Article pubs.acs.org/est
Aerobic Biodegradation Kinetics and Mineralization of Six Petrodiesel/Soybean-Biodiesel Blends Mohamad H. Yassine,† Shuyun Wu,† Makram T. Suidan,*,‡ and Albert D. Venosa§ †
Division of Environmental Engineering, School of Energy, Environmental, Biological, and Medical Engineering, University of Cincinnati, Ohio 45221, United States ‡ Faculty of Engineering and Architecture, American University of Beirut, P.O. Box 11-0236 Riad El-Solh, Beirut, Lebanon 1107 2020 § U.S. Environmental Protection Agency, 26 W. Martin Luther King Drive, Cincinnati, Ohio 45268, United States S Supporting Information *
ABSTRACT: The aerobic biodegradation kinetics and mineralization of six petrodiesel/soybean-biodiesel blends (B0, B20, B40, B60, B80, and B100), where B100 is 100% biodiesel, were investigated by acclimated cultures. The fatty acid methyl esters (FAMEs) of biodiesel were found to undergo rapid abiotic transformation in all experiments. The C10−C21 n-alkanes of petrodiesel were metabolized at significantly higher microbial utilization rates in the presence of biodiesel. The rates of mineralization of the blends were also enhanced in the presence of biodiesel; yet a similar enhancement in the extent of mineralization was not observed. Abiotic fuel-blends/aqueous-phase equilibration experiments revealed that the FAMEs of biodiesel were capable of cosolubilizing the n-alkanes of petrodiesel, a mechanism that fully explains the faster utilization and mineralization kinetics of petrodiesel in the presence of biodiesel without necessarily enhancing the extent of biomineralization. The biodegradation of six targeted aromatic compounds present in petrodiesel was also influenced by the amount of biodiesel in a blend. While toluene, o-xylene, and tetralin were not degraded in the B0 and B20 treatments, all of the targeted aromatic compounds were degraded to below detection limits in the B40 and B80 treatments. Biomass acclimated to B60, however, was unable to degrade most of the aromatic compounds. These results indicate that the amount of biodiesel in a blend significantly affects the absolute and relative abundance of the dissolved and bioavailable constituents of biodiesel and petrodiesel in a way that can considerably alter the biodegrading capacity of microbial cultures. greenhouse gases, from biodiesel combustion.4 A 2002 EPA report showed that a 20% biodiesel blend boosted NOx emissions by 2% but cut particulate matter (PM) by about 10%, hydrocarbons (HC) by about 21%, and carbon monoxide (CO) by about 11%.5 Because of its potential as a renewable energy source, as well as a method to reduce net greenhouse gas emissions, biodiesel has attracted great interest as an alternative fuel in the transportation industry. The great benefits of using biodiesel as a fuel demand a more comprehensive understanding of its behavior in the environment. Although numerous reports have confirmed that biodiesel is more readily biodegradable than petrodiesel,6−8 little is known about how biodiesel blending would affect the biodegradation of petrodiesel. There appears to be an ongoing debate on whether biodiesel would enhance the microbial utilization of petrodiesel or not. If such enhancement does exist, the mechanisms by which it is achieved are inadequately understood. Zhang et al.,8 Pasqualino et al.,9 and Chen et al.10
1. INTRODUCTION Biodiesel is defined as the monoalkyl esters of long chain fatty acids derived from transesterification of vegetable oils and animal lipids for use in compression-ignition (petrodiesel) engines. This specification is for pure (100%) biodiesel prior to blending with petrodiesel fuel. Petroleum-derived diesel is composed of 75−85% saturated hydrocarbons (primarily normal, branched, and cycloalkanes), and 15−25% aromatic hydrocarbons (mainly alkylbenzenes). The average chemical formula for common petrodiesel fuel is C12H26, ranging from approximately C10H22 to C15H32.1 A major reason for the renewed interest in fuels made from plant and animal lipids is that feedstocks for the fatty-acid methyl esters (FAMEs) of biodiesel are renewable or can be reclaimed from wastes (e.g., used animal fats, recycled frying oil). In contrast, fossil fuels are nonrenewable resources and have increasingly higher costs both financially and environmentally. An additional incentive for using biodiesel is the potential positive impact on global climate change. Various greenhouse emission models estimate lower net crop-tocombustion carbon and sulfur emissions for biodiesel compared to petrodiesel.2,3 However, several studies have confirmed a net increase of NOx emissions, which are considered indirect © 2013 American Chemical Society
Received: Revised: Accepted: Published: 4619
January 24, 2013 March 27, 2013 April 3, 2013 April 3, 2013 dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
