- Email: [email protected]
Effects of short-term weathering on the stable carbon isotope compositions of crude oils and fuel oils
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Effects of short-term weathering on the stable carbon isotope compositions of crude oils and fuel oils Yu Liua,⁎, Jixiang Xua,b, Wenjing Chena, Ying Lic a b c
College of Environmental Science and Engineering, Dalian Maritime University, Dalian 116026, China College of Environmental Science and Engineering, Maritime Safety Administration of the People's Republic of China, Beijing 100736, China College of Navigation, Dalian Maritime University, Dalian 116026, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Oil spills n-Alkanes Diagnostic ratios Stable carbon isotope Short-term weathering
A short-term simulated weathering experiment was performed on two crude oils and two heavy fuel oils under natural conditions to evaluate the effects of natural weathering processes by using gas chromatography–mass spectrometry combined with gas chromatography–isotopic ratio mass spectrometry. The results of diagnostic ratios of n-alkanes show that only odd to even predominance (OEP1, OPE2) and carbon preference index (CPI) remain stabilized during the 28 d weathering process, but they cannot effectively distinguish the four types of oils. Statistical analyses based on paired sample t-test and principal component analysis (PCA) revealed that stable carbon isotope compositions of n-alkanes in the four studied oils have no significant changes over the weathering time, and that the carbon isotope discrimination (Δδ13C) of n-alkanes is < 3‰. We have provided evidence that the stable carbon isotope compositions of n-alkanes compared to n-alkanes diagnostic ratios significantly improve the efficiency and fidelity of the oil fingerprint identification.
1. Introduction Statistical data for the past few years show that the average number of large oil-spill accidents (> 700 tons) resulting from oil exploration, shipping transportation, collisions, and unexpected events have gradually reduced, averaging at 1.8 large oil spills per year since 2010. However, the total recorded amount of oil spilled in 2015 has been > 7000 tons (ITOPF, 2015). Oil spills have physical effects and chemical toxicity, causing serious damage to marine and terrestrial ecosystems, human health, and natural resources (Beyer et al., 2016). Oil spilled into a marine environment is subjected to weathering processes, which strongly deplete the concentration of petroleum hydrocarbons and change the oil's chemical component distributions (Ezra et al., 2000; Yim et al., 2011; Samuels et al., 2013). For example, moderately biodegraded oils show an unresolved complex mixture (UCM), loss of short-chain n-alkanes (n < C15), and moderate change in alkyl naphthalene distributions (Asif et al., 2009). Prince et al. (2002) found that the maximal total extent of weathering loss of hydrocarbon from an Arctic oil spill by biodegradation can reach 87%; in this process, approximately half of the four-ring chrysene series compounds were found to be heavily degraded by photooxidation over 20 years. Gas chromatography–flame ionization detection GC–FID and gas chromatography–mass spectrometry (GC–MS) are still the most widely used
⁎
analytical techniques for oil spill fingerprinting. In these techniques, GC–FID and GC–MS profiles, source-specific target analytes, and diagnostic ratios of selected biomarkers are determined to trace spill sources (Fernández-Varela et al., 2009; Wang, Hu, He, Liu, & Zhao, 2013; Retnam et al., 2015). However, conventional chromatography cannot identify and quantify many compounds in UCMs below the baseline because of resolution and sensitivity limitations. These contain a large number of co-eluting compounds at low concentrations and overlap with the high-concentration chromatographic peaks, increasing the analytical uncertainty, especially in analyses of highly weathered oils, in which the concentration of petroleum hydrocarbons is lower than that in the original samples (Bayona et al., 2015; Wang, Chen, Zhang, He, & Zhao, 2013). Compound-specific isotope analysis (CSIA, a highly specialized and complementary analytical technology (O'Malley et al., 1997), has been extensively applied in environmental forensic investigations related to petroleum pollution. It has also been applied in organic geochemistry, being used in inferring source materials for hydrocarbons, as well as in analysis of depositional environments and maturity (Maioli et al., 2012; Al-Areeq and Maky, 2015; Li and Guo, 2010). For example, Harvey et al. (2012) distinguished four diesel fuels from different sources or locations by measuring the carbon and hydrogen isotope ratios of nalkanes, showing that compound-specific isotope analysis is a powerful
Corresponding author. E-mail address: [email protected] (Y. Liu).
http://dx.doi.org/10.1016/j.marpolbul.2017.04.003 Received 19 December 2016; Received in revised form 24 March 2017; Accepted 3 April 2017 0025-326X/ © 2017 Elsevier Ltd. All rights reserved.
Please cite this article as: Liu, Y., Marine Pollution Bulletin (2017), http://dx.doi.org/10.1016/j.marpolbul.2017.04.003
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
tool for distinguishing different fuel samples using δ13C and δD values. Hall et al. (2014) determined the origin of asphaltites with moderate or severe weathering obtained from the coast of southern Australia on the basis of biomarker ratios and n-alkane isotopic profiles, confirming they have the same source. Petroleum is mainly composed of aliphatic hydrocarbons and aromatic hydrocarbons, with aliphatic hydrocarbons (n-alkanes, branched alkanes, and cycloalkanes) usually occurring at the highest concentration in petroleum hydrocarbon, accounting for 50% of the oil (Tissot and Welte, 1984). However, the saturated components in oils are most vulnerable to weathering such as evaporation, photooxidation, and biodegradation, increasing the concentration of asphaltene and polar components in residual oil (Prince et al., 2003; Aeppli et al., 2012). In general, the loss of whole oil is mainly controlled by saturated hydrocarbons degradation, and the loss of the fuel oils is mainly dominated by the reduction of aromatic hydrocarbons (Li and Xiong, 2009). Despite its high specificity, CSIA may be limited when the oil spills show significant fractionation of stable carbon isotopes in moderate or severe biodegradation processes (Wang, Gao, Sun, Qin, Yin, & He, 2013). Studies indicate that n-alkanes are oxidized by aerobic microorganisms to oxidation products of alkyl alcohol, alkyl acid, and CO2 via a series of transformation pathways. The 12Ce12C bond usually has bond energy lower than that of 12Ce13C and 13Ce13C bonds prior to breaking in the aerobic biodegradation process, leading to 13C-enrichment in the residual saturate fraction (Xiao et al., 2012; Sun et al., 2005; Wilkes et al., 2000). Therefore, fundamental research is essential to understanding the relationship between the weathering processes and the stable isotopic composition of n-alkanes as well as to determine whether carbon isotope ratios can be used as a weatheringresistance indicator. The effects on carbon isotopic fractionation have rarely been studied by simulation experiments. The aim of this study is combine the methods gas chromatography–mass spectrometry with gas chromatography–isotopic ratio mass spectrometry to investigate the simultaneous changes in chemical composition and carbon stable isotope ratios of oil spills during the weathering process. A short-term 28-day simulated weathering experiment under natural conditions was performed on two types of oils (crude oils and fuel oils). Weathering characteristics were also analyzed and compared. Our results may provide technical support for studies on identification and characterization weathered oil spills.
Oil samples were taken periodically on the 0th, 1st, 3rd, 7th, 14th, 21st, and 28th d. All oil samples were stored in prewashed amber bottles at 4 °C in a refrigerator before experiments. In the weathering process, the average temperature, average wind speed, and average humidity were − 1.48 ± 3.55 °C, 7.23 ± 1.87 m/s, and 58 ± 4%, respectively. 2.2. Sample preparation About 0.2 g of the oil sample was dissolved in 10 mL of hexane (HPLC grade, Tedia, Fairfield, USA) and then was centrifuged at 1409g for 10 min. The fractionation of saturates was conducted in a chromatographic column (0.47 cm i.d. × 12 cm). The column was dry-packed with 10 g of activated alumina (activated for 4 h at 200 °C, 100–200 mesh, AR) and was topped with a 1 cm thick layer of anhydrous sodium sulfate (activated for 4 h at 350 °C, AR). The column was conditioned with hexane and the eluent was discarded just prior to exposure of the sodium sulfate layer to air. About 200 μL of concentrated extract was quantitatively transferred. The aliphatic hydrocarbons were eluted with 15.0 mL of hexane. The n-alkanes were further isolated from branched hydrocarbon fractions by urea adduction (Barrie et al., 2016). 2.3. GC–MS All processed samples were analyzed on mass spectrometer (ISQ, Thermo Fisher Scientific, USA) interfaced with a gas chromatograph (Thermo Fisher Scientific, CA, USA) equipped with a DB-5MS capillary column (60 m × 0.25 mm × 0.25 μm; Agilent, Santa Clara, CA, USA). He (99.999% purity) was used as the carrier gas and introduced at a flow rate of 1.2 mL/min in constant-flow mode. The GC temperature program was started at 60 °C, which was ramped to 100 °C at 20 °C/ min, held for 2 min, then from 100 to 280 °C at 6 °C/min, and held for 35 min isothermally. Injection was done with a volume of 1.0 mL in splitless mode. For MS, the injector, interface, and ion source were held at 280 °C, 250 °C, and 200 °C, respectively. The ion source was set in electron ionization mode at 70 eV. The mass spectrometer was operated in full scan mode. Scanning was done within the range of 50–550 amu at a rate of 0.2 s/scan. n-Alkanes were identified on the basis of the retention time of a standard mixture of n-alkanes (nC10 − nC32, Sigma, USA). 2.4. Stable isotope analysis
2. Materials and methods CSIA was performed using a DELTA V stable isotope ratio mass spectrometer interfaced to a gas chromatograph (Thermo Fisher Scientific). The GC instrument parameters and the temperature program are consistent with the GC–MS analytic conditions described above. Isotopic analyses were carried out at the Environmental Information Institute, Dalian Maritime University (Dalian, China). The isotope ratio for carbon was expressed as follows:
2.1. Weathering simulation experiment A simulated weathering experiment was performed on crude oils and fuel oils (Table 1). Four different oils (each oil about 8 g) were separately added to four 2000 mL beakers with 1600 mL seawater. The seawater was pre-filtered through four layers of nylon cloth (500 mesh). The beakers were kept on an outdoor open-air platform for 28 d under natural conditions in order to approach the marine field environment.
