Environ. Sci. Technol. 2001, 35, 102-107
Analysis of Long-Side-Chain Alkylaromatics in Crude Oil for Evaluation of Their Fate in the Environment TAPAN K. DUTTA AND SHIGEAKI HARAYAMA* Marine Biotechnology Institute, 3-75-1 Heita, Kamaishi, Iwate 026-0001, Japan
It has long been believed that the n-hexane fraction of crude oil only contains saturates. However, we found that onethird of its content was aromatics with long alkyl side chains and that these aromatics could be separated from the saturates by preparative thin-layer chromatography. The separated alkylaromatic fraction was characterized by UV-visible, NMR and mass spectrometries. A gas chromatographic-mass spectrometric analysis showed the presence of a homologous series of long-side-chain n-alkylaromatics, namely mono-, di-, and tri-n-alkylbenzenes in the C7-C27 range (the subscript to C indicates the total number of carbon atoms in the alkyl side chain) and di- and tri-n-alkylbenzothiophenes in the C3-C22 range. The biodegradation of these crude oil components by a natural bacterial population in seawater and their photooxidation by artificial sunlight were investigated. The n-alkylbenzenes were found to be quite susceptible to biodegradation but resistant to photooxidation, whereas the n-alkylbenzothiophenes were almost completely photooxidized and substantially biodegraded.
Introduction A vast amount of mineral oil is being released into the ocean. Upon its release, the petroleum components undergo various modifications, with biodegradation and photooxidation being considered to be the two major factors contributing to the removal of hydrocarbon molecules from the environment. A homologous series of long-side-chain alkylaromatics, including sulfur-containing compounds, comprise a significant portion of petroleum and have served as markers to identify the source of oil pollution (1-11). Although the decomposition of C10-C14 phenylalkanes (the subscript to C indicates the number of carbon atoms in the alkyl side chain), which are the raw materials for linear alkylbenzenesulfonate surfactants (LAS), in the natural environment has been intensively studied (12-16), little is known about the fate of alkylaromatics after their release into the natural environment (17, 18). The biodegradation of n-alkylbenzenes with C12-C14 alkyl side chains has been demonstrated in Acinetobacter lwoffi (19). The biodegradation of alkylbenzenes by bacterial communities and by pure cultures has been further investigated, and it was found that the combination of photooxidation and biodegradation eliminated n-nonylbenzene more quickly than biodegradation alone (20, 21). * Corresponding author phone: +81-193-26-6544; fax: +81-19326-6592; e-mail:
[email protected]. 102
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 1, 2001
The biodegradation of benzothiophenes and dibenzothiophenes possessing no or only a short-chain alkyl group has been investigated, and pathways involved in the degradation of these compounds were elucidated (22-25). The photochemical decomposition of alkylbenzothiophenes and alkyldibenzothiophenes in diesel oil has been studied for the development of desulfurization techniques (26, 27), and methylbenzothiophenes were shown to be oxidized to 2-sulfobenzoic acid via oxidation of the methyl group and cleavage of the thiophene ring (28). In this present study, a method is developed to separate and characterize long-side-chain alkylaromatics of crude oil. This method is then applied to investigate the biodegradation and photooxidation of alkylaromatics under conditions which mimic the natural marine environment.