were delivered to the bioreactors using Hamilton syringe pumps, and the feed rates were increased gradually throughout an acclimation period of six months to a final organic loading of 750 mg/L-d fuel. Adequate aeration in the bioreactors was maintained throughout the operating period by means of continuous air diffusers, operated at nominal volumetric flow rates of 2−3 L/min. The bioreactors received a combined feed of essential nutrient and vitamin minimal medium buffered at 7.5 ± 0.1 of the following final concentration: 150 mg/L MgSO4·7H2O, 156 mg/L KNO3, 48.7 mg/L CaCl2·2H2O, 37.3 mg/L FeCl2·4H2O, 1393 mg/L K2HPO4, 544 mg/L KH2PO4, 0.08 mg/L CuSO4·H2O, 0.15 mg/L Na2MoO4·2H2O, 0.13 mg/ L MnSO4·H2O, 0.23 mg/L ZnCl2, 0.42 mg/L CoCl2·6H2O, 0.015 mg/L 4-aminobenzoic acid (99%), 0.00585 mg/L biotin, 0.0003 mg/L cyanocabalamin, 0.00585 mg/L folic acid dihydrate (99%), 0.015 mg/L nicotinic acid (98%), 0.015 mg/L pantothenic acid Ca-salt hydrate (98%), 0.03 mg/L pyridoxine hydrochloride (98%), 0.015 mg/L riboflavin (98%), 0.015 mg/L thiamine hydrochloride (99%), and 0.015 mg/L thiotic acid (98%). The same nutrients growth medium was used in the batch experiments. The buffering capacity of the growth medium was designed and tested to accommodate for negligible pH drop due to accumulation of degradation products during the batch experiment even after 42 d of incubation. Certain water quality variables including effluent FAMEs and target n-alkane analytes, total and volatile suspended solids (VSS), chemical oxygen demand (COD), pH, and dissolved oxygen (DO) were monitored routinely. When the reactors achieved steady state performance, a portion of each reactor’s culture was collected, washed three times with sterile saline solution, and frozen in 10% w/w glycerol at −80 °C for consistent use in the batch experiments. 2.3. Batch Experiments. Batch experiments were carried out in sterile and surface-deactivated 160 mL glass serum bottles where 50 mL of filter-sterilized nutrient and vitamin medium was added. Biomass aliquots acclimated to each of the fuel blend, thawed, and washed three times with sterile saline solution were added to the bottles at a measured solid concentration of 100 ± 15 mg VSS/L. The bottles were sealed with PTFE septa and spiked with 10 μL of their respective fuel blends, corresponding to an initial organic substrate concentration of 175 ± 3 mg/L fuel. The substrate-to-biomass ratio used in this study (approximately ∼1.75) resembles “extant” kinetic conditions according to the terminology proposed by Grady and co-workers,21 indicating that the conditions simulated in these batch experiments could manifest the conditions in the source environment from which the biomass is obtained (i.e., inside the bioreactors). The amount of oxygen available in the headspace of the bottles was sufficient for the complete oxidation of 1.5X the amount of fuel added on COD basis. The bottles were then placed in a rotary tumbler operated at a nominal speed of 20−25 rpm. Optimum sampling frequency was determined through range finding experiments that were every 4 h for the first 24 h, every 6 h for the second 24 h, every 8 h for the next 24 h, every 12 h for the next 24 h, and every 24 h thereafter for 168 h (7 days). Separate bottles were set for the analysis of CO2 and VSS at 7 and 42 d. Abiotic controls inactivated with sodium azide (0.9% w/v) as well as biomass controls were also included to quantify abiotic substrate losses and biomass endogenous decay and respiration rates, respectively. All experiments were conducted at 22 ± 2 °C.
claimed to observe cometabolic enhancement in the mineralization of petrodiesel by unacclimated cultures when blended with various types of biodiesel feedstocks. Mariano et al.11 and Owsianiak et al.12 did not observe such cometabolic effect in the mineralization of petrodiesel/biodiesel blends by microbial enrichments isolated from contaminated sites. Prince et al.13 and DeMello et al.14 found that the biodegradation rate of FAMEs, in petrodiesel/biodiesel blends by unacclimated fresh and seawater microorganisms, respectively, was roughly the same as the biodegradation rate of the n-alkanes. Owsianiak et al.,12 in turn, found that for petrodiesel/biodiesel blends of less than 10% biodiesel, the microorganisms preferentially metabolized the FAMEs and slowed the assimilation of the n-alkanes. Miller and Mudge15 found that adding biodiesel to a sand column contaminated with crude oil significantly enhanced the mobility of the oil along with the microbial uptake of aliphatic hydrocarbons. They suggested that this apparent enhancement was probably achieved through cosolubilization rather than cometabolism. In fact, biodiesel’s capacity to mobilize oil and coal tar in sediments was tested in microcosms, mesocosms, and field scale studies,16−18 and was found to significantly enhance the mobility of the contaminants. Taylor and Jones16 and Alvarez et al.18 further found that the addition of biodiesel to the contaminated sediments markedly enhanced microbial utilization of the contaminants, compared to adding nutrients or microbial enrichment alone. In this study, we investigated the impact of soybean biodiesel blending level on the aerobic biodegradation kinetics and mineralization of petrodiesel/ biodiesel blends B0, B20, B60, B80, and B100 (where B100 is 100% unblended biodiesel) by acclimated cultures.
