DIFFERENTIATION OF CRUDE OILS, FUEL OILS, AND USED LUBRICATING OIL USING DIAGNOSTIC RATIOS Pornpetch Hattakijvilai a, Siriporn Jongpatiwut*,a,b a The Petroleum and Petrochemical College, Chulalongkorn University, Bangkok, Thailand b Center of Excellence on Petrochemical and Materials Technology, Bangkok, Thailand Keywords: Chemical fingerprinting, PAHs, Biomarker, Crude oil, Fuel oil, Used lube oil ABSTRACT Oil spills in marine are caused by several sources i.e. crude oil, fuel oil and used lube oil (ULO) by accidentally happened according to a leakage or maybe, in some cases, purposefully discharged to the sea. Normally, fuel oil is produced from the heavy fraction of crude oil which gives similar properties. Consequently, it cannot be concluded the source of the spill wherever its origin from crude oil or fuel oil. In this work, crude oils, fuel oils and used lube oil were characterized using GC-FID and GCxGC TOFMS. The statistical method will be applied to differentiate crude oils, fuel oils and ULO that are typically used in Thailand. The results from double ratio plots of alkyl-anthracene and alkylphenanthrenes showed a positive method to distinguish crude oils and those two refined products. *siriporn.j@chula.ac.th INTRODUCTION As different geological conditions in different regions such as pressure, temperature and fossil, they originated unique chemical compounds in oils which can be beneficially used to identify source of oil spills. Oil spills in marine can be probably happened by accidents like a leakage and unauthorized discharges into the sea of crude oils, fuel oils which is used as a ship fuel and used lube oil (ULO). Fuel oil is refined from the heavy fraction of crude oil which gives the similar properties to crude oil. Therefore, it might not be concluded the source of the spill wherever its origin from crude oil or fuel oil. Polycyclic aromatic hydrocarbons or PAHs are aromatic compounds containing from two to eight benzene rings condensed together. PAHs is one of an important hydrocarbon chemical fingerprinting for identification of oils as the different sources of crude oils and fuel oils give the different PAHs distribution (Zhang et al., 2016). Biomarkers or biological markers are complex hydrocarbon molecules derived from living organisms. Biomarkers are high degradationresistant in the environment as for example; pristane, phytane, steranes, triterpanes and porphyrins (Moustafa et al., 2012). The analysis of PAHs and biomarkers can be done following the NORDTEST oil spill identification (Liv-Guri Faksness et al., 2002), using the statistical theory. In this work, crude oils, fuel oils and used lube oil were characterized using GC-FID and GCxGC TOFMS. The statistical method was applied to differentiate crude oils, fuel oils and ULO that are typically used in Thailand. In addition, the ratios of alkylated PAH isomers in the same alkylation level were also be studied. Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 1
EXPERIMENTAL A. Materials Five crude oil samples used in this study are named as CO1-CO5, two different processed fuel oils were labeled as FO1 and FO2 and used marine lube oil was designated as ULO. Hexane and dichloromethane with high purity were purchased from RCI Labscan (Thailand) and Burdick Jackson (South Korea), respectively. Sodium carbonate was obtained from Asia Pacific Specialty Chemicals. Silica gel and anhydrous sodium sulfate were purchased from Merck (Germany). B. Sample preparation The extraction of oil samples method was combined from two published articles which are Mulabagal et al. (2013) and Zhang et al. (2016). Approximately 0.15 g of oil samples were placed into clear vials. Then these oil samples were extracted with 10 ml of high purity n- hexane/dichloromethane (1:1,v/v) and 0.5 g of sodium carbonate were added to remove water. The mixture