Experimental Section Crude Oil. We were interested in studying the degradation of the heavy fractions of crude oil as they are generally recalcitrant to biodegradation. To prepare crude oil rich in the heavier fractions, Arabian light crude oil (38 L) was distilled at 230 °C, and the unevaporated part (22 L) was used as the starting material for this study. This unevaporated part is subsequently called “crude oil” in this report. Fractionation of the Crude Oil. Crude oil, its biodegraded samples, or its photooxidized samples were dissolved in chloroform and subsequently dried in vacuo. The dried material was then dissolved in n-hexane to separate the n-hexane-soluble fraction (maltenes) from the fraction precipitated in n-hexane (asphaltenes). The maltene fraction was further fractionated by column chromatography with silica gel (C-200, Wako Pure Chemicals, Tokyo, Japan). The silica gel was pretreated by washing serially with 3 volumes of acetone, n-hexane and dichloromethane in a column with a coarse-porosity fitted disk, left under a fume hood overnight and then dried completely at 50 °C. The dried gel was activated at 180 °C for 20 h (29). A column containing 3 g of activated silica gel topped with 1.5 g of anhydrous sodium sulfate was conditioned with 20 mL of n-hexane, and 300 µL of n-hexane containing ∼30 mg of crude oil material was applied to the column. The column was then eluted with a 3-bed volume (21 mL) of n-hexane to collect the saturates (this fraction was subsequently found to contain long-side-chain alkylaromatics as described in the Results and Discussion section). The column was finally eluted with a 3-bed volume of n-hexane:benzene (1:1) to collect the aromatics. Each fraction was dried in vacuo. The n-hexane fraction was separated by preparative thinlayer chromatography (silica gel, 20 × 20 cm, 0.25 mm thickness; Merck, Germany), developing with n-hexane for 40 min and then by n-hexane:toluene (20:80, v/v) in the same dimension for 13 min. Two fractions, called fraction 1 (Rf > 0.57) and fraction 2a (Rf ) 0.34-0.45), were recovered by scraping the individual bands from the plate and extracting the silica gel with chloroform. Gravimetric Analysis. The weight of the crude oil components in each sample was determined with a sensitive digital balance. TLC-FID Analysis. An Iatroscan MK-5 system (Iatron Laboratories, Tokyo, Japan) was used to carry out thin-layer chromatography (TLC) and flame ionization detection (FID) to quantify the separated fractions of the crude oil (30). One µL of a chloroform extract was applied to an SIII Chromarod coated with silica, and the sample was developed first with 100% n-hexane for 20 min, then by toluene:n-hexane (80:20, v/v) for 9.45 min, and finally by dichloromethane:methanol 10.1021/es001165a CCC: $20.00
2001 American Chemical Society Published on Web 11/16/2000
(95:5, v/v) for 2.25 min. The developed sample was then analyzed by FID. GC-MS Analysis. The analysis of the crude oil fractions by gas chromatography-mass spectrometry (GC-MS) was performed by a QP-5000 instrument (Shimadzu, Kyoto, Japan) fitted with a fused silica capillary column (DB-5, 30 m × 0.25 mm; J&W Scientific, Tokyo, Japan). The temperature was held at 50 °C for 2 min, increased from 50 °C to 300 °C at a rate of 6 °C/min and then held at 300 °C for 16 min. The injection volume was 1 µL, and the carrier gas was helium at a flow rate of 1.7 mL/min. The mass selective detector was operated either in the scan mode or in the selected ion monitoring (SIM) mode. 1H- and 13C NMR Analyses. A sample was dissolved in CDCl3, and the spectra were recorded by a JNM FT-NMR system (JEOL, Tokyo, Japan) operated at 400 MHz for 1H and at 100 MHz for 13C. A pulse delay of 1.0 s was used for quantitative 13C NMR. The intensity of each chemical shift was determined relative to tetramethylsilane (TMS) as an internal standard. The aromaticity is defined as the ratio of aromatic carbon to total carbon, where the amount of aromatic and aliphatic carbons were obtained from the integrated intensity of the peaks between 100 and 170 ppm and 8-58 ppm, respectively (31). Photooxidation Treatment. The treatment by photooxidation was carried out on an oil slick. A 50-mg sample of crude oil was floated on 50 mL of distilled water in a glass Petri dish (90 mm in internal diameter) to form the slick. Samples were incubated in a chamber whose temperature was controlled at 20 °C by ventilating with air and irradiated for 4 weeks under artificial sunlight by a combination of four tin-halide lamps, four high-pressure mercury lamps and four argon lamps (Toshiba Corporation, Tokyo, Japan) whose emission spectrum and intensity (1200 µmol m-2 s-1) are very similar to those of natural sunlight. Dark control samples were prepared in a similar way and incubated in the dark. Biodegradation Treatment. Seawater was collected from a depth of 15 m in Kamaishi Bay, Japan. The seawater medium used for the biodegradation experiments consists of 800 mL of nonsterilized seawater and 200 mL of an autoclaved solution (pH 7.6) containing 5 g/L of NH4NO3, 1 g/L of K2HPO4, and 0.1 g/L of ferric citrate. Crude oil was added at a concentration of 1 g/L as the carbon source. Abiotic controls were prepared from autoclaved seawater. The medium was incubated at 20 °C for 8 weeks while constantly shaking (100 strokes/min). Each experimental sample and negative control sample was prepared in duplicate. Preparation of the Oil Samples. At appropriate times, the pH value of both the biodegraded and photooxidized samples was adjusted to 7.0, and each sample was extracted three times with half of its volume of chloroform. The chloroform extracts were evaporated in a rotary evaporator under reduced pressure. Each extract was transferred to a preweighed glass vial and left under a fume hood to evaporate further to constant weight.