2. MATERIALS AND METHODS 2.1. Chemicals. Low-sulfur petroleum diesel (B0) was purchased from a BP diesel station (Cincinnati, OH) with an nalkane mole fraction of 0.165. Unblended soybean-methyl ester biodiesel (B100) was purchased from Peter Cramer North America (Cincinnati, Ohio) with FAMEs mole fractions of 0.145 C16:0, 0.055 C18:0, 0.206 C18:1, 0.518 C18:2, and 0.0759 C18:3. All other fuel blends (B20, B40, B60, and B80) were blended in our laboratory by volumetric splash mixing. Palmitic acid methyl ester (99%) (C16:0-ME), palmitoleic acid methyl ester (99%) (C16:1-ME), stearic acid methyl ester (99%) (C18:0-ME), oleic acid methyl ester (99%) (C18:1ME), linoleic acid methyl ester (99%) (C18:2-ME), linolenic acid methyl ester (99%) (C18:3-ME), and n-alkanes standard mixture (nC10-nC30), were all purchased from Sigma Aldrich (USA). All other chemicals, if not noted, were purchased from Sigma Aldrich or Fisher Scientific (USA). 2.2. Enrichments and Acclimation. Bacterial enrichment was obtained from the activated sludge tank of Mill Creek Wastewater Treatment Plant (Cincinnati, OH) and was mixed with two cultures previously developed in our laboratory at the University of Cincinnati (Cincinnati, OH) capable of degrading gasoline19 and triglycerides.20 The mixed consortium was washed 5 times with sterile saline solution (0.85% NaCl) and was inoculated in six laboratory-scale 6-L porous pot (PP) chemostat bioreactors at initial volatile suspended solids (VSS) concentration of ∼500 mg/L. A schematic of the PP chemostat bioreactors used in this study is shown in Figure S1 in the Supporting Information (SI) section. The flow-through bioreactors were operated as completely mixed systems with a hydraulic retention time of 1 day and complete retention of biomass. Organic feeds of B0, B20, B40, B60, B80, or B100 4620
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
2.4. Chemical Analysis. The chemical analysis of the FAMEs and n-alkanes in the bioreactors’ effluent as well as the kinetic batch experiments followed the same procedure. For the batch experiments, at each sampling event, triplicate bottles of each experimental treatment were spiked with 50 μL of a standard surrogate solution of C16:1-ME and deuterated nC16 in acetone. The bottles were then completely liquid−liquid extracted twice with 1:1 dichloromethane (DCM), and the extract was filtered through anhydrous sodium sulfate and stored at −20 °C until the time of analysis. Extracts were added in 1 mL volumes in GC vials and spiked with 10 μL of an internal standard solution containing C11:0-ME, C13:0-ME, deuterated nC10, and deuterated nC20 in DCM and were analyzed on an HP Agilent 5890 Series II gas chromatograph (GC) equipped with an HP-INNOWAX capillary column (30 m long, 0.25 mm i.d., 0.25 μm film thickness), and an HP Agilent 5971 mass spectrometer detector (MSD) (HP, Palo Alto, CA). The flow rate of helium carrier gas was 1 mL/min. Both the inlet and the detector interface temperatures were held at 320 °C, and the former was operated in splitless mode. Oven temperature program was as follows: hold at 35 °C for 3 min, ramp at 5 °C/min to 250 °C, and hold at 250 °C for 5 min. Target analytes were the biodiesel FAMEs and the petrodiesel C10−C24 n-alkanes. Six additional peaks eluting within the method range and positively identified as toluene, ethylbenzene, p-xylene, m-xylene, o-xylene, and tetralin were also targeted for analysis. Detection limits were in the range of 5−25 μg/L for all the analytes. The surrogate recoveries of all extractions ranged from 80 to 120%, and were used to ensure that extraction recoveries were within the permissible limits of EPA method 8000 and were not used as internal standards for analytical quantifications. 2.5. CO2 and VSS Analysis. Two sampling events of six bottles each (biologically active + endogenous respiration controls) per fuel treatment were set for the analysis of total CO2 and VSS produced at 7 and 42 d (ultimate). Just prior to sampling time, 100 μL of HCl was spiked in the bottles through the PTFE septa, delivering a final concentration of 24 mN HCl and dropping the pH < 2. While the bottles were still in the inverted position, the needle of the syringe was slowly pulled out of the bottles to minimize leakage of gas in or out of the bottles. The septa were then wrapped with parafilm, and the bottles were tumbled again for a 24 h equilibration period. Upon equilibration, a 20 mL gastight Popper Micromate Syringe (New Hyde Park, NY) was inserted inside the headspace of each bottle, and the gas in the headspace was allowed to expand to equilibrate the pressure inside the bottle with atmospheric pressure. The volume of headspace gas expanded inside the 20 mL syringe was recorded, and while the syringe was still inside the bottle, a 0.5 mL of the headspace gas was withdrawn by a gastight Vici Precision Sampling Syringe (Baton Rouge, LA) and analyzed for gas composition using an HP 5890 Series II gas chromatograph (GC) equipped with a dual column HP 10 ft molecular sieve BX-45/60 mesh, HP 6 ft HAYESEP Q 80/100 mesh, and a thermal conductivity detector (TCD) (HP, Palo Alto, CA). The mass of CO2 in gas phase was calculated using the ideal gas law as shown in eq 1. mCO2,gas =
PCO2(Vheadspace + Vsyringe) RT
MWCO2
where mCO2,gas = mass of CO2 in the gas phase (mg), PCO2 = partial pressure of CO2 inside the gas phase of bottle (atm), and is equal to the percentage of CO2 in the headspace measured by the GC-TCD, Vheadspace and Vsyringe are the volumes of the bottle’s headspace and expanded gas in the syringe (L), respectively, MWCO2 = molecular weight of CO2 (44 000 mg/mol), R = universal gas constant 0.08205 L·atm/ K·mol, andT = room temperature (295.2 K). The mass of aqueous-phase CO2 was calculated using Henry’s law according to eq 2, assuming thermodynamic equilibrium between the gas and aqueous phase CO2 species, and negligible ionized carbonate species due to the low pH, mCO2,aq = PCO2KHVaq MWCO2
(2)
where mCO2,aq = mass of CO2 in the aqueous phase (mg), KH = Henry’s law constant for CO2 at 295.2 K (KH = 2.40 × 10−2 Molar/atm),22 and Vaq = volume of the aqueous phase (0.05 L). The total amount of CO2 produced was calculated as the sum of eqs 1 and 2 and was corrected with respect to endogenous respiration in the blanks at every sampling event. The aforementioned procedure was validated by measuring the dissolved inorganic carbon directly using a Shimadzu TOC-V CSH Analyzer (Columbia, MD) according to manufacturer’s procedures and found to be within 3% accuracy. After the CO2 analyses were completed, the bottles were opened, filled completely with sterile saline solution, resealed with the same septa, and placed in an ultrasound sonication bath for 12−16 h to detach any biomass adhered to the glass surface before measuring the VSS in the bottles according to EPA Method 160.4. Chemically oxidizable carbon in B0 and B100 fuels was experimentally determined by injecting 10 μL of the fuels in sealed serum bottles containing 20 mL of a 3% potassium dichromate in sulfuric acid (95%) solution and incubated at 150 °C for 24 h. The bottles were then cooled down to room temperature, and the headspace gas was analyzed as outlined above. The sulfuric acid phase was diluted with deionized water down to 5% and analyzed for inorganic carbon on the Shimadzu TOC analyzer. The amount of chemically oxidizable carbon in the fuels was found to be 2.346 ± 0.036 and 2.484 ± 0.0315 mg CO2/μL fuel for B0 and B100, respectively. 2.6. Fuels/Water Equilibration Experiment. To test any cosolubilizing effect between biodiesel and petrodiesel, abiotic batch experiments were conducted in which biodiesel/ petrodiesel blends were equilibrated with a sterile micronutrients and vitamins containing aqueous phase by vigorous shaking and tumbling for 1 h in an experimental setup similar to that of the biological batch experiment outlined earlier but without adding biomass. After equilibration, the samples were transferred into clean separatory funnels and allowed to settle for 1 h. The supersaturated aqueous phase was withdrawn from the bottom of the funnels and was filtered through fritted glass funnels (6 cm deep, 3 cm i.d., 1 μm pore size) filled with 3 mm surface-deactivated borosilicate glass beads (Chemglass, Vineland NJ) to filter out the particulate and colloidal oil from the aqueous phase. Filtrates were then extracted with DCM and concentrated 18-fold in a TurboVap II Evaporation System (Biotage, Charlotte NC) according to the manufacturer’s procedure, and analyzed by GC-MS as outlined earlier. Surrogate recoveries for this experiment ranged from 84% to 96%.