δ13C = [ (Rsample Rstandard ) − 1] ×10 3 where R represents 13C/12C. The δ13C value is relative to that of Vienna Pee Dee belemnite. The analytical precision of δ13C was < 0.06‰.
Table 1 Data of the oils samples. Sample
Locations
Oil species
API gravitya (°)
Viscosity (50 °C, mm2/s)
IBPb (°C)
WAFc (mg/L)
KWT AM 180# 380#
Kuwait Oman China China
Crude oil Crude oil Fuel oil Fuel oil
31.8 34.6 11.3 11.3
6.965 7.987 180 380
55 60 133 213
8.96 8.12 14.35 12.62
3. Results and discussion 3.1. Effects of weathering on the diagnostic ratios of n-alkanes In this study, nine kinds of diagnostic ratios (Table 2) were selected to monitor weathering, to interpret chemical data from oil spills, and to differentiate or determine a match between the crude oils and fuel oils. Here, diagnostic ratios parameters were calculated from the chromatographic peak areas of n-alkane components; therefore, ratios only represent an approximation of the relative concentrations of each nalkane fraction. Fig. 1 and Table 2 show the variations of nine
Note: WAF was quantified by UV spectroscopy using a UV1102 (Shanghai Techcomp Instrument Co.) according to GB 17378.4-2007. a API gravity = 141.5/specific gravity (60 °F) − 131.5. b IBP: initial boiling point. c WAF: water accommodated fraction.
2
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
Table 2 Changes in diagnostic ratios of n-alkanes in the weathering process. Sample no.
Ratios
n-C16/n-C18a n-C17/n-C18b (nC19 + nC20) / (nC19 − nC22)c LMW/HMWd OEP1e OEP2f (nC19 + nC20) / (nC21 + nC22)g (nC13 + nC14) / (nC25 + nC26)h CPIi
KWT crude oils
AM crude oils
180# fuel oils
380# fuel oils
0d
28 d
0d
28 d
0d
28 d
0d
28 d
1.35 1.28 0.57 2.36 1.09 1.04 1.35 2.76 1.28
1.26 1.24 0.57 1.84 1.08 1.03 1.34 1.99 1.22
1.17 1.14 0.56 0.38 1.00 1.04 1.27 2.84 1.27
0.96 1.05 0.53 1.54 1.02 1.03 1.14 0.98 1.21
1.09 1.11 0.64 1.87 1.00 1.01 1.77 2.31 1.23
0.69 0.87 0.46 0.78 0.99 1.10 0.86 0.58 1.16
0.79 0.93 0.53 1.67 1.02 1.05 1.17 2.07 1.24
0.70 0.90 0.47 0.94 1.00 0.99 0.87 0.59 1.18
Note: Values represents the average value of the three measurement results. d LMW/HMW = LMW(n-C11 − n-C21)/HMW(n-C22 − n-C32). e OEP1 = (C17 + 6 × C19 + C21) / (4 × C18 + 4 × C20). f OEP2 = (C21 + 6 × C23 + C25) / (4 × C10 + 4 × C24). i CPI = [C13 + C15 + C17 + C19 + C21 + C23) / (C14 + C16 + C18 + C20 + C22 + C24).
all ratios for the original and weathered oils substantially decreased, except for OEP1, OEP2, and CPI ratios (Fig. 1). The relative standard deviation (RSD) was used to evaluate the stability of diagnostic indices of saturated hydrocarbon; its acceptable value is < 5% (Ho et al., 2015). As shown in Fig. 2, the OEP1, OEP2,
diagnostic ratios for the two types of oils in the weathering process. The diagnostic ratios for two crude oils (KWT and AM) except for (nC13 + nC14) / (nC25 + nC26) and LMW/HMW ratios apparently did not change during the short-term weathering process (Table 2). In contrast, two fuel oils (180# and 380#) have distinct changes in ratios;
Fig. 1. Changes in diagnostic ratios of n-alkanes from the four oil samples over weathering period. (A) KWT crude oils; (B) AM crude oils; (C) 180# fuel oils; (D) 380# fuel oils. Ratios A ~ I represent n-C16/n-C18, n-C17/n-C18, (nC19 + nC20) / (nC19 − nC22), LMW/HMW, OEP1, OEP2, (nC19 + nC20) / (nC21 + nC22), (nC13 + nC14) / (nC25 + nC26), and CPI, respectively; means are results of triplicate analyses.
3
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
(R2 values of 0.95 and 0.83, respectively). The two fuel oils with WAF values higher than those of the two crude oils display relatively larger levels of degradation of n-alkanes (Fig. 3). Studies have shown that photodegradation and biodegradation are not the main factors that influence the loss of n-alkanes in the short-term weathering process; evaporation and dissolution have significant effects on the n-alkane components (Bacosa et al., 2015; Xiao et al., 2012; He et al., 2016). API gravity show an inverse correlation with the levels of n-alkane degradation in some of the oils (Asif et al., 2009), but the two kinds of fuel oils (180# and 380#) with the same API values have clearly different degrees of weathering in this study. This difference is probably because 180# fuel oil has solubility (14.35 mg/L) higher than that of 380# fuel oil, making it more susceptible to weathering. It shows that that in addition to evaporation, dissolution is also an important factor affecting the composition of n-alkanes in short-term weathering processes (Li et al., 2009), but the effects of dissolution on short-term degradation of alkanes need further investigation. The data suggest that the OEP1, OEP2, and CPI ratios showed little change, but the other selected ratios of n-alkanes were remarkably affected by weathering. The RSD of LMW/HMW and (nC13 + nC14) / (nC25 + nC26) are > 5% in this study, showing that they cannot be used in oil source identification of weathered oil spills. Furthermore, OEP1, OEP2, and CPI ratios do not differ significantly between the four types of oils (P > 0.05), which cannot be effectively distinguish by diagnostic ratios. Hence, the consistency of these ratios in various oils should be considered when they are used to distinguish different kinds of oil (Liu et al., 2015).
Fig. 2. RSD (%) of diagnostic ratios for n-alkanes in different oils.
and CPI ratios of both fuel oils and crude oils have low RSDs (< 5%), which indicate that they are not significantly affected by weathering. In contrast, LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios of both oils decreased markedly within 0–30 d, and their relative standard deviations are > 9.69% (Fig. 2). This finding shows that weathering has an impact on LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios; therefore, they should be used cautiously in the identification of spilled oils subjected to moderate (30 d) or serious (> 30 d) weathering in order to avoid misjudgment. Moreover, the changes in LMW/ HMW and (nC13 + nC14) / (nC25 + nC26) ratios mainly occurred in the first seven days of weathering, with the changes progressively slowing down (Fig. 1). This result agrees with the conclusions of previous studies that weathering mainly occurs in the first few days of the oil spill (Lemkau et al., 2010). In the present study, the diagnostic indices (RSDs) of n-C16/n-C18, n-C17/n-C18, (nC19 + nC20) / (nC1 ~ nC22), and (nC19 + nC20) / (nC21 + nC22) from crude oils are < 5%; in contrast, those of the fuel oils are > 5%. Since the RSDs of these indices are distinctly higher than the former, the fuel oils are more susceptible to weathering (Fig. 2). Clearly, oils of the same type have similar weathering patterns, indicating that changes in diagnostic ratios may be correlated with the types of oils (Figs. 1 and 2). We found a positive correlation between the degree of weathering for the n-alkanes and the water accommodated fraction WAF values in this study, as shown in Fig. 3. A plot of WAF values versus the RSD of LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios show a strong linear correlation
3.2. Effects of weathering on the carbon isotopic composition of the nalkanes The changes in stable carbon isotopic compositions of individual nalkanes from crude oils and fuel oils in the weathering simulation process are shown in Fig. 4. The four kinds of oils have different stable carbon isotopic compositions, including carbon stable isotope ratios (δ13C values) and carbon isotope curves of n-alkanes. The 180# and 380# fuel oils have similar stable carbon isotopic compositions (Fig. 4C and D). They have relatively flat carbon isotope profiles with increasing n-alkanes carbon number, and their distribution ranges of δ13C values are − 28‰ to −24‰ and − 29‰ to − 25‰, respectively. Oil-AM has relatively low δ13C values, ranging from − 35‰ to 28‰ almost constantly or becoming slightly larger with increasing carbon number (Fig. 4C). Fig. 4A clearly shows that the carbon isotope ratios of nalkanes in KWT crude oils are widely distributed, with δ13C values ranging from −32‰ to −21‰. δ13C values first decrease with
Fig. 3. Relationship between WAF and RSD of diagnostic ratios of n-alkanes. (A) WAF value versus the RSD of LMW/HMW; (B) WAF value versus the RSD of (nC13 + nC14) / (nC25 + nC26).
