms (46). As a significant example, hexane EC ranging from 100 to 150 g‚m-3reactor‚h-1 have been obtained with Aspergillus niger (7) in comparison to the 10-60 g‚m-3reactor‚h-1, reported with Pseudomonas (8, 9) and with activated sludge (10). Fungi are also advantageous in biofilters because of their tolerance to the low pH and moisture content that are commonly encountered in such systems (2, 11). On the other hand, fungi generally grow more slowly than bacteria, may induce high-pressure drop, and in extreme cases, such as severe drying, sporulation may be induced and spores can be liberated to the environment. Pursuing also the objective to improve mass transfer, various authors (12-14) have shown that the inclusion of a hydrophobic organic phase in the reactors can significantly enhance the transfer of pollutants to the microorganisms and, thereby, their removal. In these systems, commonly known as two-liquid-phase bioreactors (TLPBs), the organic phase is used to rapidly absorb poorly soluble organic pollutants from gaseous streams. The pollutants are then transferred by liquid-liquid partitioning to the aqueous phase at sub-inhibitory levels where they are biodegraded. Hexane EC of up to 80 g‚m-3reactor‚h-1 has been reported in a biotrickling filter percolated with a mixture of silicone oil and water (1:1) (14). Solvent selection is a crucial step to the development of TLPBs, as the organic phase must be nonbiodegradable, nonvolatile, nonmiscible, and biocompatible, and must have good partitioning properties with the target pollutants (15, 16). It can be difficult to combine these properties but silicone oil has often been used to support the biodegradation of hydrophobic pollutants for being biocompatible and highly stable (12-14). Both the use of fungi and the use of an organic phase have the potential to improve the removal of hydrophobic air pollutants. However, little data is still available to compare these technologies that have never, to the best of our knowledge, been combined together. This study was therefore conducted to systematically compare hexane removal in several bioreactor configurations inoculated with a fungal strain and supplied with silicone oil.

Introduction

Materials and Methods

Air pollution represents a serious health and environmental problem that requires effective control. In particular, many volatile organic compounds (VOCs) can cause odor nuisances and have toxic effects, and must be removed from contaminated streams (1). Compared to classical methods for air treatment, such as absorption, condensation, adsorption, or incineration, biological methods are often considered as most cost-efficient for the treatment of large volumes of air contaminated with low concentrations of organic pollutants (2).

Chemicals. Hexane (95%) was obtained from Tecsiquim Chemical Co. (Mexico City). Silicone oil [poly(dimethylsiloxane) 200 fluid with a viscosity of 20 cSt and density of 0.95 g‚mL-1], hexadecane, tetradecane, undecane, 2-undecanone, diethyl sebacate, and 1-decanol were purchased from SigmaAldrich (+99%). Microorganism and Mineral Salt Medium (MSM). Fusarium solani (CBS 117476) (17) was used for hexane biodegradation. The mineral salt medium (MSM) used for fungal growth contained (g‚L-1) the following: NaNO3, 6; KH2PO4, 1.3; MgSO4‚7H2O, 0.38; CaSO4‚2H2O, 0.25; CaCl2, 0.055; FeSO4‚7H2O, 0.015; MnSO4‚H2O, 0.012; ZnSO4‚7H2O, 0.013; CuSO4‚7H2O, 0.0023; CoCl2‚6H2O, 0.0015; and H3BO3, 0.0015. The pH was adjusted to 4 with 0.01 M HCl.

* Corresponding author phone: +52-55-58-04-6408; fax: +5255-58-04-6407; e-mail: [email protected]. † Universidad Auto ´ noma Metropolitana-Iztapalapa. ‡ Lund University. 2390

