Biodegradation of gasoline: kinetics, mass balance and fate of individual hydrocarbons

Journal of Applied Microbiology 1999, 86, 1008 1016 Biodegradation of gasoline: kinetics, mass balance and fate of individual hydrocarbons F. Solano-Serena, R. Marchal, M. Ropars, J.-M. Lebeault 1 and J.-P. Vandecasteele Institut FrançaisduPétrole, Division Chimie et Physico-chimie appliquées, Département Microbiologie, Rueil- Malmaison, and 1 Université de Technologie de Compiègne, Centre de Recherches de Royallieu, Compiègne Cedex, France 6957/11/98: received 2 November 1998, revised 19 February 1999 and accepted 23 February 1999 F. SOLANO-SERENA, R. MARCHAL, M. ROPARS, J.-P. LEBEAULT AND J.-P. VANDECASTEELE. 1999. The degradation of gasoline by a microflora from an urban waste water activated sludge was investigated in detail. Degradation kinetics were studied in liquid cultures at 30 C by determination of overall O 2 consumption and CO 2 production and by chromatographic analysis of all 83 identifiable compounds. In a first fast phase (2 d) of biodegradation, 74% of gasoline, involving mostly aromatic hydrocarbons, was consumed. A further 20%, involving other hydrocarbons, was consumed in a second slow phase (23 d). Undegraded compounds (6% of gasoline) were essentially some branched alkanes with a quaternary carbon or/and alkyl chains on consecutive carbons but cycloalkanes, alkenes and C10- and C11-alkylated benzenes were degraded. The degradation kinetics of individual hydrocarbons, determined in separate incubations, followed patterns similar to those observed in cultures on gasoline. Carbon balance experiments of gasoline degradation were performed. The carbon of degraded gasoline was mainly (61 7%) mineralized into CO 2, the remaining carbon being essentially converted into biomass. INTRODUCTION Release of gasoline in the environment is a common occurrence arising principally from leaks of storage tanks (Dowd 1994). A major risk of gasoline release is contamination of drinking water sources by the mobile gasoline components which can migrate through the soil matrix (Leahy and Colwell 1990; Alvarez and Vogel 1991; McLinn and Rehm 1995). In all cases of soil pollution by gasoline, detailed knowledge of the microbiology of gasoline degradation is needed to decide and implement appropriate steps of action. First, it is necessary to know the biodegradability properties of all components of gasoline (intrinsic biodegradability). Next, the specific degradative capacities of the autochthonous microflora present at the site considered have to be evaluated. The information will allow, in particular, the assessment of the prospects for natural attenuation and the possibilities for Correspondence to: J.-P. Vandecasteele, Institut Français du Pétrole, Division Chimie et Physico-chimie appliquées, Département Microbiologie, 1 et 4 avenue de Bois-Préau, 92852 Rueil-Malmaison Cedex, France (e-mail: j-paul.vandecasteele@ifp.fr). in situ engineered bioremediation or on-site treatment by biofiltration (Jutras et al. 1997; Wright et al. 1997) of extracted soil vapours or aquifer water. The biodegradability of the most water-soluble components of gasoline, such as benzene (Paje et al. 1997), toluene (Leahy and Olsen 1997), ethylbenzene and xylene isomers (Di Lecce et al. 1997), compounds usually termed BTEX, has been clearly established using pure strains. Studies with complex microfloras have also confirmed the degradability properties of BTEX (Mallakin and Ward 1996; Matteau and Ramsay 1997). By contrast, little is known about the degradability of the other gasoline components except for some polyalkylated benzenes (Lang 1996; Rozkov et al. 1998) and some linear and branched alkanes (Schaeffer et al. 1979). The biodegradability of whole gasoline is not easy to assess since it consists of more than 200 identifiable components and since interaction between individual components may affect, in particular, the kinetics, as already pointed out for BTEX (Alvarez and Vogel 1991). Whole gasoline degradation has been studied in soil samples (Zhou and Crawford 1995) and in liquid cultures (Yerushalmi and Guiot 1998). These 1999 The Society for Applied Microbiology