was vortexed for 5 min and the solutions were allowed to settle at room temperature for 4 h. After that, the supernatants in each vial was filtered through 0.45 μm filter and transferred to a vial containing 0.5 g of clean up mixture consisting 0.25 g of silica gel and 0.25 g of anhydrous sodium sulfate. Then, the mixture was vortexed for 2 min and allowed to settle at room temperature for 2 min. Finally, the final sample was filtered through 0.2 µm filter. C. Sample analysis The prepared samples were analyzed for their hydrocarbon fraction and normalized with C25 using a Varian CP-3800 stimulated distillation gas chromatography (SIMDIST-GC), equipped with a flame-ionization detector (FID) and HP-5 column (30 m 0.25 mm ID 0.25 µm), and the oven temperature was started at 40 ᵒC (5 min) then increased 6 ᵒC/min to 310 ᵒC (10 min). The qualitative analysis of PAHs and biomarker were constructed in selected ion monitoring mode (SIM), using comprehensive two-dimension gas chromatography (Agilent Technologies 7890) with time-of-flight mass spectrometer (LECO, Pegasus 4D TOF/MS) (GCxGC-TOF/MS) equipped with an Rxi -PAH MS (60 m 0.25 mm ID 0.1 µm) as the first column, and a non-polar Rxi -1HT (1 m 0.25 mm ID 0.1 µm) as the second column. The target ion of m/z 178,192, 206, 220, 184, 198, 212, 226 were monitored for PAHs and alkylated isomers and m/z 191 was used for hopane. D. Data analysis Data were acquired and processed using Leco ChromaTOF software. PAHs and hopanes group were set up which were resampled with a mass range 45 to 550 atomic mass units with an acquisition rate of 200 spectra/sec. Diagnostic ratios were calculated using this equation: Ratio = A/(A+B), where A and B were (area of analyte)/(area of internal standard). Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 2
RESULTS AND DISCUSSION A. Differentiation of crude oils and processed oils by aliphatic hydrocarbons GC-FID results provided basic information about hydrocarbons distribution in oils. Crude oils and fuel oils showed similar pattern on the chromatogram, but in fuel oils had less hydrocarbon in ranges of C8-C10 as illustrated in Figure 1a and 1b. Used lubricating oil could be distinguished from other oil types by the unresolved mixture complex or UCM as shown in Figure 1c. Light hydrocarbon carbons in used lube oil were contaminated by unburnt fuel in the engine. (a) (b) (c) Figure 1 GC-FID chromatogram of (a) CO1 (b) FO2 (c) ULO. B. Analysis of crude oils and processed oils by PAHs From GCxGC-TOFMS results, the distribution patterns of methyl-anthracene (MA) and methyl-phenanthrene (MP) at m/z 192 were different in most crude oils and fuel oil. The alkyl-phenanthrenes, 3-, 2-, 9- and 1-MP, were detected in all oil samples while 2-MA and 4-MP were appeared in only CO2, FO1 and FO2. Besides, both fuel oils, FO1 and FO2, had a larger of the first two peak of 3- and 2-MP than the last second peak of 9- and 1- MP. While in the most crude oils (CO1, CO3, CO4, and CO5), the first double peak of 3- and 2- Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 3
MP was smaller than the second one (Eurocrude, 1995, Liv-Guri Faksness et al., 2002). For calculating diagnostic ratios, it could be claimed that most of crude oils (expect CO2) had 4- MD/1-MD ratio less than 0.75 while fuel oil and CO2 were resulted more than 0.75. However used lube oil could not be detected 1-MD, the ratio of 4-MD/1-MD became equal one as illustrated in Table 1. The trend of 2-MP/1-MP ratio were below 0.5 in almost crude oils and above 0.5 in CO2 and processed oils. Table 1 Diagnostic ratio values of 4-MD/1-MD and 2-MP/1-MP in all samples Ratios CO1 CO2 CO3 CO4 CO5 FO1 FO2 ULO 4MD/1MD 0.72 0.94 0.58 0.61 0.63 0.83 0.77 1.00 2MP/1MP 0.47 0.59 0.46 0.46 0.48 0.73 0.62 0.89 The cross plots of C2D/C2P vs C3D/C3P in all samples are illustrated in Figure 2, the result