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

10.1021/es051512m CCC: $33.50

 2006 American Chemical Society Published on Web 02/22/2006

The fungal inoculum contained a spore suspension with an initial concentration of 2 × 107 spores‚mL-1 of MSM. The spores were obtained by growing the fungus on agar Petri plates with 5 g‚L-1 of malt extract and collecting the spores with MSM containing a few drops of Tween 80. Partition Test. Glass tubes of 55 mL capacity were filled with 2 mL of the tested organic solvents and closed with Mininert Teflon Valves (VICI Precision Sampling, Inc., Baton Rouge, LA). For each solvent tested, a duplicate set of tubes was prepared and supplied with 10, 50, 250, 750, or 1250 µL of hexane. The tubes were then vigorously shaken in a vortex for 1 min and allowed to equilibrate during 30 min at 30 °C on a rotary shaker at 150 rpm. Gas samples of 200 µL of each experiment were then withdrawn from the tubes’ headspaces to measure the hexane concentration. The amount of hexane in the silicone oil was calculated by subtracting the amount of hexane in the gas phase at the equilibrium to the initially added hexane. The hexane partition coefficient was then calculated as the ratio between its concentration in the gaseous and organic phases. Solvent Biodegradability and Toxicity. Biodegradability tests were conducted in duplicate in 155 mL glass flasks with 19 mL of MSM, 2 mL of the tested solvent, and 1 mL of fresh fungal inoculum with a concentration of 55 mgprotein‚L-1. The flasks were closed with butyl septa, sealed with aluminum caps, and incubated on a rotary shaker at 150 rpm and 30 °C. Control flasks were prepared and incubated under similar conditions but were not supplied with the organic solvent. The concentrations of O2 and CO2 in the flasks’ headspace were monitored every 3 days by GC for one month by withdrawing 200 µL gas samples with a 500 µL syringe. Solvents were considered biodegradable if the CO2 production in the flasks supplied with solvent as sole source of carbon and energy was at least five times higher than the CO2 in the controls without solvent. Toxicity tests were carried out in duplicate as described above but supplied with glucose, yeast extract, and peptone at 1, 0.02, and 0.02 g‚L-1, respectively, as easily available carbon and energy sources. Control flasks were also provided with the nutritive solution without organic phase. The concentrations of O2 and CO2 in the flasks’ headspace were monitored by GC for one month as described above. A solvent was considered toxic if no CO2 was produced. Batch Hexane Degradation Experiments. Glass flasks of 615 mL were supplied with 43 mL of mineral medium, 5 mL of silicone oil, and 2 mL of fungal inoculum to achieve a final concentration of 17 mg protein‚L-1, and 20 µL of hexane (initial gaseous concentration of 20 g‚m-3). The flasks were incubated at 30 °C on a rotary shaker at 150 rpm. Controls without silicone oil were tested under similar conditions. The concentrations of hexane, CO2, and protein were periodically monitored by withdrawing 200 µL and 600 µL samples from the gaseous and liquid phases, respectively. The hexane, CO2, and protein data were fitted with the integrated Gompertz model and the maximum volumetric rates were then calculated as reported by Acun ˜ a et al. (18). The specific rates (mgHexane‚g-1protein‚h-1 and mgCO2‚g-1protein‚h-1) were obtained by dividing the volumetric rate by the protein present. Stirred-Tank Rector (STR). A 2 L Multigen fermentor (New Brunswick Scientific Co. Inc., NJ) equipped with a double Rushton turbine operated at 400 rpm was used to evaluate hexane biodegradation at 30 °C (19). The reactor was filled with 1.35 L of mineral medium and 150 mL of silicone oil (organic/aqueous ratio of 1:9), and inoculated with F. solani to attain an initial biomass concentration of approximately 220 mg protein‚L-1. The reactor was aerated at 1 vvm (one volume of air per volume of working volume per minute) that corresponded to 1.5 L‚min-1 equivalent to an empty bed residence time (EBRT) of 1 min referred to the working