BIODEGRADATION OF GASOLINE 1009 authors reported quite extensive overall biodegradation of sizeable amounts of gasoline using microfloras from contaminated soils but their studies concerned essentially total hydrocarbons rather than individual compounds. In a previous study, we developed a methodology for evaluating the intrinsic degradative capacities of a microflora (i.e. in the absence of any limiting condition) using a gasoline model mixture composed of 23 hydrocarbons (GM23) and applied it to the microflora of a non-polluted soil. This microflora degraded over 90% of GM23. In the present study, we extended our investigation to gasoline itself in order to obtain information on the fate of all identifiable hydrocarbons during biodegradation by activated sludge. Activated sludge was selected for this work as it constitutes the reference microflora for biodegradability studies and as such could allow the evaluation of the intrinsic biodegradability of gasoline itself, i.e. of all its constitutents. The points of interest included kinetic aspects and carbon balances of biodegradation. MATERIALS AND METHODS Culture media and inocula The vitamin-supplemented mineral salt medium described by Bouchez et al. (1995) was used as the nutrient solution. A gasoline cut (400 mg l 1 ) or individual hydrocarbons (150 mg l 1 ) were added to the nutrient solution as sole carbon source. The gasoline cut was a commercial unleaded gasoline topped at 76 C by distillation. It was free of hydrocarbons having less than six carbons and contained no oxygenated compounds. Inocula were prepared from an activated aerobic sludge sampled at an urban waste water treatment plant. A microbial suspension containing about 3 g l 1 dry weight was obtained by centrifugation of the sludge at 15 000 g for 20 min and resuspension of the centrifuged biomass in the same volume of nutrient solution. The microbial suspension was used to inoculate culture media at a final concentration of 100 mg dry weight l 1 unless otherwise stated. The centrifuged biomass could be used immediately or stored at 80 C for several months without significant loss of degradation capacity. Biodegradation of topped gasoline The biodegradation tests were performed in 500 ml flasks closed with Teflon-coated stoppers and with side-arms equipped with Mininert valves (Pierce, Oud-Beijerland, The Netherlands). Gasoline (25 ml) was added to 50 ml of inoculated culture medium through the valve with a Hamilton syringe. The precise quantities of gasoline introduced in each flask were determined by the difference in weight between the full and emptied syringe. After an incubation period of 25 d at 30 C with alternative shaking (70 strokes per min), 5mlofCH 2 Cl 2, containing 600 mg ml 1 dodecane as internal standard, was introduced into the flasks through the valve and the remaining hydrocarbons were extracted for 1 h under shaking. The flasks were refrigerated overnight at 4 C before opening and the suspensions were centrifuged at 35 000 g for 30 min at 4 C. The CH 2 Cl 2 phase of each flask was then analysed by gas chromatography. Experiments were performed in duplicate and abiotic controls containing 1 g l 1 HgCl 2 were run under similar conditions. Mineralization of topped gasoline The inoculated culture medium (18 ml) was introduced into 240 ml flasks closed by Viton rubber stoppers. Gasoline portions (9 ml) were weighed as described above and then dispensed into each flask with a 10 ml syringe. After incubation at 30 C for 25 d under alternative shaking, the flasks were acidified with 0 5 ml HNO 3 (68%) and CO 2 was determined according to the gas chromatographic procedure described below. Experiments were performed in triplicate. Gasolinefree flasks were run simultaneously for the determination of endogenous respiration. The mineralization yield of gasoline was estimated from the CO 2 produced in test flasks after correction for endogenous respiration. It was calculated as the molar ratio of the carbon in the CO 2 produced to the carbon in the gasoline consumed. Kinetics of O 2 consumption The kinetics of O 2 consumption during gasoline degradation were determined in duplicate at 30 C over 25 d with the electrolytic respirometric equipment D-12 Sapromat (Voith, Heidenheim, Germany). The 500 ml stirred culture flasks (300 rev min 1 ) contained 250 ml of inoculated culture medium and 125 ml of gasoline. Control experiments without gasoline were performed. O 2 consumption was also used for monitoring the degradation kinetics of hydrocarbons. Incubation was stopped at selected times and the remaining hydrocarbons were extracted as described above and analysed by gas chromatography. Kinetics of CO 2 production from individual hydrocarbons The kinetics of CO 2 production during the degradation of some individual hydrocarbons were studied at 30 C over 16 d. Cultures were carried out in 125 ml shaken flasks with 25 ml of nutrient solution containing 70 mg l 1 of inoculum biomass and 5 ml of hydrocarbon. Flasks were closed with Teflon-coated butyl stoppers and sealed with aluminium caps. CO 2 was determined at various times by gas chromatography as described below. The kinetics of endogenous respiration of the inoculum was determined in the same conditions with flasks incubated without hydrocarbon.

