Marine Pollution Bulletin 70 (2013) 147–154

Contents lists available at SciVerse ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier .com/locate /marpolbul

Chemical fingerprinting of petroleum biomarkers in Deepwater Horizon oilspill samples collected from Alabama shoreline

V. Mulabagal, F. Yin, G.F. John, J.S. Hayworth, T.P. Clement ⇑Department of Civil Engineering, Auburn University, Auburn, AL 36849, USA

a r t i c l e i n f o a b s t r a c t

Keywords:BP oil spillDeepwater Horizon oil spillHopane analysisFingerprintingTar balls

0025-326X/$ - see front matter � 2013 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.marpolbul.2013.02.026

⇑ Corresponding author. Tel.: +1 334 844 6268.E-mail address: Clement@auburn.edu (T.P. Clemen

We compare the chromatographic signatures of petroleum biomarkers in Deepwater Horizon (DH)source oil, three other reference crude oils, DH emulsified mousse that arrived on Alabama’s shorelinein June 2010, and seven tar balls collected from Alabama beaches from 2011 to 2012. Characteristichopane and sterane fingerprints show that all the tar ball samples originated from DH oil. In addition,the diagnostic ratios of various hopanes indicate an excellent match. Quantitation data for C30ab-hopaneconcentration levels show that most of the weathering observed in DH-related tar balls found onAlabama’s beaches is likely the result of natural evaporation and dissolution that occurred duringtransport across the Gulf of Mexico prior to beach deposition. Based on the physical and biomarkercharacterization data presented in this study we conclude that virtually all fragile, sticky, brownish tarballs currently found on Alabama shoreline originated from the DH oil spill.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Understanding the fate and transport of spilled crude oil in mar-ine environments is an important environmental managementproblem. One of the largest crude oil spills in the history of oilindustry recently occurred in the Gulf of Mexico (GOM) when theDeepwater Horizon (DH) drilling platform exploded and sank atthe Macando prospect exploration site (MC252). This catastrophereleased an estimated 170 million gallons of raw crude oil intoGOM waters between April 20, 2010 and July 15, 2010 (McNuttet al., 2011). From June 2010 to the present, DH-related oil, in var-ied weathered forms such as emulsions, surface slicks and tar balls,were deposited on Alabama’s beaches (Hayworth and Clement,2011; Hayworth et al., 2011). Our team has been monitoring thetar ball activity along Alabama’s beaches and other DH-oil spill re-lated contamination issues for the past 3 years (Clement et al.,2012; Hayworth and Clement, 2012). Our field observations indi-cate that the level of tar ball activity continues to be high all alongthe Alabama shoreline, especially after a storm event (Clementet al., 2012). For example, during a recent field survey completedin September 2012, immediately after Hurricane Isaac, we ob-served thousands of tar balls and tar mat fragments scattered allalong the beaches. At one of the high deposition sites located inBon Secour National Wildlife Refuge, Alabama, we collected over150 tar balls (size ranging from 1 to 4 cm diameter) within a rela-tively small 20 m2 control region (Clement et al., 2012). The contin-

ll rights reserved.

t).

uous deposition of DH-oil-spill related tar balls is a major concernfor residents living along GOM beaches, and therefore a fundamen-tal understanding of how the concentrations of various oil compo-nents are evolving with time in these tar ball samples is needed tobetter evaluate its long term impacts on GOM ecosystems. The pri-marily objective of this study is to analyze the biomarker finger-prints of various types of DH-related weathered oil spill residuesrecovered from Alabama’s beaches.

Comparing the biomarker fingerprints of tar ball samples withthose from the original source crude oil was an established charac-terization procedure used in several oil spill assessment studies(Pauzi Zakaria et al., 2001; Peters et al., 2007; Wang et al., 2006).Biomarkers such as hopanes and steranes are stable compoundspresent in crude oils; they were accumulated as the oil was formedfrom formerly living organisms whose organic materials werepreserved within source rocks over geologic times (Peters et al.,2005; Wang et al., 2007). Their resistance to weathering allowsthem to be used as conservative markers, and they are often re-ferred to as ‘‘molecular fossils’’ (Peters and Modowan, 1993). Thechemical signatures of biomarkers such as hopanes and steranescan be used for tracing the source of a particular raw crude oil,and also for quantifying the weathering rate of an oil in naturalenvironments (Wang et al., 2001).The composition of biomarkersin crude oils can vary widely, depending primarily on the sourceand geological conditions of the reservoir (Idris et al., 2008; Peterset al., 2007). Among available biomarkers, hopanes and steranesare the most important class of compounds commonly used inoil spill management efforts (Prince et al., 1994). Studies haveshown that hopanes are abundant in highly weathered oils, and

148 V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154

can be used to quantify the degree of weathering (Wang et al.,1994a). Due to variations in geologic conditions, each crude oilwould exhibit a unique set of hopane and sterane biomarker ‘‘fin-gerprints’’. Thus, these biomarkers provide a way to differentiateoils, monitor the degree of oil weathering, and quantify biodegra-dation processes under a variety of environmental conditions(Escobar et al., 2011; Wang and Fingas, 2003).