liquid volume (mass-flow controller 60061; Cole Parmer, Vernon Hills, IL) and fed with a constant hexane load of 180 g‚m-3reactor‚h-1 with a mixture of air saturated with hexane and moist air. The pH was automatically controlled at 4.0 with HCl (0.01 M). Similar experiments were conducted without silicone under the same conditions as controls. To maintain a constant biomass concentration, 100 mL of liquid was replaced with MSM every 2 days. The collected liquid was centrifuged to separate the phases (aqueous, biomass, and oil) and the silicone oil was then reintroduced to the reactor to keep a constant medium-oil ratio. Gas samples were withdrawn at the inlet and outlet of the reactor daily to monitor the concentration of hexane. A final set of experiments was carried out by varying the hexane inlet gaseous concentrations between 1 and 26 g‚m-3 under constant EBRT to evaluate the influence of hexane load on the STR performance. Packed-Bed Reactor (PBR). The packed-bed bioreactor was similar to that reported previously (6). It consisted of a 1 m cylindrical glass column with an inner diameter of 0.07 m, filled with 2.5 L (310 g) of Perlite (average diameter of 4 mm, Hummert, Mexico). Hexane saturated air was mixed with moist air and introduced at the top of the reactor with an EBRT of 1 min (mass-flow controller 60061; Cole Parmer, Vernon Hills, IL). The reactor was inoculated during the first week of operation by recirculating an emulsion (200 mL) of silicone oil and F. solani spores in MSM at 10 mL‚min-1 for 15 min every day to ensure the absorption of the emulsion on perlite. This emulsion was prepared by homogenizing 5% of silicone oil, 0.1% of Tween 80, and 5 g‚L-1 of malt extract in fungal MSM at 5000 rpm during 8 min with an UltraTurrax (Ika-Werk Staufen, Germany). The emulsion was then supplied with F. solani to obtain a final concentration of 2 × 107 spores‚mL-1. Fresh MSM (100 mL) was sprayed every 2 days at the top to maintain the moisture content at 65% and the pH at 4. The packed-bed bioreactor was fed with a hexane inlet load of 180 g‚m-3reactor‚h-1. A final set of experiments was carried out by varying the hexane inlet gaseous concentrations between 1 and 15 g‚m-3 under constant EBRT to evaluate the influence of hexane load on the bioreactor performance. Analytical Methods. Hexane gaseous concentration was measured with a GC-FID (Gow Mac, Series 580; Bridgewater, NJ), equipped with a 6 ft × 1/8 in. × 0.085 in. SS column (2% Silar 10C, Graphpac GC 80/100; Alltech, Deerfield, IL). Nitrogen was used as a carrier gas at a flow rate of 20 mL‚min-1. The FID was supplied with hydrogen (30 mL‚min-1) and reconstituted air (300 mL‚min-1). The injector, oven, and detector temperatures were maintained at 190, 180, and 210 °C, respectively. During the solvent biodegradability and toxicity tests, CO2 and O2 concentrations were analyzed by GC-TCD (19). The CO2 concentration in gas samples from the packed bed bioreactor was measured with an infrared analyzer (3400 gas analyzer; California Analytical Instruments). The reported values correspond to the difference between the outlet of the reactor and the ambient CO2 concentrations. The protein concentration was determined using the Lowry method by hydrolyzing the sample with a 0.2 M NaOH solution and using bovine serum albumin as standard. The sample was first centrifuged at 8000 rpm for 10 min to separate the aqueous phase from the oil phase. The biomass content in the carrier was measured as the loss of volatile solids at 550 °C according to Standard Methods (20). Three samples were taken from different levels of the reactor and the measurements were made in duplicate. The reported biomass content is the average of the 3 samples. Scanning Electron Microscopy. Biofilter samples were observed in a digital scanning electron microscope (JSM5900 LV, JEOL, Japan) using 13 kV accelerating voltage. The VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2391

TABLE 1. Partition, Toxicity, and Biodegradability Tests with Fusarium solani solvent

log Powa

aqueous solubilitya

Kb

biodegradable

toxic

silicone oil hexadecane tetradecane undecane 1-decanol diethyl sebacate 2-undecanone

NKc 8.25 7.2 6.5 4.57 4.33 4.09

NK 0.0009 0.0022 0.0040 37 80 19.7

0.0034 ( 0.0003 0.0042 ( 0.0004 0.0026 ( 0.0001 0.0038 ( 0.0001 0.0073 ( 0.0008 0.0115 ( 0.0003 0.0050 ( 0.0009

+ + + -

+ + +

a Logarithm of the octanol-water partitioning ratio and aqueous solubility (mg L-1) according to the SRC PhysProp Database. b K: Partitioning coefficient of hexane in organic and gaseous phases calculated as the ratio of the hexane concentrations in gas and organic phases and expressed as the average ( 95% confidence interval calculated from 5 experimental measurements. c Not known.

samples were prepared according to the method described by Acun ˜ a et al. (18). Calculations. Results from the reactor experiments were expressed in terms of the hexane volumetric elimination capacity (EC in g‚m-3reactor‚h-1) and the hexane gaseous elimination capacity (ECg in g‚m-3gas‚h-1) according to the following formulas:

EC )