4
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
-20
(A)
(B)
-26
-22
-28 -24
-30 -26
-32 KWT-0d KWT-1d KWT-3d KWT-7d KWT-14d KWT-21d KWT-28d
-30 -32 -34
δ13
δ13
-28 AM-0d AM-1d AM-3d AM-7d AM-14d AM-21d AM-28d
-34 -36 -38
-40 C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
n-Alkanes -20
-20
(C)
(D)
-22
-22
-24
-24
-26
-26 -28 180#-0d 180#-1d 180#-3d 180#-7d 180#-14d 180#-21d 180#-28d
-30 -32 -34
δ13
δ13
-28
380#-0d 380#-1d 380#-3d 380#-7d 380#-14d 380#-21d 380#-28d
-30 -32 -34
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
Fig. 4. Carbon isotopic profiles of n-alkanes from the four oil samples and the corresponding weathering processes for the 1 d, 3 d, 7 d, 14 d, and 28 d samples (values represent mean values, n = 3). (A) KWT crude oil and its weathered oils; (B) AM crude oil and its weathered oils; (C) 180# fuel oil and its weathered oils; and (D) 380# fuel oil and its weathered oils.
increasing carbon number and then increase after the C19 alkane, resulting in a v shape in the profiles. Visually, KWT crude oil is unique among the samples; its carbon isotopic profile of n-alkanes shows a v shape with respect to carbon number. It can be distinguished from other oils on the basis of carbon isotope profiles of individual n-alkanes (Cheng et al., 2015). Oils from the same source have similar stable isotope curves. Different sources of oils have different curve shapes; thus, the rules can be used for oil–oil correlation and oil–source correlation (Murray et al., 1994; Odden et al., 2002). Results indicate that the four oil samples did not show a significant carbon isotopic fractionation with weathering. The carbon isotope discrimination Δδ13C (Δδ13C = δ13C(t) − δ13C(0)) of individual n-alkanes are < 3‰ (Fig. 4). The carbon isotope curves for the individual n-alkanes from both fuel oils and crude oils also did not extensively change with weathering process. This conclusion is in agreement with results of the three independent weathering experiments conducted by Mansuy et al. (1997), which revealed that the δ13C values of n-alkanes are minimally affected by evaporation, water washing, or biodegradation.
μ=− χ±
ts n
where μ is the population mean, where − x is the mean value of multiple parallel analysis results, s is the standard deviation of multiple parallel analysis results, n is the number of parallel analyses, and t is the t value that can be determined from a t critical value table based on the confidence limit (0.05 for a 95% confidence interval) and the degrees of freedom (n − 1), Results of the t-test of the non-weathered oil samples (KWT, AM, 180#, and 380#) and their corresponding oils weathered for 28 d are shown in Fig. 5. Most of the points are mainly distributed on a straight line (y = x); meanwhile, the error bars for all points of the original oil and the corresponding weathered oil cross the y = x straight line (CL = 95%) for each oil, indicating that the four kinds of oils and their weathered oils can be considered as the same oil sample or from the same source (Sun et al., 2009). Results of the t-test suggest that weathering and non-weathered oils can still completely match after moderate weathering, proving that stable carbon isotopes ratios can be used in the identification of moderately weathered oils. Overall, the δ13C values and the carbon isotope curves for the individual n-alkanes from the oils do not significantly change in the short-term weathering process.
3.3. t-Test analysis In order to further examine the weather resistance and applicability to oil spill source identification of the δ13C values of individual nalkanes (nC11 − nC32), a Student's t-test and scatter plot were used to analyze and compare the correlation between the original oil and 28 d weathered samples at 95% confidence limit (CL = 95%). The equation is expressed as
3.4. PCA Principal component analysis (PCA) is the most widely used and powerful statistical analysis technique in oil fingerprint identification (Wang and Fingas, 2003). In order to further study the correlation between the four kinds of oils and their weathered oils, PCA based on 5
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al. -20
-28
KWT crude oils —— y=x
-29
C13 C14 C12
C31 C16 C32
-26 C25
C28 C27 C26 C29 C17
C30
-28 C18
-30
-32 -32
C29
-31
C27 C12 C26 C13 C28 C25 C14 C11 C15
-32
C24 C21 C22 C16 C23 C18 C17
-33
-26
-24
-22
C20
-34
(A) -28
(B) -35 -35
-20
-34
-33
Weathered oil for 28 days -24
-24 -25
-28
-30
-29
-28
380# fuel oils —— y=x
C15 C16 C14 C17 C13 C28 C18 C11 C19C25 C24 C26 C21 C27 C22 C20 C12 C23
-27 -28 C29 C32 C31
-29 C30
-29
-30 -30
-31
-26
C31 C32 C15 C12C13 C16 C11 C17 C14 C19C30 C20 C26 C18 C22 C25 C29 C21 C23 C28 C27 C24
Not-weathered oil
Not-weathered oil
-25
-27
-32
Weathered oil for 28 days
180# fuel oils —— y=x
-26
C30
C31C32
C19
C24 C23C20 C22 C21 C19
-30
AM crude oils —— y=x
-30
C11 C15
-24
Not-weathered oil
Not-weathered oil
-22
-30
(C) -29
-28
-27
-26
-25
(D) -24
Weathered oil for 28 days
-31 -31
-30
-29
-28
-27
-26
-25
-24
Weathered oil for 28 days
Fig. 5. Correlation of the δ13C values of n-alkanes (nC11 − nC32) between non-weathered oil and its weathered counterpart for 28 d at 95% confidence levels: KWT samples (A), AM (B), ts , where n = 3, t = 4.303, and s is the standard deviation. 180# (C), and 380# (D). Error bars are calculated by n
corner of the double plot, with relatively heavy isotope values. Studies have indicated that the strong anti-weathered indicators OEP1, OEP2, and CPI (RSD < 5%) show no difference between 180# and 380# (Table 2), which cannot be effectively differentiated by diagnostic indicators. However, we can effectively distinguish between 180# and 380# by stable isotope composition through the cross-plot diagram, as shown in Fig. 6. The carbon isotope values of long- and short-chain nalkanes of 180# are heavier than those of 380#, while the carbon isotope ratio values of middle-carbon-number n-alkanes from 180# are lower than those of 380#. Oils and their corresponding weathered samples at different weathering periods cluster together closely, further proving that there is little isotopic fractionation during short-term weathering. Analytical data from PCA imply that the two crude oils and two heavy fuel oils, which resemblance each other in appearance, can be distinguished in the scattered plots. In conclusion, we have provided evidence that the source apportionment of spilled oils after weathering can be advanced by statistical analysis combined with CSIA.
2
1
PC2
0
-1
KWT AM 180# 380#
-2
-3 -8
-6
-4
-2
0
2
4
PC1 Fig. 6. PCA results based on the δ13C values of n-alkanes (nC11 − nC32) in non-weathered and weathered oils.