Advanced analytical instruments such as gas chromatography/flame ionization detector (GC/FID) and gas chromatography/massspectrometry (GC/MS) have been routinely used to analyze bio-markers in crude and weathered oil samples (Barakat et al.,1999; Hauser et al., 1999; Mostafa et al., 2009; Pauzi Zakariaet al., 2001; Stout, 2003; Wang et al., 1994a,b, 1999, 2001, 2004;Wang and Fingas, 2003; Yim et al., 2011). In this work, we useda GC/MS method, run in selected ion monitoring (SIM) mode, forbiomarker fingerprinting and quantification. Field samples wereprocessed using a rapid cleanup procedure and were analyzed forhopanes and steranes. An Agilent GC/MS system was used to fin-gerprint and quantify hopanes in DH source oil, emulsified mousse(which first arrived on Alabama’s beaches in June 2010), and sevendifferent tar balls and tar-mat fragments collected from Alabamabeaches between September 2011 and February 2012. In addition,steranes profiles in DH source oil and several field samples werealso fingerprinted.

2. Materials and methods

2.1. Materials

All of the organic solvents used in this study were of analyticalor higher grade and were purchased from VWR International(Suwanee, GA). Polytetrafluoroethylene membrane filters (PTFE,0.45 and 0.2 lm) were purchased from VWR International (Suwa-nee, GA). Hopane standards were purchased from Chiron, Trond-heim, Norway. Anhydrous sodium sulfate (>99.0%, granular) andsilica gel (60–200 lm) were purchased from Sigma–Aldrich (Allen-town, PA). Chem Tube-Hydromatrix was purchased from AgilentTechnologies (Wilmington, DE). Deactivated GC liners (splitless ta-pered glasswool), GC capillary columns (J&W DB-EUPAH, p/n 121-9627, 340 �C, 20 m � 180 lm � 0.14 lm; HP-5, p/n 19091 J-436,60 m � 250 lm � 0.25 lm; fused silica, p/n 160-7625, 450 �C,0.7 m � 150 lm � 0 lm) were purchased from Agilent Technolo-gies (Wilmington, DE).

2.2. Sample details

Chemical fingerprinting of hopanes in seven different tar ballsamples was completed. A sample of MC252 source oil, obtained

Fig. 1. Tar ball sampling locations along Alabama shoreline from Perdido Pass to Fort M22 miles].

from British Petroleum (BP), is designated as ‘‘DH oil’’ in this study.The emulsified form of DH oil (orange-colored mousse) collectedfrom Orange Beach, Alabama, on June 11th, 2010, is designatedas ‘‘Mousse’’. A small volume of original source crude oil(11.3 mL) was transferred into an open 4-oz jar and was evapo-rated for 6 months under fume hood (Mott Manufacturing Limited,Ontario, Canada) at a face velocity of 200 fpm. The weathering pro-cess was completed at room temperature (22 �C) and the resultingevaporated crude was designated as: ‘‘Evaporated DH oil’’ (EDH).Our research team has collected and archived several hundredtar balls and tar mat fragments over the past 3 years; details ofthese sampling trips are summarized in Clement et al. (2012).For this study, we randomly selected seven distinct tar balls col-lected between September 2011 and February 2012 over the 22-mile long study region shown in Fig. 1, and these samples are des-ignated as: TB1, TB2, TB3, TB4, TB5, TB6 and TB7. The samplinglocations of these tar balls are shown in Fig. 1, and their GPS coor-dinates and other details are summarized in Table 1. In addition tothe above samples, we also analyzed three standard referencecrude oil samples: Arabian crude (AC), Bazra crude (BC) and Vene-zuelan heavy crude (VHC) which were purchased from ONTA Inc.,Toronto, Canada.

2.3. Estimation of oil percentage in tar ball samples

To determine the amount of residual oil fraction remaining inthe tar ball samples, about 1 g of each tar ball was weighed andwas extracted with 10 mL of dichloromethane four times. The dis-solved oil fraction was decanted and the remaining solid (sand)fraction was dried and weighted. The sand fraction in these tar-ballsamples varied from 76% to 89% and these results are summarizedin Table 1.

2.4. Sample preparation procedure

About 25 mg of DH, EDH, mousse, and other crude oil samples(AC, BC, and VHC) were placed into 40 mL clear vials. To these pureoil samples we added10 mL of hexane to maintain an effective oilconcentration of 25 mg of oil phase per 10 mL of hexane. We thenadded 0.5 g of Chem-Tube-Hydromatrix and the samples were vor-texed for 5 min, and the solutions were then allowed to settle atroom temperature for 4 h. The supernatant in each vial was thenfiltered through 0.45 lm PTFE membrane filter and 4 mL of the ex-tract was transferred to a vial containing 0.5 g of cleanup mixtureconsisting of 0.25 g of silica gel 60 (60–100 lm) and 0.25 g of anhy-drous sodium sulfate. The mixture was vortexed for 2 min, allowedto settle for 2 min, and then filtered through 0.2 lm PTFEmembrane.

organ [scale: from TB1 (Orange Beach point) to TB6 (Morgan Town Blvd.) is about

Table 1Details of tar ball samples collected from various Alabama beaches.