Results and Discussion Chromatographic Analysis of the n-Hexane Fraction. The n-hexane fraction of crude oil has long been described as the fraction containing only saturated molecules (32). However, as described next, we found that n-hexane can elute not only the saturates but also long-side-chain alkylaromatics. The TLC-FID analysis of the crude oil and its n-hexane fraction demonstrates that the components in the n-hexane fraction were resolved into two peaks, 1 and 2a (Figure 1), the abundance of these two peaks estimated from the TLCFID signal intensities being in the ratio of 2:1. These two peaks were separated into fractions 1 and 2a by preparative TLC. When 95 mg of material was applied, 59.6 mg and 30.6 mg were recovered in fractions 1 and 2a, respectively,
FIGURE 1. TLC-FID chromatograms. A, B, C and D present the data for crude oil, the n-hexane fraction, fraction 1 and fraction 2a, respectively. Peaks 1, 2a, 2b, 3 and 4 are classified as saturates, long-side-chain alkylaromatics, aromatics, resins and asphaltenes, respectively.
FIGURE 2. UV-visible spectra of the n-hexane fraction and subfractions 1 and 2a. A, B and C present the absorption spectra of the n-hexane fraction, fraction 1 and fraction 2a, respectively. The oil concentration of each sample was 30 µg/mL. supporting the 2:1 estimate by TLC-FID. Since 124 mg of the n-hexane fraction was recovered from 230 mg of crude oil, fraction 2a accounted for 18% (w/w) of the initial crude oil sample. The UV-visible absorption spectrum of the total n-hexane fraction and the spectra of TLC-separated fractions 1 and 2a are presented in Figure 2. Fraction 1 did not exhibit any absorption at wavelengths above 220 nm, while fraction 2a showed an absorption spectrum with three shoulders at 270, 294, and 305 nm, suggesting the presence of several types of aromatic compounds in this fraction. The 3-fold increase in the absorption of fraction 2a compared with that of the n-hexane fraction supports the 2:1 ratio of the fraction 1 and fraction 2a components in the n-hexane fraction. The analyses described next indicate the presence of long-sidechain alkylaromatics in fraction 2a. NMR Analysis: An Average Molecular Structure of the Compounds in Fraction 2a. 1H- and 13C NMR analyses of fractions 1 and 2a also confirmed successful separation by TLC of the saturates and aromatics in the n-hexane fraction. No signal for the aromatic proton and the aromatic carbon was detected in fraction 1, while both the signals were detected in fraction 2a, confirming the presence of aromatics in this fraction. The aromaticity (the number of aromatic carbons/total number of carbons) of fraction 2a was calculated from the 13C NMR spectrum to be 0.21 (Table 1). If we consider an average chemical structure for the alkyl-substituted aromatics VOL. 35, NO. 1, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
103
FIGURE 3. GC-MS data for the n-alkylbenzenes in the SIM mode (m/z ) 92, 105 and 119). The spectra were recorded at m/z ) 92 (left), m/z ) 105 (center) and m/z ) 119 (right). A, B and C present the chromatograms for the n-hexane fractions prepared from crude oil, from biodegraded oil, and from photooxidized oil, respectively. The chromatograms for fraction 2a were very similar to those presented in this figure.
TABLE 1. Average Molecular Parameters of the n-Hexane Fraction and Subfractions of the Crude Oil Sample by Nuclear Magnetic Resonance 1H
fraction
Haro
Hr
NMRa
Hβ
13C
Hγ
Cali
NMRa
Caro
n-hexane 0.014 0.028 0.699 0.259 0.943 0.057 fraction 1 0 0 0.725 0.275 1.0 0 fraction 2a 0.069 0.107 0.535 0.289 0.79 0.21
fab(%) 5.7 0 21
a H aro, HR, Hβ and Hγ represent the fractions of aromatic protons (6.29.6 ppm), R-alkyl protons (2.0-4.2 ppm), alkyl protons, except for the R-alkyl and terminal methyl protons (1.0-2.0 ppm), and terminal methyl protons (0.4-1.0 ppm), respectively, while Cali and Caro represent aliphatic carbons (8-58 ppm) and aromatic carbons (100-170 ppm), respectively. b The aromaticity, fa, is defined as the ratio of aromatic carbons to total carbons.