the δ13C of alkanes (nC11 − nC32) was conducted by using SPSS 19.0. The PCA results are shown in Fig. 6. The first and second principal components respectively explain 77.6% and 15.3% of the variance, and the accumulated contribution ratio of the two principal components is 92.9%. Factor analysis showed that the low carbon number (n < C15) and high carbon number of n-alkanes (n > C25) in the first principal component have larger loadings, and the middle carbon number nalkanes have heavy loading in the second principal component. Fig. 3 clearly indicates that oil-AM and oil-KWT are the endpoints for the four group oils. The fuel oil samples 180# and 380# lie in the upper right
4. Conclusions The weathering simulation experiment revealed that the LMW/ HMW and (nC13 + nC14) / (nC25 + nC26) ratios are markedly dependent on weathering(RSD) > 5%, indicating that these may not be valid indices for oil source identification of the spill oil after degradation. The OEP1, OEP2, and CPI ratios change little, but they cannot effectively distinguish the four types of oil samples. High WAF values are consistent with relatively high levels of degradation of n-alkanes in the four oils. Paired sample t-test results show that the four oil samples 6
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
ITOPF (International Tanker Owners Pollution Federation), 2015. Oil Tanker Spill Statistics. (Available from: < http://www.itopf.com/knowledge-resources/datastatistics/statistics/ > . Accessed 26 December 2015). Lemkau, K.L., Peacock, E.E., Nelson, R.K., Ventura, G.T., Kovecses, J.L., Reddy, C.M., 2010. The M/V Cosco Busan spill: source identification and short-term fate. Mar. Pollut. Bull. 60 (11), 2123–2129. Li, S.M., Guo, D., 2010. Characteristics and application of compound specific isotope in oil-source identification for oils in Dongying Depression, Bohai Bay Basin. Geoscience 24, 252–258. Li, Y., Xiong, Y., 2009. Identification and quantification of mixed sources of oil spills based on distributions and isotope profiles of long-chain n-alkanes. Mar. Pollut. Bull. 58 (12), 1868–1873. Li, Y., Xiong, Y., Yang, W., Xie, Y., Li, S., Sun, Y., 2009. Compound-specific stable carbon isotopic composition of petroleum hydrocarbons as a tool for tracing the source of oil spills. Mar. Pollut. Bull. 58 (1), 114–117. Liu, X., Ma, Q., Yang, X., 2015. Study on consistency of diagnostic ratios of different oils. Aquatic Procedia 3, 231–237. Maioli, O.L.G., de Oliveira, C.R., Dal Sasso, M.A., dos Santos Madureira, L.A., de Almeida Azevedo, D., de Aquino Neto, F.R., 2012. Evaluation of the organic matter sources using the δ13C composition of individual n-alkanes in sediments from Brazilian estuarine systems by GC/C/IRMS. Estuar. Coast. Shelf Sci. 114, 140–147. Mansuy, L., Philp, R.P., Allen, J., 1997. Source identification of oil spills based on the isotopic composition of individual components in weathered oil samples. Environ. Sci. Technol. 31 (12), 3417–3425. Murray, A.P., Summons, R.E., Boreham, C.J., Dowling, L.M., 1994. Biomarker and nalkane isotope profiles for Tertiary oils: relationship to source rock depositional setting. Org. Geochem. 22 (3–5), 521IN5–542IN6. Odden, W., Barth, T., Talbot, M.R., 2002. Compound-specific carbon isotope analysis of natural and artificially generated hydrocarbons in source rocks and petroleum fluids from offshore Mid-Norway. Org. Geochem. 33 (1), 47–65. O'Malley, V.P., Burke, R.A., Schlotzhauer, W.S., 1997. Using GC–MS/combustion/IRMS to determine the 13C/12 C ratios of individual hydrocarbons produced from the combustion of biomass materials—application to biomass burning. Org. Geochem. 27 (7), 567–581. Prince, R.C., Owens, E.H., Sergy, G.A., 2002. Weathering of an Arctic oil spill over 20 years: the BIOS experiment revisited. Mar. Pollut. Bull. 44 (11), 1236–1242. Prince, R.C., Garrett, R.M., Bare, R.E., Grossman, M.J., Townsend, T., Suflita, J.M., et al., 2003. The roles of photooxidation and biodegradation in long-term weathering of crude and heavy fuel oils. Spill Sci. Technol. Retnam, A., Kassim, A.M., Ahmad, W.K.W., 2015. Fingerprinting of light fuel oil: a Malaysia case study. Prog. Environ. Sci. 30, 190–194. Samuels, W.B., Amstutz, D.E., Bahadur, R., Ziemniak, C., 2013. Development of a global oil spill modeling system. Earth Sci. Res. 2 (2), 52. Sun, Y., Chen, Z., Xu, S., Cai, P., 2005. Stable carbon and hydrogen isotopic fractionation of individual n-alkanes accompanying biodegradation: evidence from a group of progressively biodegraded oils. Org. Geochem. 36 (2), 225–238. Sun, P., Bao, M., Li, G., Wang, X., Zhao, Y., Zhou, Q., Cao, L., 2009. Fingerprinting and source identification of an oil spill in China Bohai Sea by gas chromatography-flame ionization detection and gas chromatography–mass spectrometry coupled with multistatistical analyses. J. Chromatogr. A 1216 (5), 830–836. Tissot, B.P., Welte, D.H., 1984. Petroleum Formation and Occurrence . Wang, Z., Fingas, M.F., 2003. Development of oil hydrocarbon fingerprinting and identification techniques. Mar. Pollut. Bull. 47 (9), 423–452. Wang, C., Gao, X., Sun, Z., Qin, Z., Yin, X., He, S., 2013a. Evaluation of the diagnostic ratios for the identification of spilled oils after biodegradation. Environ. Earth Sci. 68 (4), 917–926. Wang, C., Chen, B., Zhang, B., He, S., Zhao, M., 2013b. Fingerprint and weathering characteristics of crude oils after Dalian oil spill, China. Mar. Pollut. Bull. 71 (1), 64–68. Wang, C., Hu, X., He, S., Liu, X., Zhao, M., 2013c. Source diagnostic and weathering indicators of oil spills utilizing bicyclic sesquiterpanes. Acta Oceanol. Sin. 32 (8), 79–84. Wilkes, H., Boreham, C., Harms, G., Zengler, K., Rabus, R., 2000. Anaerobic degradation and carbon isotopic fractionation of alkylbenzenes in crude oil by sulphate-reducing bacteria. Org. Geochem. 31 (1), 101–115. Xiao, Q., Sun, Y., Zhang, Y., Chai, P., 2012. Stable carbon isotope fractionation of individual light hydrocarbons in the C 6–C 8 range in crude oil as induced by natural evaporation: experimental results and geological implications. Org. Geochem. 50, 44–56. Yim, U.H., Ha, S.Y., An, J.G., et al., 2011. Fingerprint and weathering characteristics of stranded oils after the Hebei Spirit oil spill. J. Hazard. Mater. 197, 60–69.
for 28d also can be well matched to its original oil samples, indicating that the stable carbon isotopic compositions did not exhibit considerable isotope fractionation in the short-term weathering process. PCA based on the δ13C of alkanes (nC11 − nC32) showed that the four oil samples at different weathering periods with corresponding original oils can cluster together in the scatter plots respectively, which cannot be distinguished by chemical composition. This demonstrated that statistical analysis combined with carbon isotope fingerprints can markedly increase the ability of oil spill characterization. In conclusion, the chemical compositions of n-alkanes are more likely to be affected than are the isotopic compositions of n-alkanes by the weathering process, and the stable isotopic compositions relative to chemical compositions are more efficient and reliable in the oil fingerprint identification. Acknowledgments This study was supported by The National Key Technologies R & D Program (2015BAD17B05) and National Marine Public Welfare Project (201305002). References Aeppli, C., Carmichael, C.A., Nelson, R.K., Lemkau, K.L., Graham, W.M., Redmond, M.C., et al., 2012. Oil weathering after the Deepwater Horizon disaster led to the formation of oxygenated residues. Environ. Sci. Technol. 46 (16), 8799–8807. Al-Areeq, N.M., Maky, A.F., 2015. Organic geochemical characteristics of crude oils and oil-source rock correlation in the Sunah oilfield, Masila Region, Eastern Yemen. Mar. Pet. Geol. 63, 17–27. Asif, M., Grice, K., Fazeelat, T., 2009. Assessment of petroleum biodegradation using stable hydrogen isotopes of individual saturated hydrocarbon and polycyclic aromatic hydrocarbon distributions in oils from the Upper Indus Basin, Pakistan. Org. Geochem. 40 (3), 301–311. Bacosa, H.P., Erdner, D.L., Liu, Z., 2015. Differentiating the roles of photooxidation and biodegradation in the weathering of Light Louisiana Sweet crude oil in surface water from the Deepwater Horizon site. Mar. Pollut. Bull. 95 (1), 265–272. Barrie, C.D., Taylor, K.W., Zumberge, J., 2016. Measurement of compound-specific carbon isotope ratios (δ13C values) via direct injection of whole crude oil samples. Rapid Commun. Mass Spectrom. 30 (7), 843–853. Bayona, J.M., Domínguez, C., Albaigés, J., 2015. Analytical developments for oil spill fingerprinting. Trends Environ. Anal. Chem. 5, 26–34. Beyer, J., Trannum, H.C., Bakke, T., Hodson, P.V., Collier, T.K., 2016. Environmental effects of the Deepwater Horizon oil spill: a review. Mod. Trends Surg. Cheng, P., Xiao, X.M., Gai, H.F., Li, T.F., Zhang, Y.Z., Huang, B.J., Wilkins, R.W.T., 2015. Characteristics and origin of carbon isotopes of n-alkanes in crude oils from the western Pearl River Mouth Basin, South China sea. Mar. Pet. Geol. 67, 217–229. Ezra, S., Feinstein, S., Pelly, I., Bauman, D., Miloslavsky, I., 2000. Weathering of fuel oil spill on the east Mediterranean coast, Ashdod, Israel. Org. Geochem. 31 (12), 1733–1741. Fernández-Varela, R., Andrade, J.M., Muniategui, S., Prada, D., Ramírez-Villalobos, F., 2009. The comparison of two heavy fuel oils in composition and weathering pattern, based on IR, GC-FID and GC–MS analyses: application to the prestige wreackage. Water Res. 43 (4), 1015–1026. Hall, P.A., McKirdy, D.M., Grice, K., Edwards, D.S., 2014. Australasian asphaltite strandings: their origin reviewed in light of the effects of weathering and biodegradation on their biomarker and isotopic profiles. Mar. Pet. Geol. 57, 572–593. Harvey, S.D., Jarman, K.H., Moran, J.J., Sorensen, C.M., Wright, B.W., 2012. Characterization of diesel fuel by chemical separation combined with capillary gas chromatography (GC) isotope ratio mass spectrometry (IRMS). Talanta 99, 262–269. He, S., Wang, C., Li, Y., Yu, H., Han, B., 2016. Evaluation of diagnostic ratios of medium and serious weathered oils from five different oil sources. Acta Oceanol. Sin. 35 (4), 1–8. Ho, S.J., Wang, C.Y., Luo, Y.M., 2015. GC–MS analysis of two types of mixed oils, a comparison of composition and weathering patterns. Mar. Pollut. Bull. 96 (1), 271–278.