Qg Vreactor

(Sin - Sout)

ECg(Stirred tank) ) EC

(1)

() 1 G

(2)

(1)

(3)

ECg(Packed bed) ) EC

where Vreactor is the reactor volume (m3reactor); Qg is the air flow (m3gas‚h-1); Sin and Sout are the inlet and outlet hexane concentrations, respectively (g‚m-3gas); G (m3gas‚m-3reactor) corresponds to the gas hold up and was determined experimentally by dividing Vg (gas volume) over the reactor volume (Vreactor); and  is the bed void fraction of the PBRs (m3gas‚m-3reactor). For the Qg of 1.5 L‚min-1, the gas volume was 0.2 L corresponding to a hold up for the STR of 0.133 m3gas‚m-3reactor. The bed void fraction for the PB reactor was 0.65 m3gas‚m-3reactor.

FIGURE 2. Time course of the hexane elimination capacities (filled symbols) and hexane removal efficiencies (open symbols) in the stirred tank bioreactor inoculated with Fusarium solani in mineral medium (squares) and with silicone oil (triangles).

Results Solvent Evaluation. Silicone oil was the only nontoxic and nonbiodegradable organic phase for F. solani (Table 1). The hexane concentration in silicone oil was approximately 300 times greater than that in the gas phase. Batch Experiments. The kinetic consumption of hexane by F. solani was evaluated in batch experiments (Figure 1).

FIGURE 3. Time course of the hexane elimination capacity (filled symbols) and hexane removal efficiency (open symbols) in the packed bed bioreactor supplied with 5% of silicone oil and inoculated with Fusarium solani.

FIGURE 1. Time course of total hexane consumption and of CO2 and protein concentration by Fusarium solani in batch experiments with 10% of silicone oil (open symbols) and without silicone oil (filled symbols). 2392

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

The specific hexane consumption and CO2 production rates were 156 mgHexane‚g-1protein‚h-1 and 307 mgCO2‚g-1protein‚h-1, respectively, for the experiment performed without silicone oil compared to 356 mgHexane‚g-1protein‚h-1 and 113 mgCO2‚g-1protein‚h-1, respectively, when adding the solvent. These data were obtained considering the relation of the total initial hexane concentration in the flask (22.3 gtotal hexane‚m-3) and the initial headspace concentration of hexane for each experiment. The protein production in the experiment with silicone oil was 33% greater than that without silicone oil. Stirred-Tank Bioreactors Maximum ECs of 120 and 50 g‚m-3reactor‚h-1 were attained in the biphasic bioreactor and in the control reactor, respectively, after 3 and 4 days of

TABLE 2. Maximum Hexane Elimination Capacities Expressed in Terms of the Reactor and Gas Volume, and the Specific Hexane Consumption Rates for Packed-Bed and Stirred-Tank Bioreactors Inoculated with Fusarium solani experiment

load g‚m-3reactor‚h-1

EC g‚m-3reactor‚h-1

RE %

ECg g‚m-3gas‚h-1

stirred tank bioreactor

one liquid phase two liquid phases

180 180 600 1575

50 120 252 580

28 67 41 37

554 900 1895 4361

packed-bed bioreactor

one liquid phasea

164 568 180 680

110 130 165 360

50 30 >90 54

170 200 277 554

two liquid phases

specific removal rate mgHexane‚g-1protein‚h-1 250 545 1145 2636 15 17.2 44.6 87

a According to the values of Arriaga et al. (6) achieved in a fungal PBR. This study can be used as a control experiment to the fungal PBR tested in this study.

FIGURE 4. Effect of hexane inlet load on elimination capacity in the packed-bed (A) and the stirred-tank (B) bioreactors containing 5% and 10% of silicone oil, respectively, and inoculated with Fusarium solani. experiment (Figure 2). In the biphasic reactor, hexane absorption in the oil occurred during the first hour of operation, and during the first week the EC was between 80 and 120 g‚m-3reactor‚h-1.The protein concentration was maintained at approximately 220 mg‚L-1 during the 12 days of operation. In both reactors, the EC decreased by 40% at day 9. The maximum specific microbial activities recorded at the days of maximum EC and with different hexane inlet loads are shown in Table 2. A maximum EC of 580 g‚m-3reactor‚h-1 was attained with a load of 1575 g‚m-3reactor‚h-1 (Table 2 and Figure 4). The critical inlet load (RE up to 95%) was around 160 g‚m-3reactor‚h-1.