in fraction 2a, the aromaticity could be expressed by the formula, 6/(6 + s × l), where s is the average number of substitutions per benzene nucleus, and l is the average carbon chain length per substitution. The s × l value was calculated to be 23 from the observed aromaticity value () 0.21). The ratio of aromatic protons to aliphatic protons observed in fraction 2a can also be expressed using s and l as (6 - s)/[s × (2l + 1)] ) Haro/(HR + Hβ + Hγ) ) 0.069/0.931. Since s × l was 23, s and l were calculated to be 2.4 and 9.6, respectively. The 1H NMR spectrum of fraction 2a (Table 1) indicates the ratio of R-CH2-, β-CH2-, and γ-CH3 in the alkyl side chains to be 0.107/2:0.535/2:0.289/3 ) 12.8:64.1:23.1. The average molecular weight per benzene nucleus of the compounds in fraction 2a was calculated to be approximately 400 by adding the average molecular weight of the substituted benzene nucleus (78-2.4) and that of the alkyl side chains [23 × (0.769 × 14 + 0.231 × 15)]. GC-MS Analyses of the Compounds in Fraction 2a. The mass spectra of mono-n-alkylbenzenes display doublet peaks (daughter ions) at m/z ) 91 and 92, the latter being the base peak (1). The GC-MS chromatograms of the n-hexane fraction (Figure 3A, left) and fraction 2a by SIM detection at m/z ) 92 were then recorded, and the presence of the C7-C27 series of mono-n-alkylbenzenes in fraction 2a was confirmed. The GC spectrum of an authentic standard, n-hexadecylbenzene, fitted the assigned peak in Figure 3A (left). The mass spectra of the di-n-alkylbenzenes show doublet peaks at m/z ) 105 and 106 with their base peak at m/z ) 105. The exceptions are ethyl-n-alkylbenzenes and meta104
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 1, 2001
n-alkyltoluene, whose base peaks have been observed at m/z ) 119 and 106, respectively (1, 4, 5). The m/z ) 105 chromatogram of the n-hexane fraction (Figure 3A, center) and fraction 2a indicates the presence of C7-C27 di-nalkylbenzenes in these fractions. Two major series of di-nalkylbenzenes are apparent in the chromatogram. These were identified to be a series of ortho-n-alkyltoluenes and one of n-propyl-n-alkylbenzenes (see below). The m/z ) 106 chromatogram shows a series of peaks present between each of the peaks of the ortho-n-alkyltoluenes and the n-propyl-n-alkylbenzenes, the intensities of these peaks being higher in the m/z ) 106 chromatogram than in the m/z ) 105 ion chromatogram. Thus, these peaks correspond to n-alkyltoluenes. These peaks were assigned as a homologous series of meta-n-alkyltoluenes based on their elution behavior (1) and mass-fragmentation patterns (data not shown). Tri-n-alkylbenzenes exhibit doublet peaks at m/z ) 119 and 120, the former being the base peak. The exception is 5-n-alkyl-meta-xylene whose base peak has been found at m/z ) 120 (6). The presence of six isomers of each n-alkylxylene in crude oil has been verified with the GC-MS SIM detection (4-6). The m/z ) 119 chromatogram of the n-hexane fraction (Figure 3A, right) and fraction 2a indicates the presence of either the C8-C24 series of tri-n-alkylbenzenes, or, as already mentioned, of ethyl-n-alkylbenzenes. Di- and tri-n-alkylbenzothiophenes could also be identified by GC-MS. Di-n-alkylbenzothiophenes show doublet peaks at m/z ) 161 (base peak) and 162, while the equivalent data for tri-n-alkylbenzothiophenes are m/z ) 175 (base peak) and 176 (8, 9). The exceptions are di- and tri-n-alkylbenzothiophenes possessing either one ethyl or one propyl side chain, whose base peaks were observed at +14 or +28 higher positions, respectively, than the above-mentioned m/z values (8). The major alkylbenzothiophenes in Rozel Point oil have been characterized as 4-n-alkyl-2-methylbenzothiophenes, 2-n-alkyl-4-methylbenzothiophenes and midchain 2,4-din-alkylbenzothiophenes by GC-MS SIM detection (8). Figure 4A shows the chromatograms for the n-hexane fraction at m/z ) 161 and m/z ) 175, representing the series of C3-C21 di-n-alkylbenzothiophenes and C3-C22 tri-n-alkylbenzothiophenes, respectively. Similar chromatograms were obtained using fraction 2a. Their identity as organic sulfur compounds was further demonstrated from their natural isotopic abundance (32S:34S ) 95:4.2). Figure 5 shows the chromatograms at m/z ) 161, 163, 175 and 177 to demonstrate the isotopic abundance in these organic sulfur
FIGURE 4. GC-MS data for the di-n-alkylbenzothiophenes and trin-alkylbenzothiophenes in the SIM mode (m/z ) 161 and 175). The spectra were recorded at m/z ) 161 (left) and m/z ) 175 (right). A, B and C present the chromatograms for the n-hexane fractions prepared from crude oil, from biodegraded oil, and from photooxidized oil, respectively. The chromatograms for fraction 2a were very similar to those presented in this figure.