1010 F. SOLANO-SERENA ET AL. Chromatographic analyses Carbon dioxide was measured with a chromatograph equipped with a thermal conductivity detector and a Porapak Q column (80/100 mesh, 2 m) using an external standard method. The carrier gas was helium and the column temperature was 50 C. The temperature of the injector and that of the detector was 100 C. Samples (250 ml) of the headspace gas of culture flasks were withdrawn with a gas-tight syringe and injected into the chromatograph for CO 2 determination. Residual gasoline was analysed with a 3400 chromatograph (Varian, Sugarland, TX, USA) equipped with a flame ionization detector and a CP-Sil Pona CB column (0 25 mm 100 m) (Chrompack, Raritan, NJ, USA) according to the French standard method NF M07 086 (Durand 1998). The carrier gas was helium. The temperature of the injector was 250 C and that of the detector was 300 C. The column temperature was first set at 35 C for 10 min and was increased to 114 C at 1 1 C min 1, then to 280 C at 1 7 C min 1 and finally set at 280 C for 40 min. The identification of the gasoline components was performed using the Carburane software (Durand et al. 1995). The detection limit for each gasoline component was 0 15 mg l 1 of culture medium. Biomass analysis The final cell biomass was recovered by filtration of the cultures in respirometric flasks and was then freeze-dried and weighed. The elemental composition of dried biomass was determined by standard methods: ASTM D5291 for carbon, hydrogen and nitrogen and ASTM D5622 for oxygen. The initial cell biomass amount and composition were determined by the same method. Chemicals n-alkanes, cyclohexane, benzene, toluene and xylenes were purchased from Prolabo (Fontenay-sous-Bois, France). Other individual hydrocarbons and vitamins were from Fluka-Sigma (Saint-Quentin-Fallavier, France). RESULTS Oxygen consumption The kinetics of gasoline degradation were first investigated by means of O 2 consumption. The results of a 25 day experiment are presented in Fig. 1. Two main degradation phases were observed. First, a fast degradation phase (FDP), which started after an 18 h lag period and lasted until the 40th hour. The maximal rate of oxygen consumption was 44 0 mg l 1 h 1 and the average rate was 24 5 mg l 1 h 1. During this phase, a temporary Fig. 1 Kinetics of oxygen consumption during gasoline degradation. The oxygen consumption was determined for 25 d at 30 C using the electrolytic respirometric equipment as described in the Materials and Methods slow-down of the oxygen consumption rate could be noticed after 34 h of incubation. Chromatographic analyses showed that this decrease correlated with the exhaustion of toluene and ethylbenzene (data not shown) which were two of the most abundant compounds in the gasoline cut since they represented together up to 36 4% (w/w) of the total mixture. The FDP was then followed by a slow degradation phase (SDP) where the rate of oxygen consumption slowed down steadily from the 40th hour to the 25th day. The average rate, 15 mg l 1 d 1, was about 40 times lower than in the FDP. Degradation capacity of the microflora for gasoline In order to correlate oxygen consumption with substrate degradation, the residual components of gasoline were analysed at the end of the FDP and the SDP. The chromatographic patterns obtained are presented in Fig. 2. A large portion, 74% (w/w), of the initial amount of gasoline, corresponding to 24 identifiable components, was found to be consumed in the FDP, but only 20%, corresponding to 45 identifiable components, was degraded in the SDP. Degradation of all hydrocarbon classes (Table 1) began during the FDP. Aromatic hydrocarbons were mostly degraded in the FDP whereas linear, branched and cyclic alkanes were principally consumed in the SDP. This result clearly indicated that the degradation of aromatics was faster than that of linear, branched or cyclic alkanes. Table 2 shows the residual amounts of the identifiable hydrocarbons in culture media after 25 d of incubation for abiotic and for test flasks. All aromatics were extensively degraded, including the compounds bearing two alkyl chains in the ortho position (o-xylene) and tri- and tetra-substituted aromatics (Table 2). Linear alkanes, except in the case of