The effect of hexane load on EC in the biphasic PBR was evaluated after 45 days of operation and the experiment was carried out during 2 weeks (Figure 4). The critical inlet load was approximately 150 g‚m-3reactor‚h-1 and a maximum EC of 360 g‚m-3reactor‚h-1 was attained at a load of 680 g‚m-3reactor‚h-1 (Table 2). Scanning Electron Microscopy. Samples from the packedbed bioreactor showed fungal morphologies. It was also possible to observe mycelia adhered to silicone oil and some bacteria on the hyphae (Figure 5).

Packed-Bed Bioreactor. The PBR with 5% of silicone oil was operated for 45 days under a constant inlet hexane load of 180 g‚m-3reactor‚h-1 (Figure 3). The EC remained at approximately 40 g‚m-3reactor‚h-1 during the first 10 days of experimentation. It then reached a maximum of 180 g‚m-3reactor‚h-1 at day 14 that was maintained during one week before stabilizing around 140 g‚m-3reactor‚h-1 by day 28. Integration of the data from all the experiments yielded a total hexane consumption and CO2 production of around 360 g (300 g C) and 496 g (135 g C), respectively. The average final biomass concentration was 177 ( 23 mgbiomass‚g-1dry perlite with a protein content of about 23%. According to these results, approximately 45% of the hexane introduced was mineralized and 10% of the hexane carbon consumed was incorporated as biomass (assuming 50% carbon in the biomass). The specific microbial activity of the PBR calculated from the average EC was 44.6 mg hexane‚g-1protein‚h-1.

The use of silicone oil to increase the mass transport and bioavailability of hexane to F. solani was an effective strategy (Table 2). As a significant example, the specific total hexane consumption rate obtained in batch culture with silicone oil was 2.3 times greater than that value without silicone. The use of fungi might improve the removal of hydrophobic compounds because their large aerial mycelia are in direct contact with the gas (4-6). This study shows that fungal activity can be further improved by the addition of a hydrophobic phase to the system. Hence, hexane uptake by fungi in biphasic systems likely undergoes direct assimilation from the gas phase as well as through the observed contact between the fungi and silicone oil. Both uptake mechanisms strongly reduce the mass transfer limitations. In our study, the observed microbial adhesion to silicone oil (Figure 5) was likely favored by the production of surface active molecules, such as the hydrophobins (21), by the fungi, which

Discussion

VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2393

FIGURE 5. SEM microphotographs for fungal packed bed bioreactors: (a) mycelia over silicone oil particle; (b) hyphae and bacteria. also stabilized the emulsion. This further increased hexane bioavailability and could have allowed direct uptake at the organic aqueous interface (22, 23). Besides increasing the hexane mass transfer, the use of silicone oil might also have improved hexane biodegradation by increasing oxygen transfer to the microorganisms as the solubility of oxygen in silicone oil (250 mg‚L-1) is around 30 times higher than in water (24). The selection of an appropriate and economical organic phase is an essential parameter for the performance of twophase bioreactors. The most significant characteristics of the organic phase are nonbiodegradability, biocompatibility, high solubilization capacity for the substrate, and nonmiscibility in water (12). Silicone oil was chosen on this basis as it was nonbiodegradable and nontoxic for F. solani and it exhibited an affinity for hexane 104 times greater than water (partition coefficient of hexane in water 30.9) (1). The susceptibility of F. solani to the solvents tested may result from the hydrophobic nature of hyphae, because the toxicity of organic solvents is partially related to their concentration in the cell membrane, which should increase with the hydrophobicity of the membrane in case of hydrophobic solvents (25). However, toxicity also depends on the molecular structure of the solvent used and the partition coefficient of octanolwater (12, 26). Considering that undecane was consumed, it was unexpected to find that 1-decanol inhibited the fungi. However, the solubility of 1-decanol in water is around 104 times higher than undecane and consequently may attain inhibitory levels. In fact, 1-decanol was not toxic to F. solani at a concentration 100 times lower than its solubility in water (37 mg‚L-1) (27). The hexane EC (165 g‚m-3reactor‚h-1) and RE (>90%) obtained with the biphasic PBR were greater than the 10150 g‚m-3reactor‚h-1 EC and the 40-60% RE previously obtained in bacterial and fungal biofilters (6, 8, 14). The 15% reduction of the EC recorded after day 30 of operation could be related with the instability of the emulsion and the leaching of the 2394