FIGURE 5. Selected ion chromatograms for the di-n-alkylbenzothiophenes (m/z ) 161, 163) and tri-n-alkylbenzothiophenes (m/z ) 175, 177) showing the natural isotope abundance of sulfur. The chromatogram for the monoalkylbenzenes (m/z ) 92) is also shown to demonstrate that these molecules were not detected in the m/z ) 94 chromatogram. compounds. The peak heights of the m/z ) 161 and 175 chromatograms are more than 20 times those of the corresponding peaks of the m/z ) 163 and 177 chromatograms, reflecting the natural isotopic abundance of 32S and 34S. When the m/z ) 92 chromatogram was examined as a negative control, no peak corresponding to the +2 isotope species could be detected in the m/z ) 94 chromatogram. Further information on these long-side-chain alkylaromatics could be obtained by examining the molecular ion species. Figure 6 shows the m/z ) 260 chromatogram of fraction 2a, presenting the molecular ions of several C13 alkylbenzene and C9 alkylbenzothiophene isomers. One of the peaks in the m/z ) 260 chromatogram (O in Figure 6) was also observed in the m/z ) 92 chromatogram that detects mono-n-alkylbenzenes. This peak was thus found to correspond to n-tridecylbenzene. The two most abundant peaks found in both the m/z ) 260 and m/z ) 105 chromatograms (4 in Figure 6) would be di-n-alkylbenzene. The larger peak was judged to be ortho-n-dodecyltoluene according to its elution pattern with respect to n-tridecylbenzene (1). The possibility of 2-phenyltridecane (33) for the smaller peak was ruled out by the presence of daughter ions at m/z ) 133/134 and m/z ) 231/232 in the mass spectrum of this peak. By comparing the elution profiles of the di-n-alkylbenzenes
FIGURE 6. Selected ion chromatograms (m/z ) 260, 92, 105, 119, 161 and 175). m/z ) 260 depicts the C13 alkylbenzenes and C9 alkylbenzothiophenes. The open circle represents mono-n-alkylbenzene as it was also detected at m/z ) 92. The open triangles and squares represent the isomers of di- and tri-n-alkylbenzenes, respectively, as they were respectively detected at m/z ) 105 and 119. The solid squares and triangles represent the isomers of diand tri-n-alkylbenzothiophenes, as they were detected at m/z ) 161 and 171, respectively. previously reported (1), the smaller peak was assigned to be an isomer of n-propyl-n-decylbenzene. The three most abundant peaks in the m/z ) 119 chromatogram detecting tri-n-alkylbenzenes had their counterparts in the m/z ) 260 chromatogram (0 in Figure 6). These molecular species did not exhibit any characteristic fragment ions (doublets) other than m/z ) 119/120. Therefore, these peaks were judged to be the isomers of n-undecylxylenes. The two peaks detected in the m/z ) 161 chromatogram (9 in Figure 6) may be the two isomers of methyl-n-octylbenzothiophenes as they did not show any characteristic daughter ions other than m/z ) 161/162 in their mass fragmentation spectra. Similarly, the two peaks found in the m/z ) 175 chromatogram (2 in Figure 6) are presumed to have been isomers of dimethyl-nheptylbenzothiophenes as they also did not show any characteristic daughter ions other than m/z ) 175/176 in their mass fragmentation spectra. Another possible interpretation for these peaks is the presence of ethyl-n-heptylbenzothiophenes showing their base peaks at m/z ) 175. Elution Profile of the Alkylaromatics by Silica Gel Chromatography. The presence of a variety of long-sidechain alkylaromatics in fraction 2a was thus confirmed by the GC-MS analyses (Table 2). When the n-hexane:benzene fraction, which is considered to collect the aromatics, was analyzed by GC-MS, neither long-side-chain alkylbenzenes nor long-side-chain alkylbenzothiophenes could be detected, indicating that the long-side-chain alkylaromatics had been exclusively collected into the n-hexane fraction. VOL. 35, NO. 1, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
105
TABLE 2. Fate of the Long-Side-Chain Alkylaromatics after Biodegradation or Photooxidation Σ abundances
mono-n-alkylbenzenes di-n-alkylbenzenes tri-n-alkylbenzenes di-n-alkylbenzothiophenes tri-n-alkylbenzothiophenes
Σ abundances
selected ion (m/z)
control
biodegradation
depletion (%)
control
photooxidation
depletion (%)
92 105 119 161 175
2017 9606 9599 3063 2800
144 1613 1007 636 649
92.9 83.2 89.5 79.2 76.8
1780 8820 8774 2694 2519
1712 8387 8297 136 139
3.8 4.9 5.7 95.0 94.5
TABLE 3. Mass Balance of Fraction 2a after Biodegradation or Photooxidation (A) amt of n-hexane frac to the init amt of crude oila (%) initial crude oil biodegraded oil photooxidized oil