showed a good trend to distinguish crude oils and refined products. The ratios of C2D/C2P and C3D/C3P that higher than 0.7 was belonged to most crude oils, excepted CO2 which contains lowest sulfur percentage, had these ratios below 0.15. While the ratios observed on fuel oils and ULO were lower than 0.4 and 0.15, respectively. Figure 3 presents the cross plot of 2-MA/2-MP vs 2-MA/3-MP, 2-MA/1-MP vs 2-MA/9-MP, 2- MA/(3+2)-MP vs 2-MA/(9+1)-MP, (3+2)-MP/(9+1)-MP vs 2-MA/ MP. The results of these cross plots showed that most of crude oils (excluding CO2) had no 2-MA therefore, the ratios were equal zero excluding the ratio of (3+2)- MP/(9+1)-MP had value lower than 0.4. Also with in ULO there was no 2-MA as same as in the most crude oils, but the ratio of (3+2)-MP/(9+1)-MP was above 0.6. Comparing CO2 and fuel oils, the ratios of CO2 were below 0.15 in 2-MA/2-MP and 2-MA/3-MP and lower than 0.2 in 2-MA/1-MP and 2- MA/9-MP while they were above 0.15 and 0.2 in both fuel oils, respectively. (a) CO1 (d) CO4 (f) FO1 (b) CO2 (e) CO5 3-MP 2-MP 9-MP 1-MP (g) FO2 2-MA 4-MP (c) CO3 (h) ULO m/z 192 Figure 2 The distribution pattern of methyl-phenanthrene (MP) and methyl-anthracene (MA) with target ion 192 of all samples. Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 4
Figure 2 The cross plot of C2D/C2P vs C3D/C3P of all samples. (a) (b) (c) (d) Figure 3 The cross plot of (a) 2-MA/2-MP vs 2-MA/3-MP (b) 2-MA/1-MP vs 2-MA/9-MP (c) 2-MA/(3+2)-MP vs 2-MA/(9+1)-MP (d) (3+2)-MP/(9+1)-MP vs 2-MA/ MP of all samples. C. Analysis of crude oils and processed oils by biomarkers The result from GCxGC-TOFMS with target ion 191 (m/z 191) was used to monitor hopanes. The distribution patterns of hopane were different between crude oils, fuel oils and ULO as the refining processes had an effect on hopane biomarkers, somes were Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 5
generated, somes were degraded due to the conditions of the refining unit such as vacuum distillation, hydrocracking and visbreaking unit (Russell et al., 2004). The example of correlation plots of diagnostic ratio of hopane biomarkers between each crude oil feedstock (CO1 and CO5) and FO1 using 95% confidence limit that illustrated in Figure 4 indicated non-match as several keys not overlapping the straight line x=y. Figure 4 The correlation plot of diagnostic ratios of hopanes biomarkers between fuel oil 1 and its crude oil feedstock using 95% confidence limit. CONCLUSIONS Used lube oil could be identified by the unique UCM GC-FID chromatogram. The cross plots of C2D/C2P vs C3D/C3P and (3+2)-MP/(9+1)-MP vs 2-MA/ MP could be used to differentiate crude oils and processed oils according to the boundaries of ratios in each oil types. The refining processes affected the hopane distribution pattern thus, the correlation plots between each crude oil feedstocks and their fuel oils performed as non-match. ACKNOWLEDGEMENTS The authors would like to thank the Petroleum and Petrochemical College, Chulalongkorn University, Thailand, Department of marine science, Faculty of Science, Chulalongkorn University, Center of Excellence on Petrochemical and Materials Technology, PTT Global Chemical Public Company Limited, Star Petroleum Refining Public Company Limited and Inter Marine Lube Company Limited for supports to this research work. REFERENCES Faksness, L. G., Weiss, H. M., & Daling, P. S. (2002). SINTEF Applied Chemistry. STF66 A02028. Moustafa, Y. M., & Morsi, R. E. (2012). Biomarkers. INTECH Open Access Publisher. Mulabagal, V., Yin, F., John, G. F., Hayworth, J. S., & Clement, T. P. (2013). Marine pollution bulletin. 70(1), 147-154. Zhang, H., Wang, C., Zhao R., Yin, X., Zhou, H., Tan, L., and Wang, J. (2016). Marine Pollution Bulletin. 106, 58-61. Petrochemical and Materials Technology Tuesday May 23, 2017, Pathumwan Princess Hotel, Bangkok, Thailand Page 6