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006

oil with the added MSM. The maximum EC of 360 g‚m-3reactor‚h-1 achieved in the PBR was also higher than the maximum EC of 130 g‚m-3reactor‚h-1 previously reported in a similar system not provided with silicone oil (17). The mineralization obtained (45%) was lower than the 79% reported previously (17) but the carbon balance may be influenced by the solubilization of CO2 (24) and potential hexane oxidation intermediaries (hexanol, hexanal, and hexanoic acid) in the silicone oil phase. This was also observed in the batch tests as CO2 production was more important in the experiment without silicone oil. The maximum EC (120 g‚m-3reactor‚h-1) obtained in the biphasic STR at a load of 180 g hexane‚m-3reactor‚h-1 was similar to the values reported in classical biofilters, although the removal efficiencies (67%) and the maximum EC (580 g‚m-3reactor‚h-1) achieved in this study were greater. When considering the short effective gas residence time applied in the STR, the maximum ECg achieved in this system was much higher than ECg achieved in the PBs or by other authors such as Kibazohi (10), who reported an ECg of 94 g‚m-3gas‚h-1 for a bacterial biofilter packed with perlite ( ) 0.65). Hence, these results represent, to the best of our knowledge, the highest hexane ECg yet reported in the literature, confirming the potential of two-liquid-phase bioreactors (TLPB) for the abatement of hydrophobic air pollutants (15, 16). Davison and Daugulis (28) even obtained an EC up to 233 g‚m-3reactor‚h-1 for toluene in TLPB and predicted that a maximum EC of 1290 g‚m-3reactor‚h-1 could be achieved in such systems, which is much higher than the toluene EC of 145 and 350 g‚m-3reactor‚h-1 reported for conventional bacterial (29) or fungal biofilters (5), respectively. The lower EC obtained here with hexane was likely explained by the lower aqueous solubility of hexane (approximately 50 times lower than that of toluene) and to the different operating conditions tested. In our study, the reactor was operated at an organic/ aqueous phase rate of 1:9 and an agitation speed of 400 rpm, while Daugulis and co-workers (15, 16) used a ratio of 3.4:6.6 and an agitation speed of 800 rpm. The stirring rate is indeed a crucial parameter in the performance of STRs as it controls the transfer coefficients and the interfacial area of the emulsion. This was illustrated here by the fact that specific consumption rates obtained in batch experiments agitated at 150 rpm were lower than the values achieved in the STR agitated at 400 rpm (Table 2). However, high stirring rates might not be convenient in the case of fungal STRs since shear stress may affect the integrity of the fungal mycelia (30). The lower ECg recorded in the biphasic PBR than in the biphasic STR can therefore be explained by the lack of agitation in packed bed bioreactors, which would promote the coalescence of the oil particles and a reduction of the aqueous-organic interfacial area. Furthermore, laminar gas flow in biofilters generates reduced transfer coefficients (31). In the long term, macroscopic heterogeneities in biofilters may induce zone clogging or channeling and a further reduction in mass transfer. In contrast, stirring in baffled reactors promotes emulsion dispersion and increased mass transfer rates. These results therefore suggest that two-liquidphase systems are able to exceed the performance of PBRs when treating low- and high-VOCs gas streams. In addition, the use of TLPBs improves the mass transfer of VOC and oxygen to the microorganism, effectively utilizes the entire reactor volume, and allows a higher ECg (i.e., smaller units to treat the same mass loading). However, this improvement is accompanied by higher energy input that increases the operation costs. The specific hexane removal rate obtained in the STR with silicone oil was approximately 50% greater than that without oil and can be compared with the value of 339 mgHexane‚g-1protein‚h-1 reported by Morales et al. (32) for the