BIODEGRADATION OF GASOLINE 1011 chain, the faster the degradation. Branched alkanes with a methyl group on the second or third carbon were degraded at similar velocities (Fig. 3b). Degradation of mono- and dimethyl alkanes was faster than the degradation of trimethyl alkanes. Concerning aromatics (Fig. 3c), those with no or only one alkyl chain were readily degraded, even in the case of ethyl, n-propyl and iso-propyl substituents. Degradation of the polysubstituted aromatics bearing neighbouring substituents seemed to be slowed down by steric hindrance. For instance, 1-methyl 2-ethylbenzene degradation was slower than that of 1-methyl 3-ethylbenzene. A similar situation prevailed for 1,2,3-trimethylbenzene with respect to 1,2,4- trimethylbenzene which bears an isolated methyl group on the fourth carbon. For polyalkylated aromatics with the same number and localization of substituting groups, the compounds with an ethyl chain were less easily consumed than those with a methyl chain (1-methyl 2-ethylbenzene compared with o-xylene). For comparison, the degradative capacities of the microflora were tested using individual hydrocarbons as sole carbon source. The kinetics of degradation observed in these conditions were similar to those observed in incubations on gasoline. As shown in Fig. 4, the kinetics of CO 2 production clearly confirmed that BTEX (except o-xylene) were more readily mineralized than alkanes and alkenes. Different initial periods of adaptation of the microflora depending on the hydrocarbon used were consistently observed. Fig. 2 Chromatographic pattern of residual hydrocarbons of gasoline after 2 and 25 d of incubation. (a) Abiotic control (500 ml shaken flask) after 25 d; (b) test flask (respirometric flask) after 2 d; (c) test flask (500 ml shaken flask) after 25 d undecane which is somewhat puzzling, alkenes with five to nine carbons, cyclohexane and substituted cyclopentanes were also widely degraded. Finally, the residual components of gasoline were mostly branched alkanes. In particular, trimethyl alkanes with either a quaternary carbon (2,2,4-trimethylpentane, 2,2,5-trimethylhexane) or alkyl chains on consecutive carbons (2,3,4-trimethylpentane) or both (2,3,3- trimethylpentane) appeared to be the most recalcitrant compounds of gasoline. Hydrocarbons belonging to the same class were found to be consumed at various rates (Fig. 3). For linear and branched alkanes, except for undecane, the longer the main carbon Carbon balance Extensive degradation of gasoline (94%) was observed after 25 d of incubation. The carbon balance of gasoline degradation (Table 3) shows that 61 7% of gasoline degraded was mineralized into CO 2 and that microbial cell production accounted for the remaining carbon of gasoline degraded. The biomass yield based on degraded carbon was 0 63 mg mg 1, i.e. 0 37 Cmol Cmol 1. The rate of carbon recovery was 0 98 indicating that production of metabolites of incomplete degradation was, at the most, a minor phenomenon in the process. As topped gasoline had a carbon content of 893 mg g 1, it could be represented by the mean molecular formula C 7 H 10. The stoichiometric equation for biomass production from gasoline, with the assumptions that NH 3 was the sole nitrogen source and that biomass had the relative composition suggested by McCarty (1972), could then be written: C 7 H 10 5 2 O 2 7 5 NH 3 : 7 5 C 5H 7 NO 2 11 5 H 2O (1) The stoichiometric equation for CO 2 production is: C 7 H 10 19 2 O 2 : 7CO 2 5H 2 O (2)