FIGURE 7. Elution profile of the crude oil components by n-hexane. The eluate was collected at 3-mL intervals for each fraction. Fraction 2 contained saturates, while fraction 3 contained both saturates and alkylaromatics. Fractions 4 to 7 contained alkylaromatics, while fractions 9 and 10 contained methylnaphthalenes, dibenzothiophene and derivatives. Since the presence of alkylaromatics in the n-hexane fraction has not previously been reported, we examined the elution profile of the crude oil components by n-hexane from silica gel chromatography. In this experiment, the n-hexane eluate was sampled each 3 mL (Figure 7), and the molecular species in each fraction were determined by GC-MS. No hydrocarbon was eluted in the first 3 mL, while only saturates (detected at m/z ) 85, 191) were eluted in the next 3 mL. In the fraction of 6-9 mL, both saturates and long-side-chain alkylaromatics (detected at m/z ) 92, 105, 119, 161 and 175), largely the alkylbenzenes were eluted. The ratio of saturates to long-side-chain alkylaromatics in this fraction was found to be 1:14 from TLC-FID analysis. After 9 mL, no saturates could be detected, but alkylaromatics were eluted in the fraction of 9-21 mL. In the fraction of 21-24 mL, only a minute amount of hydrocarbons was eluted, while a trace amount of aromatics exemplified by methylnaphthalenes and (methyl)dibenzothiophenes started to be eluted when more than 24 mL of n-hexane had been applied. This experiment thus demonstrates the difficulty in separating saturates and long-side-chain alkylaromatics by using only silica gel column chromatography and suggests the possible contamination of alkylaromatics in the n-hexane fractions from previously published studies. The present method for separating long-side-chain alkylaromatics in the n-hexane fraction by silica gel chromatography and subsequent preparative TLC in two steps seems to be simpler than the methods previously described. These previous methods involved either fractionating the alkylaromatics in the aromatic fraction by adsorption chromatography (eluting with n-hexane:dichloromethane ) 9:1, v/v) and subsequent preparative TLC, separating by two-step adsorption chromatography, or separating the organosulfur compounds possessing long-side-chain alkyl groups by multiple-step column chromatography (1, 4, 6, 8, 33). Fate of the Long-Side-Chain Alkylaromatics after Biodegradation and Photooxidation. We used this newly developed separation method to analyze the fate of the longside-chain alkylaromatics in crude oil after biodegradation or photooxidation. This method was equally suitable for the separation of biodegraded or photooxidized samples: fraction 2a of biodegraded or photooxidized samples could be 106
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 1, 2001
54 30.4 31.2
(C) amt of (D) amt of (B) amt of frac 28 to frac 2a to frac 2a in the init amt the init amt n-hexane of crude of frac 2ad fracb (%) oilc (%) (%) 33 50 8.3
17.8 15.2 2.6
100 85.3 14.5
a Determined gravimetrically (34). b Determined by TLC/FID: (amount of fraction 2a)/[(amount of fraction 1) + (amount of fraction 2a)] × 100. c [Column (A) × column (B)]/100. d Column (D)/0.178.
eluted by n-hexane by the same procedure, and the Rf values of the fraction 2a from these samples were identical to that of the original crude oil. The n-hexane fraction of biodegraded and photooxidized crude oil contained 30.4% and 31.2%, respectively, of the initial amount of crude oil (34). The TLC/FID chromatograms of the biodegraded and photooxidized samples respectively show ratios of peak 1 to peak 2a of 1:1 and 11:1. Thus, the concentrations of long-side-chain alkylaromatics in the biodegraded and photooxidized samples were estimated to be 15.2% and 2.6%, respectively, of the initial amount (Table 3). As already described, the initial concentration of longside-chain alkylaromatics in crude oil was 17.8%. We can thus conclude that a majority of the long-side-chain alkylaromatics in peak 2a are photosensitive, although they are rather resistant to biodegradation. Figures 3 and 4 (panel B & C each) show the GC-MS chromatograms of different series of long-side-chain alkylaromatics after their treatment by either biodegradation or photooxidation. The concentration of each compound in the n-hexane fraction was determined by GC-MS operated in SIM mode, using 17R(H),21β(H)hopane as an internal standard (Table 2). The resistance of 17R(H),21β(H)hopane toward biodegradation and photooxidation had already been demonstrated in the previous communications (34, 35). Their concentrations in fraction 2a could not be normalized by the hopane concentration as it was eluted in fraction 1. All the n-alkylbenzenes analyzed by the SIM chromatography at m/z ) 92, 105 and 119 were susceptible to biodegradation but were resistant to photooxidation (Table 2). Alkylbenzenes absorb UV light with λmax around 260 nm. The light used for the photooxidation treatment had an emission spectrum similar to that of solar radiation and did not emit short-UV light. The inert nature of the alkylbenzenes to photooxidation treatment can be easily explained. The small reduction in the amount of alkylbenzenes after the photooxidation treatment was mainly due to evaporation of the lower members of the series which was also observed in the dark control samples. A small reduction in the amount of long-side-chain alkylbenzenes with respect to the dark control was, however, apparent. This may have been due to photosensitized oxidation, because alkylbenzenes may be indirectly oxidized by light absorbed by other molecular species.