hexane degradation by cometabolism with methyl tert-butyl ether. The lower specific rate recorded in the PBR can be explained by the higher biomass content achieved in this system (approximately 180 mgbiomass‚g-1dry support), which could have induced inactive zones (6). This illustrates again the disadvantage of poorly homogenized bioreactors. Summarizing, the use of silicone oil enhanced hexane removal by F. solani in both stirred-tank and packed-bed bioreactors, which increased the performance of the biphasic systems by approximately 50% compared to their respective controls. This study presents the first reported case of the use of fungi in two-liquid-phase reactors, which broadens the application range of this technology. Solvent selection for the fungal strain was based on the same characteristics and methods proposed for bacteria and the critical log Pow below which the solvent was found toxic was between 4.6 and 6.5, which is in the same range of 5-7 reported for gram positive bacteria (33). The combined use of fungi and organic solvent led to a significant process improvement, allowing the highest hexane removal rates reported in the literature hitherto. Further work is needed to select less expensive organic phases, to increase stability, to optimize the solvent phase ratios, to model the process, and to demonstrate the feasibility of the system for other fungal species.

Acknowledgments We thank CONACYT (SEMARNAT-120-2001) for financing this work, and also Dr. Jose´ Sepulveda for SEM support.

Nomenclature EC: hexane elimination capacity (g‚m-3reactor‚h-1) ECg: hexane elimination capacity (g‚m-3gas‚h-1) Qg air flow (m3‚h-1) Sin and Sout inlet and outlet hexane concentration respectively (g‚m-3) Vreactor: reactor volume (m3reactor) Vg gas volume (m3gas) Greek Letters  bed void fraction (m3gas‚m3reactor) G gas hold up (m3gas‚m3reactor)

Literature Cited (1) Card, T. R. Fundamentals: Chemistry and Characteristics of Odors and VOCs; McGraw-Hill: New York, 1998. (2) Shareefdeen, Z.; Singh, A. Biotechnology for Odour and Air Pollution; Springer-Verlag: Heidelberg, Germany, 2005. (3) Van Groenestijn, J. W.; Hesselink, P. G. M. Biotechniques for air pollution control. Biodegradation 1993, 4, 283-301. (4) Van Groenestijn, J. W.; Liu, J. X. Removal of alpha-pinene from gases using biofilters containing fungi. Atmos. Environ. 2002, 36, 5501-5508. (5) Garcı´a-Pen ˜ a, E. I.; Herna´ndez, S.; Favela-Torres, E.; Auria, R.; Revah, S. Toluene biofiltration by the fungus Scedosporium apiospermum TB1. Biotechnol. Bioeng. 2001, 76, 61-69. (6) Arriaga, S.; Revah, S. Improving hexane removal by enhancing fungal development in a microbial consortium biofilter. Biotechnol. Bioeng. 2005, 90, 107-115. (7) Spigno, G.; Pagella, C.; Fumi, M. D.; Molteni, R.; De Faveri, D. M. VOCs removal from waste gases: gas-phase bioreactor for the abatement of hexane by Aspergillus niger. Chem. Eng. Sci. 2003, 58, 739-746. (8) Budwill, K.; Coleman, R. Project Report; Sustainable Forest Management Network. University of Alberta: Alberta, Canada, 1999. http://sfm1.biology.ualberta.ca/english/pubs/PDF/PR_19991.pdf/. (9) Paca, J.; Halecky, M.; Klapkova, E.; Maryska, M. Removal of aliphatic gasoline compounds in perlite biofilter. Proceedings of the 95th Annual Conference and Exhibiton; AWMA: Baltimore, MD, 2002. (10) Kibazohi, O.; Yun, S.; Anderson, W. A. Removal of hexane in biofilters packed with perlite and a peat-perlite mixture. World J. Microbiol. Biotechnol. 2004, 20, 337-343.