1012 F. SOLANO-SERENA ET AL. Degradation rate (%)* Amount in gasoline Hydrocarbon classes (mg g 1 ) After 2 d After 25 d Aromatics 789 88 99 2 1 Branched alkanes 165 14 74 2 5 Linear alkanes 23 17 92 2 1 Cyclic alkanes 17 10 99 2 1 Alkenes 6 71 99 2 1 * The degradation rate was determined as the ratio of the amount of hydrocarbon degraded to the initial amount. Mean value of two respirometric flasks. Mean value of five shaken flasks. The standard deviations of the data are indicated. Table 1 Degradation of hydrocarbon classes after 2 and 25 d of incubation The theoretical ratios of O 2 consumed to carbon degraded (O 2 /C) for biomass formation (eqn (1)) and CO 2 production (eqn (2)) are, respectively, 0 95 and 3 62 mg O 2 mg C 1. Experimental ratios measured for the FDP and the SDP being, respectively, 1 97 and 3 43, it could be concluded that microbial growth mainly occurred during the FDP. Furthermore, from the respective contributions of growth and mineralization (data of Table 3) and from the corresponding theoretical O 2 /C ratios, a theoretical overall O 2 consumption could be calculated for the experiment presented in Fig. 1. The ratio of this theoretical O 2 consumption to the experimental O 2 consumption was 0 94, indicating that eqns (1) and (2) and their respective contributions to the overall degradation phenomenon gave a satisfactory description of gasoline degradation. DISCUSSION Gasoline was found to be degraded up to 94% under nonlimiting conditions using as the microbial inoculum activated sludge from a waste water treatment plant. Carbon balances showed that CO 2 and biomass were the main products. The different biodegradation velocities of gasoline components appeared to account for the occurrence of two separate stages in the culture kinetics, a fast one, the FDP, followed by a slow one, the SDP. These two stages were made clearly apparent by respirometry. The ratio (O 2 /C) of O 2 consumed to carbon degraded was substantially lower in the FDP than in the SDP confirming that biomass formation occurred along with mineralization mainly in the first phase whereas during the second phase essentially mineralization occurred. Considering the overall degradation process, mineralization was the prevailing phenomenon (61 7% of carbon converted into CO 2 ). The overall yield in biomass (0 63 mg mg 1 or 0 37 Cmol Cmol 1 ) was consistent with those previously reported for various hydrocarbons (Wodzinski and Johnson 1968; Geerdink et al. 1996; Yerushalmi and Guiot 1998). The study of the fate of individual hydrocarbons during gasoline biodegradation required a methodology for quantitative determination of each of them in incubation mixtures. The convenient headspace sampling method used by other authors (Zhou and Crawford 1995; Yerushalmi and Guiot 1998) was acceptable for total hydrocarbons but not for detailed analysis of individual hydrocarbons since the values of the Henry s law constants which relate the amounts of hydrocarbons in the gas phase to their concentrations in the aqueous phase can differ from one hydrocarbon to another by several orders of magnitude (Nirmalakhandan et al. 1997). The methodology used was more complex and time-consuming but yielded reproducible results. These results emphasized the extensive degradation capacities of activated sludge already apparent in the mineralization yields observed. All hydrocarbon classes, i.e. aromatic compounds, linear, branched and cyclic alkanes as well as alkenes, were extensively degraded after 25 d. For each hydrocarbon class, degradation occurred at a specific rate. The aromatics were found to be the most readily consumed in agreement with the results reported for some individual aromatics, such as benzene, toluene, ethylbenzene, m- and p-xylene (Zhou and Crawford 1995; Nielsen et al. 1996; Yerushalmi and Guiot 1998). For other polyalkylated aromatics, the nature and relative positions of alkyl chains on the ring significantly influenced the rate of biodegradation. Concerning alkanes, cyclic alkanes were found to be degraded although Ridgway et al. (1990) noted the low occurrence of pure strains able to grow with cyclic alkanes as sole carbon source. Moreover, the simultaneous degradation of iso- and n-alkanes gave further evidence of the high efficiency of the activated sludge microflora compared with microfloras of other origins where degradation of iso-alkanes took place only after total exhaustion of n- alkanes (Geerdink et al. 1996; Solano-Serena et al. 1998).