On the other hand, n-alkylbenzothiophenes monitored by the m/z ) 161 and 175 chromatograms (Table 2) were found to be quite susceptible to both photooxidation and biodegradation. The susceptibility to photooxidation of alkylbenzothiophenes was expected as they absorb light above 300 nm. The biodegradability of alkylaromtics of lower molecular weight was less than that of higher molecular weight alkylaromatics. One possibility to explain this result is that the terminal part of the n-alkyl side chain is more accessible to a catabolic enzyme if it is more distant from the benzene or benzothiophene nucleus (13, 14). Thus, the homologous series of n-alkylbenzenes and n-alkylbenzothiophenes were found to be susceptible to biodegradation. However, the overall biodegradation of alkylaromatics analyzed by TLC-FID was not extensive: the initial concentration of alkylaromatics in crude oil was 17.8%, while it was reduced to 15.2% after biodegradation (Table 3). Since a large proportion of alkylaromatics had been readily photooxidized, it was expected that the majority of the molecules would absorb light above 300 nm. In other words, a large part of this fraction may have been polycyclic aromatics with long alkyl side chains. Attempts were then made to detect components other than long-side-chain alkylbenzenes and alkylbenzothiophenes by GC using the SIM mode. However, we were not able to detect long-side-chain alkylnaphthalenes, alkylphenanthrenes, etc. We suppose that this fraction contains a large number of alkylaromatics, and the concentration of each molecular species is too low to be detected by GC-MS. The accurate estimation of the amounts of long-sidechain alkylbenzenes and alkylbenzothiophenes in fraction 2a was not possible due to the unavailability of suitable standards. Nevertheless, the amounts of long-side-chain alkylbenzenes and alkylbenzothiophenes in fraction 2a were roughly estimated to be 7% and 2%, respectively, by assuming that the relative response factors of these compounds in SIM chromatograms were equal to those of two commercially available standards, n-undecylbenzene and n-hexadecylbenzene. Then, it could be expected that long-side-chain alkylbenzenes and alkylbenzothiophenes (9% of fraction 2a) account for a major part of biodegraded compounds (14.7%, see Table 3) and that long-side-chain alkylbenzenes (7% of fraction 2a) account for a major part of photooxidationresistant compounds (14.5%, see Table 3). In conclusion, it has been shown that the long-side-chain alkylaromatics present in crude oil could be analyzed by using its n-hexane fraction with or without further separation by TLC. This simple method enabled the biodegradation and photooxidation of long-side-chain alkylaromatics to follow easily. One of the advantages of using the n-hexane fraction for quantifying individual long-side-chain alkylaromatic compounds is the availability of hopane that is present in the same fraction as an internal standard.
Acknowledgments This work was supported by New Energy and Industrial Technology Development Organization (NEDO).
Literature Cited (1) Sinninghe Damste´, J. S.; Kock-van Dalen, A. C.; Albrecht, P. A.; De Leeuw, J. W. Geochim. Cosmochim. Acta 1991, 55, 36773683.