(11) Kennes, C.; Thalasso, F. Waste gas biotreatment technology. J. Chem. Technol. Biotechnol. 1998, 72, 303-319. (12) De´ziel, E.; Comeau, Y.; Villemur, E. Two-liquid-phase bioreactors for enhanced degradation of hydrophobic/toxic compounds. Biodegradation 1999, 10, 219-233. (13) Guieysse, B.; Cirne, M. d. D. T. G.; Mattiasson, B. Microbial degradation of phenanthrene and pyrene in a two-liquid phasepartitioning bioreactor. Appl. Microbiol. Biotechnol. 2001, 56, 796-802. (14) Van Groenestijn, J. W.; Lake, M. E. Elimination of alkanes from off-gases using biotrickling filters containing two liquid phases. Environ. Prog. 1999, 18, 151-155. (15) Yeom, S. H.; Daugulis, A. J. Development of a novel bioreactor system for treatment of gaseous benzene. Biotechnol. Bioeng. 2000, 72, 156-164. (16) Bruce, L. J.; Daugulis, A. J. Solvent selection strategies for extractive biocatalysis. Biotechnol. Prog. 1991, 7, 116-124. (17) Arriaga, S.; Revah, S. Removal of n-hexane by Fusarium solani with a gas-phase biofilter. J. Ind. Microbiol. Biotechnol. 2005, 32, 548-553. (18) Acun ˜ a, M. E.; Pe´rez, F.; Auria, R.; Revah, S. Microbiological and kinetic aspects of a biofilter for the removal of toluene from waste gases. Biotechnol. Bioeng. 1999, 63, 175-184. (19) Mun ˜ oz, R.; Guieysse, B.; Mattiasson, B. Phenanthrene biodegradation by an algal-bacterial consortium in two-phase partitioning bioreactors. Appl. Microbiol. Biotechnol. 2003, 61, 261267. (20) Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association/American Water Works Association/Water Environment Federation: Washington, DC, 1998. (21) Wo¨esten, H. A. B.; van Wetter, M. A.; Lugones, L. G.; van der Mei, H. C.; Busscher, H. J.; Wessels, J. G. H. How a fungus escapes the water to grow into the air. Curr. Biol. 1999, 9, 85-88. (22) Desai, J. D.; Banat, I. M. Microbial production of surfactants and their commercial potential. Microbiol. Mol. Biol. 1997, 61, 47-64. (23) Rosenberg, M. Basic and applied aspects of microbial adhesion at the hydrocarbon:water interface. Crit. Rev. Microbiol. 1991, 18, 159-173. (24) Jakob, F.; Jakob, K.; Perjanik, N. Silicone Dielectric Fluids; Analytical ChemTech Iternational, Inc.; 2004. http:// sfm-1.biology.ualberta.ca/english/pubs/PDF/PR_1999-1.pdf. http://www.weidman-acti.com/u/library/silicone_dielectric_ fluids.pdf. (25) Vermue¨, M.; Sikkema, J.; Verheul, A.; Bakker, R.; Tramper, J. Toxicity of homologous series of organic solvents for the grampositive bacteria Arthrobacter and Nocardia sp. and the gramnegative bacteria Acinetobacter and Pseudomonas sp. Biotechnol. Bioeng. 1993, 42, 747-758. (26) De Bont, J. A. M. Solvent-tolerant bacteria in biocatalysis. Trends Biotechnol. 1998, 16, 493-499. (27) Abraham, M. H.; Le, J. The correlation and prediction of the solubility of compounds in water using an atended solvation energy relationship. J. Pharm. Sci. 1999, 88, 868-880. (28) Davison, C. T.; Daugulis, A. J. Addressing biofilter limitations: A two-phase partitioning bioreactor process for the treatment of benzene and toluene contaminated gas streams. Biodegradation 2003, 14, 415-421. (29) Auria, R.; Frere, G.; Morales, M.; Acun ˜ a, M. E.; Revah, S. Influence of mixing and water addition on the removal rate of toluene vapors in a biofilter. Biotechnol. Bioeng. 2000, 68, 448-455. (30) Cui, Y. Q.; van der Lans, R. G. J. M.; Luyben, K. C. A. M. Effect of agitation intensities on fungal morphology of submerged fermentation. Biotechnol. Bioeng. 1997, 55, 715-726. (31) Blanch, H. W.; Clark, D. S. Biochemical Engineering; Marcel Dekker: New York, 1997. (32) Morales, M.; Vela´zquez, E.; Jan, J.; Revah, S.; Gonza´lez, U.; RazoFlores, E. Methyl tert-butyl ether biodegradation by microbial consortia obtained from soil samples of gasoline-polluted sites in Mexico. Biotechnol. Lett. 2004, 26, 269-275. (33) Inoue, A.; Horikoshi, K. Estimation of solvent-tolerance of bacteria by the solvent parameter log P. J. Ferment. Bioeng. 1991, 71, 194-196.

Received for review August 1, 2005. Revised manuscript received January 23, 2006. Accepted January 25, 2006. ES051512M VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2395