(1) 4621
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
Figure 1. nC10−nC21 alkanes biological degradation data in B0 (red ●), B20 (green ▲), B40 (yellow ■), B60 (purple ◆), and B80 (red ▼). Error bars are ±1 standard deviation of triplicate samples. Solid lines are regression fits by eq 3.
3. RESULTS AND ANALYSIS
Stot = Stot0 −
3.1. n-Alkanes. The results for n-alkanes from the kinetic batch experiments were fitted to a mechanistic kinetic model for the biodegradation of poorly soluble organic materials developed elsewhere by Yassine et al.23 In this model, each nalkane of a given carbon chain length is treated individually, and the term “substrate” is used to refer to each individual compound in the oil, not the oil in its entirety. Similarly, the term “biomass” refers to that arbitrary portion of the total biomass that is actively degrading a given n-alkane with a specific carbon number. The model is given in eq 3 below,
kCX 0 (e(YkC − kd)t − 1) (YkC − Kd)
(3)
where Stot = the experimentally observed total mass of residual substrate (i.e., n- alkane) divided by the volume of aqueous phase in the bottle (mg/L) at time t (h), X0 = the initial concentration of biomass degrading a specific n-alkane (mg/L), Y = biomass yield coefficient (g-biomass produced/g-alkane consumed), kd = biomass endogenous decay rate coefficient (h−1), k = maximum specific substrate utilization rate (mgalkane/mg-biomass·h), and C = a constant and is equal to Ssat/ (Ks+ Ssat),23 where Ssat and Ks are the substrate saturation 4622
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
between the nonaqueous phase liquid (NAPL) and the aqueous phase, prior to the transformation reaction, which rate was probably enhanced inside the aqueous phase. The initial rate of abiotic transformation of the FAMEs (Figure S4) appeared to be also influenced by the amount of petrodiesel in the blend. The FAMEs in B60, followed by B40, seemed to have the highest initial slope of abiotic transformation. A wide range of petrodiesel fuel additives have been reported to include metals such as Ce, Mo, and Mn among others, which have been shown to significantly enhance the oxidation kinetics and efficiency of the petroleum fuel in internal combustion engines27,28 and are also likely to affect the rate of abiotic transformation of the FAMEs in biodiesel/petrodiesel blends. Metal analysis using ICP-MS, however, did not confirm the presence of such a catalyst in B0. There was no significant difference in the metal concentrations between B0 and B100. 3.3. Aromatic Compounds. Although the targeted aromatic compounds were present in B0 at very low concentrations (0.35−0.70 ppm for toluene, ethylbenzene, and xylene isomers), analyzing these compounds in the batch experiments allowed for a better understanding of how the amount of biodiesel in the oil blend could affect the biodegradation capability of microbial cultures for this class of compounds. A substantial enhancement in the biodegradation of aromatic compounds occurred when biodiesel was present in the fuel (Figure S5 in SI Section). While toluene, oxylene, and tetralin were not degraded in B0 and B20 treatments, all of the aromatic compounds were completely degraded to below detection limits in B40 and B80. In B60 treatment, however, the cultures were unable to assimilate any of the xylene isomers, toluene, or tetralin during the 7-d sampling period (i.e., only ethylbenzene was degraded). Repeated batch experiments with the same biomass produced the same results. It should also be noted that the biodegradation of o-xylene and tetralin in B40 and B80 was accompanied by the development of an intense yellow coloration of the aqueous phase, most likely due to partial oxidation products of aromatic biodegradation. 3.4. Fuel Blend Biomineralization. The biomineralization of the fuel blends at 7 and 42 d is shown in Figure 4. The extent of biomineralization of the fuel blends, with respect to the chemically oxidizable carbon, was found to increase linearly with the fraction of biodiesel in the blend from 23.08% and 49.17% for B0 to 64.13% and 74.78% for B100, at 7 and 42 d, respectively. Analysis of the mineralization data also indicated that the rates of mineralization of the fuel blends were significantly enhanced in the presence of the biodiesel constituents (p < 0.05). The fraction of carbon mineralized during the first 7 d period, given by the ratio of carbon mineralized at 7 days to ultimate carbon mineralized at 42 d increased from 46.94% for B0, to 67.71% for B20, 63.52% for B40, 76.21% for B60, 71.75% for B80, and 85.76% for B100. In order to test for any synergetic enhancement attributable to biodiesel on the mineralization efficiency of fuel blends, the scatter plots in Figure 4 were constructed by linearly interpolating the biomineralization CO2 data of B0 and B100, independently, at each of the other fuel mixtures. Results indicated no statistically significant difference between the actual mineralization and interpolated data. The mineralization of the fuel mixtures (B20, B40, B60, and B80) was found to be equivalent to that of the linearly interpolated data from B0 and B100.