(2) Ellis, L.; Singh, R. K.; Alexander, R.; Kagi, R. I. Geochim. Cosmochim. Acta 1995, 59, 5133-5140. (3) Solli, H.; Larter, S. R.; Douglas, A. G. In Advances in Organic Geochemistry; Douglas, A. G., Maxwell, R. J., Eds.; Pergamon: 1980; pp 591-597. (4) Ostroukhov, S. B.; Aref’yev, O. A.; Pustil’nikova, S. D.; Zabrodina, M. N.; Petrov, A. A. Pet. Chem. U.S.S.R. 1983, 23, 1-12. (5) Ellis, L.; Kagi, R. I.; Alexander, R. Org. Geochem. 1992, 18, 587593. (6) Dong, J.; Vorkink, W. P.; Lee, M. L. Geochim. Cosmochim. Acta 1993, 57, 837-849. (7) Radke, M.; Willsch, H. Fuel 1993, 72, 1103-1107. (8) Sinninghe Damste´, J. P.; de Leeuw J. W.; Kock-van Dalen, A. C.; de Zeeuw, M. A.; de Lange, F.; Rijpstra, W. I. C.; Schenck, P. A. Geochim. Cosmochim. Acta 1987, 51, 2369-2391. (9) Sinninghe Damste´, J. P.; Rijpstra, W. I. C.; de Leeuw J. W.; Schenck, P. A. Geochim. Cosmochim. Acta 1989, 53, 1323-1341. (10) Ogata, M.; Miyake, Y. J. Chromtogr. Sci. 1980, 18, 594-605. (11) Ogata, M.; Fujisawa, K. Bull. Environ. Contam. Toxicol. 1990, 44, 884-891. (12) Eganhouse, R. P. Int. J. Environ. Anal. Chem. 1986, 26, 241-263. (13) Bayona, J. M.; Albaiges, J.; Solanas, A. M.; Griffol, M. Chemosphere 1986, 15, 595-598. (14) Takada, H.; Ishiwatari, R. Environ. Sci. Technol. 1990, 24, 8691. (15) Eganhouse, R. P.; Blumfield, D, L.; Kaplan, I. R. Environ. Sci. Technol. 1983, 17, 523-530. (16) Sigoillot, J.-C.; Nguyen, M.-H. Appl. Environ. Microbiol. 1992, 58, 1308-1312. (17) Davis, J. B.; Raymond, R. L. Appl. Microbiol. 1961, 9, 383-388. (18) Sariaslani, F. S.; Harper, D. B.; Higgins, I. J. Biochem. J. 1974, 140, 31-45. (19) Amund, O. O.; Higgins, I. J. Antonie van Leeuwenhoek 1985, 51, 45-56. (20) Rontani, J. F.; Rambeloarisoa, E.; Bertrand, J. C.; Giusti, G. Mar. Environ. Res. 1985, 16, 301-314. (21) Rontani, J. F.; Bonin, P.; Giusti, G. Mar. Chem. 1987, 22, 1-12. (22) Kropp, K. G.; Fedorak, P. M. Can. J. Microbiol. 1998, 44, 605622. (23) Oldfield, C.; Wood, N. T.; Gilbert, S. C.; Murray, F. D.; Faure, F. R. Antonie Van Leeuwenhoek 1998, 74, 119-132. (24) Frassinetti, S.; Setti, L.; Corti, A.; Farrinelli, P.; Montevecchi, P.; Vallini, G. Can. J. Microbiol. 1998, 44, 289-297. (25) Eaton, R. W.; Nitterauer, J. D. J. Bacteriol. 1994, 176, 39924002. (26) Shiraishi, Y.; Hara, H.; Hirai, T.; Komasawa, I. Ind. Eng. Chem. Res. 1999, 38, 1589-1595. (27) Shiraishi, Y.; Taki, Y.; Hirai, T.; Komasawa, I. Ind. Eng. Chem. Res. 1999, 38, 3310-3318. (28) Andersson, J. A.; Bobinger, S. Polycyclic Aromat. Compd. 1996, 11, 145-151. (29) Wang, Z.; Fingas, M.; Li, K. J. Chromatogr. Sci. 1994, 32, 361366. (30) Pollard, S. J.; Hrudey, S. E.; Fuhr, B. J.; Alex, R. F.; Holloway, L. R.; Tosto, F. Environ. Sci. Technol. 1992, 26, 2528-2534. (31) Sugiura, K.; Ishihara, M.; Shimauchi, T.; Harayama, S. Environ. Sci. Technol. 1997, 31, 45-51. (32) Wang, Z.; Fingas, M.; Blenkinsopp, S.; Sergy, G.; Landriault, M.; Sigouin, L.; Foght, J.; Semple, K.; Westlake, D. W. S. J. Chromatography A 1998, 809, 89-107. (33) Ellis, L.; Langworthy, T. A.; Winans, R. Org. Geochem. 1996, 24, 57-69. (34) Dutta, T. K.; Harayama, S. Environ. Sci. Technol. 2000, 34, 15001505. (35) Garrett, R. M.; Pickering, I. J.; Haith, C. E.; Prince, R. C. Environ. Sci. Technol. 1998, 32, 3719-3723.
Received for review April 7, 2000. Revised manuscript received September 13, 2000. Accepted September 28, 2000. ES001165A
VOL. 35, NO. 1, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
107