7
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Effects of short-term weathering on the stable carbon isotope compositions of crude oils and fuel oils Yu Liua,⁎, Jixiang Xua,b, Wenjing Chena, Ying Lic a b c
College of Environmental Science and Engineering, Dalian Maritime University, Dalian 116026, China College of Environmental Science and Engineering, Maritime Safety Administration of the People's Republic of China, Beijing 100736, China College of Navigation, Dalian Maritime University, Dalian 116026, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Oil spills n-Alkanes Diagnostic ratios Stable carbon isotope Short-term weathering
A short-term simulated weathering experiment was performed on two crude oils and two heavy fuel oils under natural conditions to evaluate the effects of natural weathering processes by using gas chromatography–mass spectrometry combined with gas chromatography–isotopic ratio mass spectrometry. The results of diagnostic ratios of n-alkanes show that only odd to even predominance (OEP1, OPE2) and carbon preference index (CPI) remain stabilized during the 28 d weathering process, but they cannot effectively distinguish the four types of oils. Statistical analyses based on paired sample t-test and principal component analysis (PCA) revealed that stable carbon isotope compositions of n-alkanes in the four studied oils have no significant changes over the weathering time, and that the carbon isotope discrimination (Δδ13C) of n-alkanes is < 3‰. We have provided evidence that the stable carbon isotope compositions of n-alkanes compared to n-alkanes diagnostic ratios significantly improve the efficiency and fidelity of the oil fingerprint identification.
1. Introduction Statistical data for the past few years show that the average number of large oil-spill accidents (> 700 tons) resulting from oil exploration, shipping transportation, collisions, and unexpected events have gradually reduced, averaging at 1.8 large oil spills per year since 2010. However, the total recorded amount of oil spilled in 2015 has been > 7000 tons (ITOPF, 2015). Oil spills have physical effects and chemical toxicity, causing serious damage to marine and terrestrial ecosystems, human health, and natural resources (Beyer et al., 2016). Oil spilled into a marine environment is subjected to weathering processes, which strongly deplete the concentration of petroleum hydrocarbons and change the oil's chemical component distributions (Ezra et al., 2000; Yim et al., 2011; Samuels et al., 2013). For example, moderately biodegraded oils show an unresolved complex mixture (UCM), loss of short-chain n-alkanes (n < C15), and moderate change in alkyl naphthalene distributions (Asif et al., 2009). Prince et al. (2002) found that the maximal total extent of weathering loss of hydrocarbon from an Arctic oil spill by biodegradation can reach 87%; in this process, approximately half of the four-ring chrysene series compounds were found to be heavily degraded by photooxidation over 20 years. Gas chromatography–flame ionization detection GC–FID and gas chromatography–mass spectrometry (GC–MS) are still the most widely used
⁎
analytical techniques for oil spill fingerprinting. In these techniques, GC–FID and GC–MS profiles, source-specific target analytes, and diagnostic ratios of selected biomarkers are determined to trace spill sources (Fernández-Varela et al., 2009; Wang, Hu, He, Liu, & Zhao, 2013; Retnam et al., 2015). However, conventional chromatography cannot identify and quantify many compounds in UCMs below the baseline because of resolution and sensitivity limitations. These contain a large number of co-eluting compounds at low concentrations and overlap with the high-concentration chromatographic peaks, increasing the analytical uncertainty, especially in analyses of highly weathered oils, in which the concentration of petroleum hydrocarbons is lower than that in the original samples (Bayona et al., 2015; Wang, Chen, Zhang, He, & Zhao, 2013). Compound-specific isotope analysis (CSIA, a highly specialized and complementary analytical technology (O'Malley et al., 1997), has been extensively applied in environmental forensic investigations related to petroleum pollution. It has also been applied in organic geochemistry, being used in inferring source materials for hydrocarbons, as well as in analysis of depositional environments and maturity (Maioli et al., 2012; Al-Areeq and Maky, 2015; Li and Guo, 2010). For example, Harvey et al. (2012) distinguished four diesel fuels from different sources or locations by measuring the carbon and hydrogen isotope ratios of nalkanes, showing that compound-specific isotope analysis is a powerful
Corresponding author. E-mail address: [email protected] (Y. Liu).
http://dx.doi.org/10.1016/j.marpolbul.2017.04.003 Received 19 December 2016; Received in revised form 24 March 2017; Accepted 3 April 2017 0025-326X/ © 2017 Elsevier Ltd. All rights reserved.
Please cite this article as: Liu, Y., Marine Pollution Bulletin (2017), http://dx.doi.org/10.1016/j.marpolbul.2017.04.003
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
tool for distinguishing different fuel samples using δ13C and δD values. Hall et al. (2014) determined the origin of asphaltites with moderate or severe weathering obtained from the coast of southern Australia on the basis of biomarker ratios and n-alkane isotopic profiles, confirming they have the same source. Petroleum is mainly composed of aliphatic hydrocarbons and aromatic hydrocarbons, with aliphatic hydrocarbons (n-alkanes, branched alkanes, and cycloalkanes) usually occurring at the highest concentration in petroleum hydrocarbon, accounting for 50% of the oil (Tissot and Welte, 1984). However, the saturated components in oils are most vulnerable to weathering such as evaporation, photooxidation, and biodegradation, increasing the concentration of asphaltene and polar components in residual oil (Prince et al., 2003; Aeppli et al., 2012). In general, the loss of whole oil is mainly controlled by saturated hydrocarbons degradation, and the loss of the fuel oils is mainly dominated by the reduction of aromatic hydrocarbons (Li and Xiong, 2009). Despite its high specificity, CSIA may be limited when the oil spills show significant fractionation of stable carbon isotopes in moderate or severe biodegradation processes (Wang, Gao, Sun, Qin, Yin, & He, 2013). Studies indicate that n-alkanes are oxidized by aerobic microorganisms to oxidation products of alkyl alcohol, alkyl acid, and CO2 via a series of transformation pathways. The 12Ce12C bond usually has bond energy lower than that of 12Ce13C and 13Ce13C bonds prior to breaking in the aerobic biodegradation process, leading to 13C-enrichment in the residual saturate fraction (Xiao et al., 2012; Sun et al., 2005; Wilkes et al., 2000). Therefore, fundamental research is essential to understanding the relationship between the weathering processes and the stable isotopic composition of n-alkanes as well as to determine whether carbon isotope ratios can be used as a weatheringresistance indicator. The effects on carbon isotopic fractionation have rarely been studied by simulation experiments. The aim of this study is combine the methods gas chromatography–mass spectrometry with gas chromatography–isotopic ratio mass spectrometry to investigate the simultaneous changes in chemical composition and carbon stable isotope ratios of oil spills during the weathering process. A short-term 28-day simulated weathering experiment under natural conditions was performed on two types of oils (crude oils and fuel oils). Weathering characteristics were also analyzed and compared. Our results may provide technical support for studies on identification and characterization weathered oil spills.
Oil samples were taken periodically on the 0th, 1st, 3rd, 7th, 14th, 21st, and 28th d. All oil samples were stored in prewashed amber bottles at 4 °C in a refrigerator before experiments. In the weathering process, the average temperature, average wind speed, and average humidity were − 1.48 ± 3.55 °C, 7.23 ± 1.87 m/s, and 58 ± 4%, respectively. 2.2. Sample preparation About 0.2 g of the oil sample was dissolved in 10 mL of hexane (HPLC grade, Tedia, Fairfield, USA) and then was centrifuged at 1409g for 10 min. The fractionation of saturates was conducted in a chromatographic column (0.47 cm i.d. × 12 cm). The column was dry-packed with 10 g of activated alumina (activated for 4 h at 200 °C, 100–200 mesh, AR) and was topped with a 1 cm thick layer of anhydrous sodium sulfate (activated for 4 h at 350 °C, AR). The column was conditioned with hexane and the eluent was discarded just prior to exposure of the sodium sulfate layer to air. About 200 μL of concentrated extract was quantitatively transferred. The aliphatic hydrocarbons were eluted with 15.0 mL of hexane. The n-alkanes were further isolated from branched hydrocarbon fractions by urea adduction (Barrie et al., 2016). 2.3. GC–MS All processed samples were analyzed on mass spectrometer (ISQ, Thermo Fisher Scientific, USA) interfaced with a gas chromatograph (Thermo Fisher Scientific, CA, USA) equipped with a DB-5MS capillary column (60 m × 0.25 mm × 0.25 μm; Agilent, Santa Clara, CA, USA). He (99.999% purity) was used as the carrier gas and introduced at a flow rate of 1.2 mL/min in constant-flow mode. The GC temperature program was started at 60 °C, which was ramped to 100 °C at 20 °C/ min, held for 2 min, then from 100 to 280 °C at 6 °C/min, and held for 35 min isothermally. Injection was done with a volume of 1.0 mL in splitless mode. For MS, the injector, interface, and ion source were held at 280 °C, 250 °C, and 200 °C, respectively. The ion source was set in electron ionization mode at 70 eV. The mass spectrometer was operated in full scan mode. Scanning was done within the range of 50–550 amu at a rate of 0.2 s/scan. n-Alkanes were identified on the basis of the retention time of a standard mixture of n-alkanes (nC10 − nC32, Sigma, USA). 2.4. Stable isotope analysis
2. Materials and methods CSIA was performed using a DELTA V stable isotope ratio mass spectrometer interfaced to a gas chromatograph (Thermo Fisher Scientific). The GC instrument parameters and the temperature program are consistent with the GC–MS analytic conditions described above. Isotopic analyses were carried out at the Environmental Information Institute, Dalian Maritime University (Dalian, China). The isotope ratio for carbon was expressed as follows:
2.1. Weathering simulation experiment A simulated weathering experiment was performed on crude oils and fuel oils (Table 1). Four different oils (each oil about 8 g) were separately added to four 2000 mL beakers with 1600 mL seawater. The seawater was pre-filtered through four layers of nylon cloth (500 mesh). The beakers were kept on an outdoor open-air platform for 28 d under natural conditions in order to approach the marine field environment.