BIODEGRADATION OF GASOLINE 1013 Table 2 Residual hydrocarbons in cultures after gasoline degradation Amount in Amount in Retention time abiotic flasks test flasks Components (min)* (mg l 1 ) (mg l 1 ) Hexane 10 61 0 16 0 00 Heptane 19 47 5 06 0 00 Octane 33 00 1 59 0 00 Nonane 48 69 0 41 0 00 Decane 64 52 0 34 0 00 Undecane 79 70 0 69 0 70 2,2-Dimethylpentane 12 21 0 11 0 00 3,3-Dimethylpentane 12 71 0 57 0 14 2,4-Dimethylpentane 14 79 0 28 0 03 2-Methylhexane 15 98 3.32 0 00 2,3-Dimethylpentane 16 14 1 64 0 12 3-Methylhexane 16 84 4 82 0 03 3-Ethylpentane 17 90 0 64 0 00 2,2,4-Trimethylpentane 18 21 18 45 9 17 2,2-Dimethylhexane 22 11 0 37 0 00 2,5-Dimethylhexane 23 39 2 78 0 22 2,4-Dimethylhexane 23 63 3 30 0 26 3,3-Dimethylhexane 24 51 0 32 0 00 2,3,4-Trimethylpentane 25 65 4 85 1 48 2,3,3-Trimethylpentane 26 20 4 38 3 16 2,3-Dimethylhexane 27 18 2 23 0 05 2-Methyl 3-ethylpentane 27 33 0 17 0 00 2-Methylheptane 28 05 1 79 0 00 4-Methylheptane 28 26 1 14 0 00 3-Methylheptane 29 11 1 89 0 00 3-Ethylhexane 29 24 0 61 0 00 2,2,5-Trimethylhexane 30 67 0 81 0 31 2,3,5-Trimethylhexane 35 40 0 16 0 02 2,4-Dimethylheptane 36 59 0 17 0 00 2,5-Dimethylheptane 38 60 0 22 0 00 4-Methyloctane 43 04 0 26 0 00 2-Methyloctane 43 20 0 27 0 00 3-Methyloctane 44 25 0 40 0 00 Methyl ethyloctane 72 55 0 54 0 00 C12-Isoparaffin 73 31 0 70 0 22 C12-Isoparaffin 74 39 0 96 0 00 1-Methylcyclopentane 12 42 0 37 0 00 Cyclohexane 15 11 0 68 0 00 1,3-Dimethylcyclopentane 17 76 0 36 0 00 1,2-Dimethylcyclopentane 18 03 0 31 0 00 1-Methylcyclohexane 21 72 1 06 0 03 1-Ethylcyclopentane 23 19 0 30 0 00 1,2,3-Trimethylcyclopentane 25 25 0 19 0 00 1,2,4-Trimethylcyclopentane 28 43 0 67 0 00 Benzene 14 32 2 20 0 00 Toluene 26 20 106 82 0 00 Ethylbenzene 40 14 16 15 0 00 m-xylene 41 51 40 34 0 14 p-xylene 41 67 17 16 0 00 o-xylene 44 98 21 66 0 03 iso-propylbenzene 50 26 0 98 0 00

1014 F. SOLANO-SERENA ET AL. Amount in Amount in Retention time abiotic flasks test flasks Components (min)* (mg l 1 ) (mg l 1 ) n-propylbenzene 54 93 4 03 0 00 1-Methyl 3-ethylbenzene 56 17 13 93 0 00 1-Methyl 4-ethylbenzene 56 47 6 13 0 00 1,3,5-Trimethylbenzene 57 40 7 17 0 39 1-Methyl 2-ethylbenzene 58 85 5 08 0 62 1,2,4-Trimethylbenzene 61 24 20 97 0 00 1,2,3-Trimethylbenzene 65 40 4 24 0 00 1-Methyl 4-isopropylbenzene 65 77 0 35 0 00 C10-Aromatic 71 36 1 29 0 00 1,3-Dimethyl 4-ethylbenzene 74 15 0 84 0 11 1-Methylindane 74 71 0 50 0 21 1,2-Dimethyl 4-ethylbenzene 75 29 1 68 0 00 1,2-Dimethyl 3-ethylbenzene 78 20 0 46 0 00 1,2,3,5-Tetramethylbenzene 80 04 1 07 0 00 1,2,4,5-Tetramethylbenzene 80 53 1 33 0 00 C11-Aromatic 82 85 0 42 0 00 4-Methylindane 84 25 0 62 0 00 C11-Aromatic 84 92 0 42 0 00 Naphthalene 88 03 1 04 0 00 Pentene 6 51 0 26 0 00 C7-Alkene 19 28 0 27 0 00 C7-Alkene 19 71 0 25 0 00 C7-Alkene 19 83 0 15 0 00 C7-Alkene 19 83 0 15 0 00 C7-Alkene 20 05 0 16 0 00 C7-Alkene 20 57 0 33 0 00 C7-Alkene 21 00 0 17 0 00 C8-Alkene 31 29 0 13 0 00 C8-Alkene 32 00 0 15 0 00 C8-Alkene 33 54 0 14 0 00 C9-Alkene 49 15 0 17 0 00 The final amounts of identified hydrocarbons were determined after 25 d of incubation at 30 C. The detection limit of each individual component was 0 15 mg l 1. The mean value of total amount of detected compounds after 25 d was 19 7 2 3 5 mg l 1. * Retention times are mentioned for each identifiable component according to chromatographic analyses described in the Materials and Methods. Mean value of two abiotic shaken flasks. Mean value of five shaken flasks. Table 2 Continued As n-alkanes and iso-alkanes have a low solubility in aqueous media, the gaseous phase of each flask constitutes a stock of substrate which is not directly available to the microflora. Nevertheless, as biodegradation proceeds, transfer of the hydrocarbon from the gas phase to the liquid phase occurs. The more volatile the hydrocarbons are, the more slowly they are transferred to the liquid. In this