concentration in the aqueous phase and the Monod halfsaturation constant (mg/L), respectively. Using the statistical software R, nonlinear least-squares minimization was used to fit the experimental data of each individual n-alkane to eq 3 and estimates for the microbial net specific growth rate (YkC − kd), according to eq 3, were obtained for all n-alkanes in the different fuel blend treatments (i.e., B0, B20, B40, B60, and B80) and are all presented in Figure 1. Figure 1 indicates that the n-alkanes utilization rates for fuel blends B20 and B40 were faster than the other fuel blends. To investigate this observation further, the actual microbial specific utilization rate (kC) for each n-alkane was approximated from the regression parameter (YkC − kd), using the stoichiometric yield coefficient Ystoich assuming a bacterial cell formula of C60H87O23N12P24 and measured endogenous decay rate kd of 0.020−0.038 d−1 for all the cultures. The microbial specific utilization rates (kC) of n-alkanes in the fuel blends are compiled in Figure 2. For almost all the n-alkanes,
Figure 2. Actual microbial utilization rates (kC’s) of the n-alkanes in B0, B20, B40, B60, and B80 as color coded in figure key. Error bars are 1 standard error of the estimates.
the microbial specific utilization rates (kC) in B20 were significantly higher than those in B0 (p < 0.005) and B80 (p < 0.05). Similarly, the specific utilization rates of the n-alkanes were significantly higher in B40 compared to B0 (p < 0.005) and B80 (p < 0.05). Although the differences in the n-alkanes specific utilization rates of B20 and B40 were statistically insignificant, as well as the difference in the n-alkanes specific rates of B60 and B80, Figure 2 points out a decreasing trend in the n-alkanes kC values in the order B20 > B40 > B60 > B80 > B0, which becomes more noticeable for the n-alkanes with longer carbon chains. 3.2. FAMEs. All of the FAMEs in all the fuel blends were found to undergo rapid abiotic transformation in the control treatments (Figure 3, dotted lines). FAMEs, and lipids in general, are widely reported to naturally undergo a range of abiotic transformation processes such as hydrolysis, autoxidation, and polymerization in the presence of a wide range of catalysts.25,26 The abiotic transformation of the FAMEs was so fast that accurate estimation of biodegradation rates for these compounds was not possible in most of the biological treatments (see Figures S2 and S3 in the SI Section). Further examination of the FAMEs abiotic control data (Figure S4 in the SI Section) revealed substantial deviation from the widely adopted first-order kinetics for such reactions, especially for the more hydrophobic FAMEs (i.e., C16:0-ME and C18:0-ME). Such a behavior can be accounted for if dynamic equilibrium partitioning of these compounds is assumed to take place 4623
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
Figure 3. FAMEs biological degradation data in all of the petrodiesel/biodiesel blends (solid symbols), and abiotic controls (hollow symbols). C16:0-ME C16:0-ME (● and ○), C18:0-ME (red ▼ and ▽), C18:1-ME (green ■ and □), C18:2-ME (yellow ◇ and ◆), C18:3-ME (blue ▲ and △). Error bars are ±1 standard deviation of triplicate samples, and dotted lines connect abiotic controls.
cometabolism of these compounds since the total amount of CO2 produced in the blends did not increase beyond the amount of CO2 produced either in B0 or B100 separately. In addition, the phenomenon of cometabolism, in principle, does not apply for the case of FAMEs and alkanes. Co-metabolism occurs when a nongrowth substrate, a substrate that microorganisms cannot utilize as a sole carbon and/or energy source, is co-oxidized in the presence of a growth substrate.29 Clearly, both the FAMEs and the n-alkanes are not only growth substrates, but are also metabolized by a metabolically similar mechanism, namely β-oxidation.30,31 But since biodiesel is a better growth substrate compared to petrodiesel,32 biodiesel could have enhanced the rate of microbial degradation of the nalkanes by increasing the actual growth yield coefficient (Yactual) of the biomass. Such an enhancement, however, would increase the total degradation rate and would not affect the specific utilization rate of the substrate as modeled by Monod kinetics.33 Fundamentally, an increase in the actual specific utilization rate (kC) can be attributed to an increase in either the maximum specific utilization rate (k), or the substrate aqueous saturation concentration (Ssat) acting through the constant C (C = Ssat/[Ks + Ssat]).23 While the rate of dissolution of nalkanes from the NAPL into the aqueous phase is believed to be faster than the rate of biological assimilation of the substrate in the aqueous phase,34−41 and since the value of k for a biodegradable, poorly soluble material is usually orders of magnitude higher than the value of C for the same material (i.e., high k and low Ssat), the actual specific utilization rate of nalkanes is more likely to be controlled by their poor aqueous solubility rather than their mass transfer rate.
Figure 4. 7 d (black) and ultimate (gray) mineralization efficiency of the petrodiesel/biodiesel blends. Bar plot is the observed experimental results (error bars are ±1 standard deviation of triplicate samples); scatter plot is linear interpolation from B0 and B100 experimental data (error bars are error-propagated), and dashed lines connect the scatter plot.