δ13C = [ (Rsample Rstandard ) − 1] ×10 3 where R represents 13C/12C. The δ13C value is relative to that of Vienna Pee Dee belemnite. The analytical precision of δ13C was < 0.06‰.
Table 1 Data of the oils samples. Sample
Locations
Oil species
API gravitya (°)
Viscosity (50 °C, mm2/s)
IBPb (°C)
WAFc (mg/L)
KWT AM 180# 380#
Kuwait Oman China China
Crude oil Crude oil Fuel oil Fuel oil
31.8 34.6 11.3 11.3
6.965 7.987 180 380
55 60 133 213
8.96 8.12 14.35 12.62
3. Results and discussion 3.1. Effects of weathering on the diagnostic ratios of n-alkanes In this study, nine kinds of diagnostic ratios (Table 2) were selected to monitor weathering, to interpret chemical data from oil spills, and to differentiate or determine a match between the crude oils and fuel oils. Here, diagnostic ratios parameters were calculated from the chromatographic peak areas of n-alkane components; therefore, ratios only represent an approximation of the relative concentrations of each nalkane fraction. Fig. 1 and Table 2 show the variations of nine
Note: WAF was quantified by UV spectroscopy using a UV1102 (Shanghai Techcomp Instrument Co.) according to GB 17378.4-2007. a API gravity = 141.5/specific gravity (60 °F) − 131.5. b IBP: initial boiling point. c WAF: water accommodated fraction.
2
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
Table 2 Changes in diagnostic ratios of n-alkanes in the weathering process. Sample no.
Ratios
n-C16/n-C18a n-C17/n-C18b (nC19 + nC20) / (nC19 − nC22)c LMW/HMWd OEP1e OEP2f (nC19 + nC20) / (nC21 + nC22)g (nC13 + nC14) / (nC25 + nC26)h CPIi
KWT crude oils
AM crude oils
180# fuel oils
380# fuel oils
0d
28 d
0d
28 d
0d
28 d
0d
28 d
1.35 1.28 0.57 2.36 1.09 1.04 1.35 2.76 1.28
1.26 1.24 0.57 1.84 1.08 1.03 1.34 1.99 1.22
1.17 1.14 0.56 0.38 1.00 1.04 1.27 2.84 1.27
0.96 1.05 0.53 1.54 1.02 1.03 1.14 0.98 1.21
1.09 1.11 0.64 1.87 1.00 1.01 1.77 2.31 1.23
0.69 0.87 0.46 0.78 0.99 1.10 0.86 0.58 1.16
0.79 0.93 0.53 1.67 1.02 1.05 1.17 2.07 1.24
0.70 0.90 0.47 0.94 1.00 0.99 0.87 0.59 1.18
Note: Values represents the average value of the three measurement results. d LMW/HMW = LMW(n-C11 − n-C21)/HMW(n-C22 − n-C32). e OEP1 = (C17 + 6 × C19 + C21) / (4 × C18 + 4 × C20). f OEP2 = (C21 + 6 × C23 + C25) / (4 × C10 + 4 × C24). i CPI = [C13 + C15 + C17 + C19 + C21 + C23) / (C14 + C16 + C18 + C20 + C22 + C24).
all ratios for the original and weathered oils substantially decreased, except for OEP1, OEP2, and CPI ratios (Fig. 1). The relative standard deviation (RSD) was used to evaluate the stability of diagnostic indices of saturated hydrocarbon; its acceptable value is < 5% (Ho et al., 2015). As shown in Fig. 2, the OEP1, OEP2,
diagnostic ratios for the two types of oils in the weathering process. The diagnostic ratios for two crude oils (KWT and AM) except for (nC13 + nC14) / (nC25 + nC26) and LMW/HMW ratios apparently did not change during the short-term weathering process (Table 2). In contrast, two fuel oils (180# and 380#) have distinct changes in ratios;
Fig. 1. Changes in diagnostic ratios of n-alkanes from the four oil samples over weathering period. (A) KWT crude oils; (B) AM crude oils; (C) 180# fuel oils; (D) 380# fuel oils. Ratios A ~ I represent n-C16/n-C18, n-C17/n-C18, (nC19 + nC20) / (nC19 − nC22), LMW/HMW, OEP1, OEP2, (nC19 + nC20) / (nC21 + nC22), (nC13 + nC14) / (nC25 + nC26), and CPI, respectively; means are results of triplicate analyses.
3
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
(R2 values of 0.95 and 0.83, respectively). The two fuel oils with WAF values higher than those of the two crude oils display relatively larger levels of degradation of n-alkanes (Fig. 3). Studies have shown that photodegradation and biodegradation are not the main factors that influence the loss of n-alkanes in the short-term weathering process; evaporation and dissolution have significant effects on the n-alkane components (Bacosa et al., 2015; Xiao et al., 2012; He et al., 2016). API gravity show an inverse correlation with the levels of n-alkane degradation in some of the oils (Asif et al., 2009), but the two kinds of fuel oils (180# and 380#) with the same API values have clearly different degrees of weathering in this study. This difference is probably because 180# fuel oil has solubility (14.35 mg/L) higher than that of 380# fuel oil, making it more susceptible to weathering. It shows that that in addition to evaporation, dissolution is also an important factor affecting the composition of n-alkanes in short-term weathering processes (Li et al., 2009), but the effects of dissolution on short-term degradation of alkanes need further investigation. The data suggest that the OEP1, OEP2, and CPI ratios showed little change, but the other selected ratios of n-alkanes were remarkably affected by weathering. The RSD of LMW/HMW and (nC13 + nC14) / (nC25 + nC26) are > 5% in this study, showing that they cannot be used in oil source identification of weathered oil spills. Furthermore, OEP1, OEP2, and CPI ratios do not differ significantly between the four types of oils (P > 0.05), which cannot be effectively distinguish by diagnostic ratios. Hence, the consistency of these ratios in various oils should be considered when they are used to distinguish different kinds of oil (Liu et al., 2015).
Fig. 2. RSD (%) of diagnostic ratios for n-alkanes in different oils.
and CPI ratios of both fuel oils and crude oils have low RSDs (< 5%), which indicate that they are not significantly affected by weathering. In contrast, LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios of both oils decreased markedly within 0–30 d, and their relative standard deviations are > 9.69% (Fig. 2). This finding shows that weathering has an impact on LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios; therefore, they should be used cautiously in the identification of spilled oils subjected to moderate (30 d) or serious (> 30 d) weathering in order to avoid misjudgment. Moreover, the changes in LMW/ HMW and (nC13 + nC14) / (nC25 + nC26) ratios mainly occurred in the first seven days of weathering, with the changes progressively slowing down (Fig. 1). This result agrees with the conclusions of previous studies that weathering mainly occurs in the first few days of the oil spill (Lemkau et al., 2010). In the present study, the diagnostic indices (RSDs) of n-C16/n-C18, n-C17/n-C18, (nC19 + nC20) / (nC1 ~ nC22), and (nC19 + nC20) / (nC21 + nC22) from crude oils are < 5%; in contrast, those of the fuel oils are > 5%. Since the RSDs of these indices are distinctly higher than the former, the fuel oils are more susceptible to weathering (Fig. 2). Clearly, oils of the same type have similar weathering patterns, indicating that changes in diagnostic ratios may be correlated with the types of oils (Figs. 1 and 2). We found a positive correlation between the degree of weathering for the n-alkanes and the water accommodated fraction WAF values in this study, as shown in Fig. 3. A plot of WAF values versus the RSD of LMW/HMW and (nC13 + nC14) / (nC25 + nC26) ratios show a strong linear correlation
3.2. Effects of weathering on the carbon isotopic composition of the nalkanes The changes in stable carbon isotopic compositions of individual nalkanes from crude oils and fuel oils in the weathering simulation process are shown in Fig. 4. The four kinds of oils have different stable carbon isotopic compositions, including carbon stable isotope ratios (δ13C values) and carbon isotope curves of n-alkanes. The 180# and 380# fuel oils have similar stable carbon isotopic compositions (Fig. 4C and D). They have relatively flat carbon isotope profiles with increasing n-alkanes carbon number, and their distribution ranges of δ13C values are − 28‰ to −24‰ and − 29‰ to − 25‰, respectively. Oil-AM has relatively low δ13C values, ranging from − 35‰ to 28‰ almost constantly or becoming slightly larger with increasing carbon number (Fig. 4C). Fig. 4A clearly shows that the carbon isotope ratios of nalkanes in KWT crude oils are widely distributed, with δ13C values ranging from −32‰ to −21‰. δ13C values first decrease with
Fig. 3. Relationship between WAF and RSD of diagnostic ratios of n-alkanes. (A) WAF value versus the RSD of LMW/HMW; (B) WAF value versus the RSD of (nC13 + nC14) / (nC25 + nC26).