work, the degradation rate of n-alkanes was found to increase with the carbon chain length. As vapour pressure decreases with carbon number, the transfer of the substrate from the gas phase to the aqueous phase might therefore be the rate-limiting step of the degradation of n-alkanes in the flasks. For branched alkanes, the incomplete biodegradation of some of them indicated that, besides substrate transfer, metabolic activity could limit the degradation phenomenon. The intrinsic degradative capacities of the microflora used, which were globally excellent, clearly suffered from a limitation for structures with a quaternary carbon or/and alkyl chains on consecutive carbons

BIODEGRADATION OF GASOLINE 1015 Fig. 4 Kinetics of carbon dioxide production during degradation of individual hydrocarbons. (R), toluene; (), benzene; (r), hexane; (E), cyclohexane; (Ž), 3-methylpentane; (), trans-2- pentene; (ž), o-xylene; (e), control without hydrocarbon. Each hydrocarbon was incubated separately in 125 ml shaken flasks as described in the Materials and Methods Table 3 Carbon balance of gasoline degradation Initial amount Final amount Substrate or products (mg C l 1 ) (mgcl 1 ) Gasoline 357 18 Biomass* 39 165 CO 2 0 204 Total carbon 396 387 The final amounts were determined after 25 d of incubation at 30 C as described in the Materials and Methods. * The initial and final proportions of carbon in biomass were respectively 47% and 51 5%. Mean value of five shaken flasks. The standard deviation of the data was 24. Mean value of three shaken flasks. The standard deviation of the data was 230. Fig. 3 Degradation of some representative hydrocarbons during gasoline degradation. (a) Linear alkanes: (r), hexane; (), heptane; (e), octane; (), nonane; (t), decane. (b) Branched alkanes: (r), 2,2,4-trimethylpentane; (), 2-methylhexane, 3- methylhexane; (), 2,5-dimethyloctane; (T), 3-methyloctane. (c) Aromatics: (), 1,2,3-trimethylbenzene; (), 1-methyl 2-ethylbenzene; (e), o-xylene; (r), benzene, toluene, ethylbenzene, m-xylene, 1,2,4-trimethylbenzene, 1-methyl 3- ethylbenzene. Cultures were performed in parallel in respirometric flasks stopped after 2 d of incubation and in 500 ml shaken flasks incubated for 25 d which were more recalcitrant. Such a limitation has already been observed for a soil microflora but the existence of specific iso-alkane-degrading microfloras has also been demonstrated (Solano-Serena et al. 1998). Thus, because of its limitation concerning iso-alkane degradation, the activated sludge microflora used did not express the whole intrinsic biodegradability of gasoline. From an applied point of view, the results indicate that, for remediation purposes, the microbiological treatment has to take into account the slow degradation of the branched alkanes. In fact, reinforcement of the microbial population with a specific iso-alkane-degrading microflora has been shown to be possible (Solano-Serena et al.

1016 F. SOLANO-SERENA ET AL. 1998) and should be considered whatever the remediation process (Van Limbergen et al. 1998). ACKNOWLEDGEMENTS The authors thank F. Haeseler for advice on the use of the respirometric equipment. REFERENCES Alvarez, P.J.J. and Vogel, T.M. (1991) Substrate interactions of benzene, toluene and para-xylene during microbial degradation by pure cultures and mixed culture aquifer slurries. Applied and Environmental Microbiology 57, 2981 2985. Bouchez, M., Blanchet, D. and Vandecasteele, J.P. (1995) Degradation of polycyclic aromatic hydrocarbons by pure strains and by defined strain associations: inhibition phenomena and cometabolism. Applied Microbiology and Biotechnology 43, 156 164. Di Lecce, C., Accarino, M., Bolognese, F., Galli, E. and Barbieri, P. (1997) Isolation and metabolic characterization of a Pseudomonas stutzeri mutant able to grow on the three isomers of xylene. Applied and Environmental Microbiology 63, 3279 3281. Dowd, R.M. (1994) Leaking underground storage tanks. Environmental Science and Technology 18, 10. Durand, J.P. (1998) Le rôle de la CPG dans l