4. DISCUSSION Both the specific microbial utilization rates of the n-alkanes (kC’s) and the rates of mineralization of the fuel blends were significantly enhanced in the presence of biodiesel, but a similar enhancement in the extent of biomineralization of these fuel blends was not observed. Specifically, 85.76% of the biomineralized carbon in B100 occurred during the first 7 d of the experiment, compared to 46.94% for B0 during the same period. These results suggest that the enhancement of biodegradation kinetics of the n-alkanes was not due to 4624
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
solubility) inside the aqueous phase in the presence of the FAMEs. This cosolubilizing behavior as well as other controlling mechanisms of the partitioning of biodiesel/ petrodiesel blends in water were thoroughly investigated by Yassine et al. elsewhere.42 Although the FAMEs were found to disappear abiotically in all of the petrodiesel/biodiesel blends, this abiotic transformation did not appear to appreciably affect the amount of CO2 produced at 7 and 42 d (shown in Figure 4), which indicates that abiotic transformation byproducts were still amenable to biodegradation. Such an observation is supported by the findings of Yassine et al.43 in which the chemical nature and ecotoxicological profiles of a biodiesel autoxidation products were investigated in a simulated oil spill scenario. The deviation of the kinetics of FAMEs abiotic transformation from the first order behavior, especially for C16:0 and C18:0 methyl esters (Figure S4), suggests dynamic partitioning of these compounds between the NAPL and the aqueous phase prior to the abiotic reaction taking place in the aqueous phase. The lower the amount of FAMEs in the NAPL, the lower is the amount of FAMEs being partitioned on the oil/water interface and subsequently dissolving into the aqueous phase. Thus, the flattening of the FAMEs abiotic transformation curves is observed with the depletion of these FAMEs (Figure S4). Although metal analysis did not confirm the presence of an assumed catalyst in B0, the findings of the cosolubilization experiment (Figure 5) could partially explain the observed differences in the initial rates of abiotic transformation of the FAME in the different fuel blends. Figure 5a shows that the highest cosolubilization of n-alkanes was observed in B60 followed by B40, and Figure S4 shows that the highest initial slope of abiotic transformation of the FAMEs was also observed in B60 followed by B40. Yet, the aqueous equilibrium concentration of the FAMEs in the petrodiesel/biodiesel blends (Figure 5b) did not show preferential cosolubilization of the FAMEs in B60 and B40. However, polymerized insoluble deposits of the autoxidation the FAMEs were visible in the cosolubilization experiment. This indicates that the measured equilibrium concentrations of the FAMEs presented in Figure 5b were not necessarily the true solubility limit of the FAMEs but were most likely reduced by autoxidation and polymerization. Monitoring the aromatic compounds during the batch experiments provided a different aspect on the effect of biodiesel/petrodiesel blending on the cultures’ acclimation and biodegradation capabilities. It is unclear why most of the aromatics in the B60 treatment resisted biodegradation while they were utilized in the other blends. Complete degradation of all aromatic compounds to below detection limits occurred in the B40 and B80 treatments, and toluene, o-xylene, and tetralin in B0 and B20 resisted biodegradation. These observations suggest a direct effect for the oil blend on the development and acclimation of the degrading culture. Since only the bioavailable fraction of the oil blends is what the microorganisms are exposed to in the dissolved phase, it is reasonable to conclude that these observations regarding the biodegradation of the aromatic compounds are affected by the cosolubilization behavior of biodiesel/petrodiesel blends depicted in Figure 5. In other words, as both the absolute and relative concentrations of the dissolved constituents of petrodiesel/biodiesel blends significantly and nonlinearly varied as shown in Figure 5, these variances may have significantly shifted the composition of the microbial communities degrading the oils to preferentially
Therefore, we tested the hypothesis that biodiesel is able to cosolubilize the n-alkanes in biodiesel/petrodiesel blends by conducting an abiotic fuel-blends/aqueous-phase equilibration experiment. Miller and Mudge15 suggested the possibility of cosolubilization of crude oil by biodiesel in sand columns. However, they did not provide any direct measurement of such cosolubilization. The results from the oil-blends/aqueous-phase equilibration experiment are summarized in Figure 5. Although
Figure 5. Results of the cosolubilization experiment. Equilibrium aqueous concentrations of the dissolved phase n-alkanes (a), FAMEs (b), and aromatic compounds (c) for B0, B20, B40, B60, B80, and B100, as color coded in the figures keys. Error bars are ±1 standard deviation of triplicate samples.
the total amount of petrodiesel in the fuel blends decreases as the total amount of biodiesel increases, Figure 5 shows that the equilibrium aqueous concentrations of the n-alkanes were significantly higher (p < 0.005) when biodiesel was added to the fuel blends. Figure 5 provides clear evidence that the FAMEs of biodiesel are capable of cosolubilizing the n-alkanes of petrodiesel, possibly by reducing the interfacial surface tension between the aqueous phase and the n-alkanes and, thus, enhancing the actual specific utilization rate (kC) for these compounds due to their increased bioavailability (i.e., 4625
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