4
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
-20
(A)
(B)
-26
-22
-28 -24
-30 -26
-32 KWT-0d KWT-1d KWT-3d KWT-7d KWT-14d KWT-21d KWT-28d
-30 -32 -34
δ13
δ13
-28 AM-0d AM-1d AM-3d AM-7d AM-14d AM-21d AM-28d
-34 -36 -38
-40 C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
n-Alkanes -20
-20
(C)
(D)
-22
-22
-24
-24
-26
-26 -28 180#-0d 180#-1d 180#-3d 180#-7d 180#-14d 180#-21d 180#-28d
-30 -32 -34
δ13
δ13
-28
380#-0d 380#-1d 380#-3d 380#-7d 380#-14d 380#-21d 380#-28d
-30 -32 -34
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
C11 C13 C15 C17 C19 C21 C23 C25 C27 C29 C31
n-Alkanes
Fig. 4. Carbon isotopic profiles of n-alkanes from the four oil samples and the corresponding weathering processes for the 1 d, 3 d, 7 d, 14 d, and 28 d samples (values represent mean values, n = 3). (A) KWT crude oil and its weathered oils; (B) AM crude oil and its weathered oils; (C) 180# fuel oil and its weathered oils; and (D) 380# fuel oil and its weathered oils.
increasing carbon number and then increase after the C19 alkane, resulting in a v shape in the profiles. Visually, KWT crude oil is unique among the samples; its carbon isotopic profile of n-alkanes shows a v shape with respect to carbon number. It can be distinguished from other oils on the basis of carbon isotope profiles of individual n-alkanes (Cheng et al., 2015). Oils from the same source have similar stable isotope curves. Different sources of oils have different curve shapes; thus, the rules can be used for oil–oil correlation and oil–source correlation (Murray et al., 1994; Odden et al., 2002). Results indicate that the four oil samples did not show a significant carbon isotopic fractionation with weathering. The carbon isotope discrimination Δδ13C (Δδ13C = δ13C(t) − δ13C(0)) of individual n-alkanes are < 3‰ (Fig. 4). The carbon isotope curves for the individual n-alkanes from both fuel oils and crude oils also did not extensively change with weathering process. This conclusion is in agreement with results of the three independent weathering experiments conducted by Mansuy et al. (1997), which revealed that the δ13C values of n-alkanes are minimally affected by evaporation, water washing, or biodegradation.
μ=− χ±
ts n
where μ is the population mean, where − x is the mean value of multiple parallel analysis results, s is the standard deviation of multiple parallel analysis results, n is the number of parallel analyses, and t is the t value that can be determined from a t critical value table based on the confidence limit (0.05 for a 95% confidence interval) and the degrees of freedom (n − 1), Results of the t-test of the non-weathered oil samples (KWT, AM, 180#, and 380#) and their corresponding oils weathered for 28 d are shown in Fig. 5. Most of the points are mainly distributed on a straight line (y = x); meanwhile, the error bars for all points of the original oil and the corresponding weathered oil cross the y = x straight line (CL = 95%) for each oil, indicating that the four kinds of oils and their weathered oils can be considered as the same oil sample or from the same source (Sun et al., 2009). Results of the t-test suggest that weathering and non-weathered oils can still completely match after moderate weathering, proving that stable carbon isotopes ratios can be used in the identification of moderately weathered oils. Overall, the δ13C values and the carbon isotope curves for the individual n-alkanes from the oils do not significantly change in the short-term weathering process.
3.3. t-Test analysis In order to further examine the weather resistance and applicability to oil spill source identification of the δ13C values of individual nalkanes (nC11 − nC32), a Student's t-test and scatter plot were used to analyze and compare the correlation between the original oil and 28 d weathered samples at 95% confidence limit (CL = 95%). The equation is expressed as
3.4. PCA Principal component analysis (PCA) is the most widely used and powerful statistical analysis technique in oil fingerprint identification (Wang and Fingas, 2003). In order to further study the correlation between the four kinds of oils and their weathered oils, PCA based on 5
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al. -20
-28
KWT crude oils —— y=x
-29
C13 C14 C12
C31 C16 C32
-26 C25
C28 C27 C26 C29 C17
C30
-28 C18
-30
-32 -32
C29
-31
C27 C12 C26 C13 C28 C25 C14 C11 C15
-32
C24 C21 C22 C16 C23 C18 C17
-33
-26
-24
-22
C20
-34
(A) -28
(B) -35 -35
-20
-34
-33
Weathered oil for 28 days -24
-24 -25
-28
-30
-29
-28
380# fuel oils —— y=x
C15 C16 C14 C17 C13 C28 C18 C11 C19C25 C24 C26 C21 C27 C22 C20 C12 C23
-27 -28 C29 C32 C31
-29 C30
-29
-30 -30
-31
-26
C31 C32 C15 C12C13 C16 C11 C17 C14 C19C30 C20 C26 C18 C22 C25 C29 C21 C23 C28 C27 C24
Not-weathered oil
Not-weathered oil
-25
-27
-32
Weathered oil for 28 days
180# fuel oils —— y=x
-26
C30
C31C32
C19
C24 C23C20 C22 C21 C19
-30
AM crude oils —— y=x
-30
C11 C15
-24
Not-weathered oil
Not-weathered oil
-22
-30
(C) -29
-28
-27
-26
-25
(D) -24
Weathered oil for 28 days
-31 -31
-30
-29
-28
-27
-26
-25
-24
Weathered oil for 28 days
Fig. 5. Correlation of the δ13C values of n-alkanes (nC11 − nC32) between non-weathered oil and its weathered counterpart for 28 d at 95% confidence levels: KWT samples (A), AM (B), ts , where n = 3, t = 4.303, and s is the standard deviation. 180# (C), and 380# (D). Error bars are calculated by n
corner of the double plot, with relatively heavy isotope values. Studies have indicated that the strong anti-weathered indicators OEP1, OEP2, and CPI (RSD < 5%) show no difference between 180# and 380# (Table 2), which cannot be effectively differentiated by diagnostic indicators. However, we can effectively distinguish between 180# and 380# by stable isotope composition through the cross-plot diagram, as shown in Fig. 6. The carbon isotope values of long- and short-chain nalkanes of 180# are heavier than those of 380#, while the carbon isotope ratio values of middle-carbon-number n-alkanes from 180# are lower than those of 380#. Oils and their corresponding weathered samples at different weathering periods cluster together closely, further proving that there is little isotopic fractionation during short-term weathering. Analytical data from PCA imply that the two crude oils and two heavy fuel oils, which resemblance each other in appearance, can be distinguished in the scattered plots. In conclusion, we have provided evidence that the source apportionment of spilled oils after weathering can be advanced by statistical analysis combined with CSIA.
2
1
PC2
0
-1
KWT AM 180# 380#
-2
-3 -8
-6
-4
-2
0
2
4
PC1 Fig. 6. PCA results based on the δ13C values of n-alkanes (nC11 − nC32) in non-weathered and weathered oils.
the δ13C of alkanes (nC11 − nC32) was conducted by using SPSS 19.0. The PCA results are shown in Fig. 6. The first and second principal components respectively explain 77.6% and 15.3% of the variance, and the accumulated contribution ratio of the two principal components is 92.9%. Factor analysis showed that the low carbon number (n < C15) and high carbon number of n-alkanes (n > C25) in the first principal component have larger loadings, and the middle carbon number nalkanes have heavy loading in the second principal component. Fig. 3 clearly indicates that oil-AM and oil-KWT are the endpoints for the four group oils. The fuel oil samples 180# and 380# lie in the upper right
4. Conclusions The weathering simulation experiment revealed that the LMW/ HMW and (nC13 + nC14) / (nC25 + nC26) ratios are markedly dependent on weathering(RSD) > 5%, indicating that these may not be valid indices for oil source identification of the spill oil after degradation. The OEP1, OEP2, and CPI ratios change little, but they cannot effectively distinguish the four types of oil samples. High WAF values are consistent with relatively high levels of degradation of n-alkanes in the four oils. Paired sample t-test results show that the four oil samples 6
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Y. Liu et al.