industrie pétrolière et la pétrochimie. De l analyse détaillée des hydrocarbures à la distillation simulée. Analusis Magazine 26, M17 M21. Durand, J.P., Béboulène, J.J. and Ducrozet, A. (1995) Detailed characterization of petroleum products with capillary GC analyzers. Analusis 23, 481 483. Geerdink, M.J., van Loosdrecht, M.C.M. and Luyben, K.Ch.A.M. (1996) Biodegradability of diesel oil. Biodegradation 7, 73 81. Jutras, E.M., Smari, C.M., Rupert, R., Pepper, I.L. and Miller, R.M. (1997) Field-scale biofiltration of gasoline vapors extracted from beneath a leaking underground storage tank. Biodegradation 8, 31 42. Lang, E. (1996) Diversity of bacterial capabilities in utilizing alkylated benzenes and other aromatic compounds. Letters in Applied Microbiology 23, 257 260. Leahy, J.G. and Colwell, R.R. (1990) Microbial degradation of hydrocarbons in the environment. Microbiological Review 54, 305 315. Leahy, J.G. and Olsen, R.H. (1997) Kinetics of toluene degradation by toluene-oxidizing bacteria as a function of oxygen concentration, and the effect of nitrate. FEMS Microbiology Ecology 23, 23 30. Mallakin, A. and Ward, O.P. (1996) Degradation of BTEX compounds in liquid media and in peat biofilters. Journal of Industrial Microbiology 16, 309 318. Matteau, Y. and Ramsay, B. (1997) Active biofiltration of toluene. Biodegradation 8, 135 141. McCarty, P.L. (1972) Energetics of organic matter degradation. In Water Pollution Microbiology ed. Mitchell, R. pp. 91 118. New- York: Wiley Interscience. McLinn, E.L. and Rehm, B.W. (1995) Biodegradation of petroleum hydrocarbons in fractured, unsatured dolomite at a field site. Ground Water Monitoring Review 15, 73 80. NF07 086 Détermination des Teneurs en Familles Chimiques d Hydrocarbures dans les Essences pour Moteur Automobile à Partir de l Analyse Détaillée. Paris: Association Française de Normalisation. Nielsen, P.H., Bjerg, P.L., Nielsen, P., Smith, P. and Christensen, T.H. (1996) In situ and laboratory determined first-order degradation rate constants of specific organic compounds in an aerobic aquifer. Environmental Science and Technology 30, 31 37. Nirmalakhandan, N., Brennan, R.A. and Speece, R.E. (1997) Predicting Henry s law constant and the effect of temperature on Henry s law constant. Water Research 31, 1471 1481. Paje, M.L.F., Neilan, B.A. and Couperwhite, I. (1997) A Rhodococcus species that thrives on medium saturated with liquid benzene. Microbiology 143, 2975 2981. Ridgway, H.F., Safarik, J., Phipps, D., Carl, P. and Clark, D. (1990) Identification and catabolic activity of well-derived gasolinedegrading bacteria from a contaminated aquifer. Applied and Environmental Microbiology 56, 3565 3575. Rozkov, A., Käärd, A. and Vilu, R. (1998) Biodegradation of dissolved jet fuel in chemostat by a mixed bacterial culture isolated from a heavily polluted site. Biodegradation 8, 363 369. Schaeffer, T.L., Cantwell, S.G., Brown, J.L., Watt, D.S. and Fall, R.R. (1979) Microbial growth on hydrocarbons: Terminal branching inhibits biodegradation. Applied and Environmental Microbiology 38, 742 746. Solano-Serena, F., Marchal, R., Blanchet, D. and Vandecasteele, J.P. (1998) Intrinsic capacities of soil microflor% for gasoline degradation. Biodegradation, 9, 319 326. Van Limbergen, H., Top, E.M. and Verstaete, W. (1998) Bioaugmentation in activated sludge: current features and future perspectives. Applied Microbiology and Biotechnology 50, 16 23. Wodzinski, R.S. and Johnson, M.J. (1968) Yields of bacterial cells from hydrocarbons. Applied Microbiology 16, 1886 1891. Wright, W.F., Schroeder, E.D., Chang, D.P.Y. and Romstad, K. (1997) Performance of a pilot-scale compost biofilter treating gasoline. Journal of Environmental Engineering 123, 547 555. Yerushalmi, L. and Guiot, S.R. (1998) Kinetics of biodegradation of gasoline and its hydrocarbon constituents. Applied Microbiology and Biotechnology 49, 475 481. Zhou, E. and Crawford, R.L. (1995) Effects of oxygen, nitrogen and temperature on gasoline biodegradation in soil. Biodegradation 6, 127 140.