(6) Demirbaş, A. Biodegradability of biodiesel and petrodiesel fuels. Energy Sources, Part A 2009, 31 (2), 169−174. (7) Sendzikiene, E.; Makarewiciene, V.; Janulis, P.; Makareviciute, D. Biodegradability of biodiesel fuel of animal and vegetable origin. Eur. J. Lipid Sci. Technol. 2007, 109 (5), 493−497. (8) Zhang, X.; Peterson, C.; Reece, D.; Haws, R.; Möller, G. Biodegradability of biodiesel in the aquatic environment. Trans. ASABE 1998, 41 (5), 1423−1430. (9) Pasqualino, J. C.; Montané, D.; Salvadó, J. Synergic effects of biodiesel in the biodegradability of fossil-derived fuels. Biomass Bioenergy 2006, 30 (10), 874−879. (10) Chen, J.; Wang, X.; Hu, E.; Xu, Y.; Hu, X.; Pan, L.; Jiang, S. Evaluation of biodegradability of biodiesel prepared from cotton seed oil. Nongye Gongcheng Xuebao/Trans. Chin. Soc. Agric. Eng. 2010, 26 (1), 301−304. (11) Mariano, A. P.; Tomasella, R. C.; De Oliveira, L. M.; Contiero, J.; De Angelis, D. D. F. Biodegradability of diesel and biodiesel blends. Afr. J. Biotechnol. 2008, 7 (9), 1323−1328. (12) Owsianiak, M.; Chrzanowski, Å.; Szulc, A.; Staniewski, J.; Olszanowski, A.; Olejnik-Schmidt, A. K.; Heipieper, H. J. Biodegradation of diesel/biodiesel blends by a consortium of hydrocarbon degraders: Effect of the type of blend and the addition of biosurfactants. Bioresour. Technol. 2009, 100 (3), 1497−1500. (13) Prince, R. C.; Haitmanek, C.; Lee, C. C. The primary aerobic biodegradation of biodiesel B20. Chemosphere 2008, 71 (8), 1446− 1451. (14) DeMello, J. A.; Carmichael, C. A.; Peacock, E. E.; Nelson, R. K.; Samuel Arey, J.; Reddy, C. M. Biodegradation and environmental behavior of biodiesel mixtures in the sea: An initial study. Mar. Pollut. Bull. 2007, 54 (7), 894−904. (15) Miller, N. J.; Mudge, S. M. The effect of biodiesel on the rate of removal and weathering characteristics of crude oil within artificial sand columns. Spill Sci. Technol. Bull. 1997, 4, 17−33. (16) Taylor, L. T.; Jones, D. M. Bioremediation of coal tar PAH in soils using biodiesel. Chemosphere 2001, 44 (5), 1131−1136. (17) Pereira, M. G.; Mudge, S. M. Cleaning oiled shores: Laboratory experiments testing the potential use of vegetable oil biodiesels. Chemosphere 2004, 54 (3), 297−304. (18) Fernandez-Alvarez, P.; Vila, J.; Garrido, J. M.; Grifoll, M.; Feijoo, G.; Lema, J. M. Evaluation of biodiesel as bioremediation agent for the treatment of the shore affected by the heavy oil spill of the Prestige. J. Hazard. Mater. 2007, 147 (3), 914−922. (19) Zein, M. M.; Suidan, M. T.; Venosa, A. D. Bioremediation of groundwater contaminated with gasoline hydrocarbons and oxygenates using a membrane-based reactor. Environ. Sci. Technol. 2006, 40 (6), 1997−2003. (20) Campo, P.; Zhao, Y.; Suidan, M. T.; Venosa, A. D.; Sorial, G. A. Biodegradation kinetics and toxicity of vegetable oil triacylglycerols under aerobic conditions. Chemosphere 2007, 68 (11), 2054−2062. (21) Grady, C. P. L.; Smets, B. F.; Barbeau, D. S. Variability in kinetic parameter estimates: a review of possible causes and a proposed terminology. Water Res. 1996, 30 (3), 742−748. (22) Sander, R. Compilation of Henry’s law Constants for Inorganic and Organic species of potential importance in environment chemistry; MaxPlanck Institute of Chemistry: Mainz, Germany, 1999. (23) Yassine, M. H.; Suidan, M. T.; Venosa, A. D. Microbial kinetic model for the degradation of poorly soluble organic materials. Water Res. 2013, 47 (4), 1585−1595. (24) Hoover, S. R.; Porges, N. Assimilation of Dairy wastes by activated sludge: II. The equation of synthesis and rate of oxygen utilization. Sewage Ind. Wastes 1952, 24 (3), 306−312. (25) Brodnitz, M. H. Autoxidation of saturated fatty acids. A review. J. Agric. Food Chem. 1968, 16 (6), 994−999. (26) Fang, H. L.; McCormick, R. L. Spectroscopic study of biodiesel degradation pathways. SAE Tech. Pap. Ser. 2006, 2006 (01), 3300. (27) Ribeiro, N. b. M.; Pinto, A. C.; Quintella, C. M.; da Rocha, G. O.; Teixeira, L. S. G.; Guarieiro, L. L. N.; do Carmo Rangel, M.; Veloso, M. r. C. C.; Rezende, M. J. C.; Serpa da Cruz, R.; de Oliveira, A. M.; Torres, E. A.; de Andrade, J. B. The role of additives for diesel
metabolize the oil fractions that are more bioavailable at each specific cosolublized oil blend. Finally, the bright yellow metabolite observed in the B40 and B80 treatments and not B0 or B20 indicates that this metabolite could be a product of the oxidation of either o-xylene or tetralin. In fact, a similar colored metabolite was observed in the biodegradation of crude oil in the presence of biodiesel, but not in the biodegradation of crude oil alone, according to Mudge and Pereira.44 These authors, however, failed to associate this metabolite with a specific group of parent compounds. Wrenn and Venosa45 associated similar colored metabolites with the accumulation of partial oxidation products of polycyclic aromatic hydrocarbons (PAHs). Similarly, other researchers46−48 reported that during microbial degradation of PAHs, yellow-brown colored products are synthesized that have quinone-like structures from the condensation of phenolic compounds in mineral media. Further research investigating the effect of biodiesel on the microbial metabolism of aromatic compounds is desirable, as these compounds pose greater risks in the environment.
■
ASSOCIATED CONTENT
S Supporting Information *
Figure S1. Schematic diagram of the porous pot (PP) chemostat bioreactors used in this study; Table S1. List of the target ions and retention times of the target analytes, internal standards, and surrogates used in this study; Figure S2. FAME biological degradation data with model regression; Figure S3. Microbial utilization rates of FAMEs; Figure S4. Deviation of abiotic transformation kinetics of the FAMEs from first-order behavior; and Figure S5. Biodegradation of targeted aromatic compounds. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This research was supported, in part, by the United States Environmental Protection Agency (US EPA) Oil Spill Program, National Risk Management Research Laboratory, Land Remediation and Pollution Control Division, under Contract No. EP-C-11-006, Work Assignment 1-19.