ITOPF (International Tanker Owners Pollution Federation), 2015. Oil Tanker Spill Statistics. (Available from: < http://www.itopf.com/knowledge-resources/datastatistics/statistics/ > . Accessed 26 December 2015). Lemkau, K.L., Peacock, E.E., Nelson, R.K., Ventura, G.T., Kovecses, J.L., Reddy, C.M., 2010. The M/V Cosco Busan spill: source identification and short-term fate. Mar. Pollut. Bull. 60 (11), 2123–2129. Li, S.M., Guo, D., 2010. Characteristics and application of compound specific isotope in oil-source identification for oils in Dongying Depression, Bohai Bay Basin. Geoscience 24, 252–258. Li, Y., Xiong, Y., 2009. Identification and quantification of mixed sources of oil spills based on distributions and isotope profiles of long-chain n-alkanes. Mar. Pollut. Bull. 58 (12), 1868–1873. Li, Y., Xiong, Y., Yang, W., Xie, Y., Li, S., Sun, Y., 2009. Compound-specific stable carbon isotopic composition of petroleum hydrocarbons as a tool for tracing the source of oil spills. Mar. Pollut. Bull. 58 (1), 114–117. Liu, X., Ma, Q., Yang, X., 2015. Study on consistency of diagnostic ratios of different oils. Aquatic Procedia 3, 231–237. Maioli, O.L.G., de Oliveira, C.R., Dal Sasso, M.A., dos Santos Madureira, L.A., de Almeida Azevedo, D., de Aquino Neto, F.R., 2012. Evaluation of the organic matter sources using the δ13C composition of individual n-alkanes in sediments from Brazilian estuarine systems by GC/C/IRMS. Estuar. Coast. Shelf Sci. 114, 140–147. Mansuy, L., Philp, R.P., Allen, J., 1997. Source identification of oil spills based on the isotopic composition of individual components in weathered oil samples. Environ. Sci. Technol. 31 (12), 3417–3425. Murray, A.P., Summons, R.E., Boreham, C.J., Dowling, L.M., 1994. Biomarker and nalkane isotope profiles for Tertiary oils: relationship to source rock depositional setting. Org. Geochem. 22 (3–5), 521IN5–542IN6. Odden, W., Barth, T., Talbot, M.R., 2002. Compound-specific carbon isotope analysis of natural and artificially generated hydrocarbons in source rocks and petroleum fluids from offshore Mid-Norway. Org. Geochem. 33 (1), 47–65. O'Malley, V.P., Burke, R.A., Schlotzhauer, W.S., 1997. Using GC–MS/combustion/IRMS to determine the 13C/12 C ratios of individual hydrocarbons produced from the combustion of biomass materials—application to biomass burning. Org. Geochem. 27 (7), 567–581. Prince, R.C., Owens, E.H., Sergy, G.A., 2002. Weathering of an Arctic oil spill over 20 years: the BIOS experiment revisited. Mar. Pollut. Bull. 44 (11), 1236–1242. Prince, R.C., Garrett, R.M., Bare, R.E., Grossman, M.J., Townsend, T., Suflita, J.M., et al., 2003. The roles of photooxidation and biodegradation in long-term weathering of crude and heavy fuel oils. Spill Sci. Technol. Retnam, A., Kassim, A.M., Ahmad, W.K.W., 2015. Fingerprinting of light fuel oil: a Malaysia case study. Prog. Environ. Sci. 30, 190–194. Samuels, W.B., Amstutz, D.E., Bahadur, R., Ziemniak, C., 2013. Development of a global oil spill modeling system. Earth Sci. Res. 2 (2), 52. Sun, Y., Chen, Z., Xu, S., Cai, P., 2005. Stable carbon and hydrogen isotopic fractionation of individual n-alkanes accompanying biodegradation: evidence from a group of progressively biodegraded oils. Org. Geochem. 36 (2), 225–238. Sun, P., Bao, M., Li, G., Wang, X., Zhao, Y., Zhou, Q., Cao, L., 2009. Fingerprinting and source identification of an oil spill in China Bohai Sea by gas chromatography-flame ionization detection and gas chromatography–mass spectrometry coupled with multistatistical analyses. J. Chromatogr. A 1216 (5), 830–836. Tissot, B.P., Welte, D.H., 1984. Petroleum Formation and Occurrence . Wang, Z., Fingas, M.F., 2003. Development of oil hydrocarbon fingerprinting and identification techniques. Mar. Pollut. Bull. 47 (9), 423–452. Wang, C., Gao, X., Sun, Z., Qin, Z., Yin, X., He, S., 2013a. Evaluation of the diagnostic ratios for the identification of spilled oils after biodegradation. Environ. Earth Sci. 68 (4), 917–926. Wang, C., Chen, B., Zhang, B., He, S., Zhao, M., 2013b. Fingerprint and weathering characteristics of crude oils after Dalian oil spill, China. Mar. Pollut. Bull. 71 (1), 64–68. Wang, C., Hu, X., He, S., Liu, X., Zhao, M., 2013c. Source diagnostic and weathering indicators of oil spills utilizing bicyclic sesquiterpanes. Acta Oceanol. Sin. 32 (8), 79–84. Wilkes, H., Boreham, C., Harms, G., Zengler, K., Rabus, R., 2000. Anaerobic degradation and carbon isotopic fractionation of alkylbenzenes in crude oil by sulphate-reducing bacteria. Org. Geochem. 31 (1), 101–115. Xiao, Q., Sun, Y., Zhang, Y., Chai, P., 2012. Stable carbon isotope fractionation of individual light hydrocarbons in the C 6–C 8 range in crude oil as induced by natural evaporation: experimental results and geological implications. Org. Geochem. 50, 44–56. Yim, U.H., Ha, S.Y., An, J.G., et al., 2011. Fingerprint and weathering characteristics of stranded oils after the Hebei Spirit oil spill. J. Hazard. Mater. 197, 60–69.
for 28d also can be well matched to its original oil samples, indicating that the stable carbon isotopic compositions did not exhibit considerable isotope fractionation in the short-term weathering process. PCA based on the δ13C of alkanes (nC11 − nC32) showed that the four oil samples at different weathering periods with corresponding original oils can cluster together in the scatter plots respectively, which cannot be distinguished by chemical composition. This demonstrated that statistical analysis combined with carbon isotope fingerprints can markedly increase the ability of oil spill characterization. In conclusion, the chemical compositions of n-alkanes are more likely to be affected than are the isotopic compositions of n-alkanes by the weathering process, and the stable isotopic compositions relative to chemical compositions are more efficient and reliable in the oil fingerprint identification. Acknowledgments This study was supported by The National Key Technologies R & D Program (2015BAD17B05) and National Marine Public Welfare Project (201305002). References Aeppli, C., Carmichael, C.A., Nelson, R.K., Lemkau, K.L., Graham, W.M., Redmond, M.C., et al., 2012. Oil weathering after the Deepwater Horizon disaster led to the formation of oxygenated residues. Environ. Sci. Technol. 46 (16), 8799–8807. Al-Areeq, N.M., Maky, A.F., 2015. Organic geochemical characteristics of crude oils and oil-source rock correlation in the Sunah oilfield, Masila Region, Eastern Yemen. Mar. Pet. Geol. 63, 17–27. Asif, M., Grice, K., Fazeelat, T., 2009. Assessment of petroleum biodegradation using stable hydrogen isotopes of individual saturated hydrocarbon and polycyclic aromatic hydrocarbon distributions in oils from the Upper Indus Basin, Pakistan. Org. Geochem. 40 (3), 301–311. Bacosa, H.P., Erdner, D.L., Liu, Z., 2015. Differentiating the roles of photooxidation and biodegradation in the weathering of Light Louisiana Sweet crude oil in surface water from the Deepwater Horizon site. Mar. Pollut. Bull. 95 (1), 265–272. Barrie, C.D., Taylor, K.W., Zumberge, J., 2016. Measurement of compound-specific carbon isotope ratios (δ13C values) via direct injection of whole crude oil samples. Rapid Commun. Mass Spectrom. 30 (7), 843–853. Bayona, J.M., Domínguez, C., Albaigés, J., 2015. Analytical developments for oil spill fingerprinting. Trends Environ. Anal. Chem. 5, 26–34. Beyer, J., Trannum, H.C., Bakke, T., Hodson, P.V., Collier, T.K., 2016. Environmental effects of the Deepwater Horizon oil spill: a review. Mod. Trends Surg. Cheng, P., Xiao, X.M., Gai, H.F., Li, T.F., Zhang, Y.Z., Huang, B.J., Wilkins, R.W.T., 2015. Characteristics and origin of carbon isotopes of n-alkanes in crude oils from the western Pearl River Mouth Basin, South China sea. Mar. Pet. Geol. 67, 217–229. Ezra, S., Feinstein, S., Pelly, I., Bauman, D., Miloslavsky, I., 2000. Weathering of fuel oil spill on the east Mediterranean coast, Ashdod, Israel. Org. Geochem. 31 (12), 1733–1741. Fernández-Varela, R., Andrade, J.M., Muniategui, S., Prada, D., Ramírez-Villalobos, F., 2009. The comparison of two heavy fuel oils in composition and weathering pattern, based on IR, GC-FID and GC–MS analyses: application to the prestige wreackage. Water Res. 43 (4), 1015–1026. Hall, P.A., McKirdy, D.M., Grice, K., Edwards, D.S., 2014. Australasian asphaltite strandings: their origin reviewed in light of the effects of weathering and biodegradation on their biomarker and isotopic profiles. Mar. Pet. Geol. 57, 572–593. Harvey, S.D., Jarman, K.H., Moran, J.J., Sorensen, C.M., Wright, B.W., 2012. Characterization of diesel fuel by chemical separation combined with capillary gas chromatography (GC) isotope ratio mass spectrometry (IRMS). Talanta 99, 262–269. He, S., Wang, C., Li, Y., Yu, H., Han, B., 2016. Evaluation of diagnostic ratios of medium and serious weathered oils from five different oil sources. Acta Oceanol. Sin. 35 (4), 1–8. Ho, S.J., Wang, C.Y., Luo, Y.M., 2015. GC–MS analysis of two types of mixed oils, a comparison of composition and weathering patterns. Mar. Pollut. Bull. 96 (1), 271–278.
7