■
REFERENCES
(1) Agency for Toxic Substances and Disease Registry (ASTDR). Toxicological profile for fuel oils; U.S. Department of Health and Human Services: Atlanta, GA, 1995. (2) Fulton, L.; Howes, T.; Hardy, J. International Energy, A. Biofuels for transport: an international perspective; OECD, International Energy Agency: Paris, 2004. (3) Hill, J.; Nelson, E.; Tilman, D.; Polasky, S.; Tiffany, D. Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proc. Natl. Acad. Sci. U.S.A. 2006, 103 (30), 11206−11210. (4) Crutzen, P. J.; Mosier, A. R.; Smith, K. A.; Winiwarter, W. N2O release from agro-biofuel production negates global warming reduction by replacing fossil fuels. Atmos. Chem. Phys. Discuss. 2007, 7 (4), 11191−11205. (5) United States Environmental Protection Agency (USEPA). A comparative analysis of biodiesels impacts on exhaust emissions, Report EPA420-P-02-001; EPA: Washington, DC, 2002. 4626
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
Environmental Science & Technology
Article
and diesel blended (ethanol or biodiesel) fuels: a review. Energy Fuels 2007, 21 (4), 2433−2445. (28) Jung, H.; Kittelson, D. B.; Zachariah, M. R. The influence of a cerium additive on ultrafine diesel particle emissions and kinetics of oxidation. Combust. Flame 2005, 142 (3), 276−288. (29) Dalton, H.; Stirling, D. I.; Quayle, J. R. Co-Metabolism [and Discussion]. Philos. Trans. R. Soc. London, Ser. B 1982, 297 (1088), 481−496. (30) Wentzel, A.; Ellingsen, T.; Kotlar, H.-K.; Zotchev, S.; ThroneHolst, M. Bacterial metabolism of long-chain n-alkanes. Appl. Microbiol. Biotechnol. 2007, 76 (6), 1209−1221. (31) Fulco, A. J. Fatty acid metabolism in bacteria. Prog. Lipid Res. 1983, 22 (2), 133−160. (32) Sørensen, G.; Pedersen, D. V.; Nørgaard, A. K.; Sørensen, K. B.; Nygaard, S. D. Microbial growth studies in biodiesel blends. Bioresour. Technol. 2011, 102 (8), 5259−5264. (33) Monod, J. The growth of bacterial cultures. Annu. Rev. Microbiol. 1949, 3 (1), 371−394. (34) Mihelcic, J.; Lueking, D.; Mitzell, R.; Stapleton, J. M. Bioavailability of sorbed- and separate-phase chemicals. Biodegradation 1993, 4 (3), 141−153. (35) Rivett, M. O.; Feenstra, S. Dissolution of an emplaced source of DNAPL in a natural aquifer setting. Environ. Sci. Technol. 2004, 39 (2), 447−455. (36) Miller, C. T.; Poirier-McNeil, M. M.; Mayer, A. S. Dissolution of trapped nonaqueous phase liquids: mass transfer characteristics. Water Resour. Res. 1990, 26 (11), 2783−2796. (37) Miller, M. E.; Alexander, M. Kinetics of bacterial degradation of benzylamine in a montmorillonite suspension. Environ. Sci. Technol. 1991, 25 (2), 240−245. (38) Ramaswami, A.; Luthy, R. G. Mass transfer and bioavailability of PAH compounds in coal tar NAPL slurry systems. 1. Model development. Environ. Sci. Technol. 1997, 31 (8), 2260−2267. (39) Ramaswami, A.; Ghoshal, S.; Luthy, R. G. Mass transfer and bioavailability of PAH compounds in coal tar NAPL slurry systems. 2. Experimental evaluations. Environ. Sci. Technol. 1997, 31 (8), 2268− 2276. (40) Chu, M.; Kitanidis, P. K.; McCarty, P. L. Dependence of lumped mass transfer coefficient on scale and reactions kinetics for biologically enhanced NAPL dissolution. Adv. Water Resour. 2007, 30 (6−7), 1618−1629. (41) Seagren, E. A.; Rittmann, B. E.; Valocchi, A. J. Quantitative evaluation of the enhancement of NAPL-pool dissolution by flushing and biodegradation. Environ. Sci. Technol. 1994, 28 (5), 833−839. (42) Yassine, M. H.; Wu, S.; Suidan, M. T.; Venosa, A. D. Partitioning behavior of petrodiesel/biodiesel blends in water. Environ. Sci. Technol. 2012, 46 (14), 7487−7494. (43) Yassine, M. H.; Wu, S.; Suidan, M. T.; Venosa, A. D. Microtox aquatic toxicity of petrodiesel and biodiesel blends: The role of biodiesel’s autoxidation products. Environ. Toxicol. Chem. 2012, 31 (12), 2757−2762. (44) Mudge, S. M.; Pereira, G. Stimulating the biodegradation of crude oil with biodiesel preliminary results. Spill Sci. Technol. Bull. 1999, 5 (5−6), 353−355. (45) Wrenn, B. A.; Venosa, A. D. Selective enumeration of aromatic and aliphatic hydrocarbon degrading bacteria by a most-probablenumber procedure. Can. J. Microbiol. 1996, 42 (3), 252−258. (46) Stieber, M.; Haeseler, F.; Werner, P.; Frimmel, F. H. A rapid screening method for micro-organisms degrading polycyclic aromatic hydrocarbons in microplates. Appl. Microbiol. Biotechnol. 1994, 40 (5), 753−755. (47) Hedges, J. Polymerization of humic substances in natural environments. In Humic substances and their role in the environment; Frimmel, F. H., Christman, R. F., Eds.; Wiley Interscience: Chichester, 1988; pp 5−58. (48) Stevenson, F. J. Humus chemistry: genesis, composition, reactions; Wiley-Interscience: New York, 1982.
4627
dx.doi.org/10.1021/es400360v | Environ. Sci. Technol. 2013, 47, 4619−4627