Sample name Latitude/N Longitude/W Sample date Location details of the sampling point Sand content (%)

TB1 30�16014.980 0 87�3401.40 0 9/6/2011 Orange beach opposite to St. Thomas church 84TB2 30�14028.650 0 87�4403.280 0 9/8/2011 Lagoon Pass 83TB3 30�14012.430 0 87�45046.940 0 9/24/2011 2432 West Beach Blvd. (near Lee’s landing marker) 89TB4 30�13044.900 0 87�49041.260 0 11/12/2011 Mobile Street (Bon Secour National Wildlife Refuge) 86TB5 30�14047.760 0 87�41027.320 0 1/25/2012 Gulf Shores public beach near Hangout Point 85TB6 30�13050.380 0 87�54033.690 0 2/18/2012 Morgantown Blvd. 88TB7 30�14030.230 0 87�43041.900 0 2/19/2012 Lagoon Pass 76

V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154 149

As shown in Table 1, our initial oil extraction data indicated thatthe tar balls contained considerable amounts of sand ranging from76% to 89%. Based on these data, we extracted the tar ball sampleswith appropriate amount of hexane to maintain a concentration of25 mg weathered oil per 10 mL of solvent. The extract was thensubject to all the cleanup procedures discussed above. The finalsamples were spiked with the internal standard C30bb-hopane(IS, 17b(H),21b(H)-hopane, 100 ng/mL) prior to analysis. All sam-ples were extracted and prepared in triplicate and each samplewas analyzed in duplicate. To quantify extraction and cleanuprecoveries, field samples were spiked with a known amount ofC30ab-hopane (17a(H),21b(H)-hopane) and analyzed using GC/MS. The cleanup steps used for sterane analysis also followed asimilar procedure.

2.5. Calibration curve

The calibration curve used to quantify hopane concentrationlevels was generated using a commercially available C30ab-hopa-

Fig. 2. Comparison of hopane distribution in Deepwater Horizon (DH) oil, evaporatedtrisnorneohopane; Tm: 17a(H)-22,29,30-trisnorhopane; C29: 17a(H),21b(H)-30-norhopC32:17a(H),21b(H)-32-bishomohopane-22S/22R; C33: 17a(H),21b(H)-33-trishomohopanedard (17b(H),21b(H)-hopane).

ne(17a(H), 21b(H)) standard (purchased from Chiron). Variousdilutions ranging from 50 to 400 ng/mL of 17a(H), 21b(H)-hopane,spiked with the internal standard C30bb-hopane (17b(H),21b(H)-hopane of 100 ng/mL), were prepared to develop the calibrationcurve.

2.6. GC/MS analysis of hopanes and steranes in crude and weatheredtar ball samples

Hopane analysis was performed using an Agilent Technologiestriple quadrupole (7000B) GC/MS/MS system run in the GC/MSSIM mode. Chromatographic separation for hopanes was achievedusing a DB-EUPAH (J&W Agilent Technologies) column(20 m � 180 lm � 0.14 lm) in constant pressure mode. The initialGC oven temperature (50 �C maintained for initial 2 min) wasramped to 310 �C (for 1 min) at 6 �C/min and held for 15 minresulting in a 60.3 min total run time. A post-run step was com-pleted using a back-flush column, a novel technology available inour Agilent GC. The back-flush post-run step allowed reverse flush-

Deepwater Horizon (EDH) oil, mousse and tar ball samples. Ts: 18a(H)-22,29,30-ane; C30: 17a(H),21b(H)-hopane; C31: 17a(H),21b(H)-31-homohopane-22S/22R;-22S/22R; C33: 17a(H),21b(H)-33-tetrakishomohopane-22S/22R; IS) internal stan-

Fig. 3. Hopane patterns in seven tar ball samples collected from Alabama beaches.

150 V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154

ing of residual compounds retained within the column. The back-flush run was performed for 4 min at 310 �C. The ion source tem-perature was maintained at 280 �C and the quad temperatureswere set at 180 �C. Helium was used as carrier gas, and the heliumflow rate was set at1 mL/min. The inlet pressure was 24.7 psi andinlet temperature was set at 280 �C; sample injection (1 lL) wasperformed in the pulsed splitless mode. Target hopane analysis(in crude, weathered oil, and tar ball samples) was performedusing a characteristic precursor ion at m/z 191 in SIM mode. Meth-od extraction and cleanup recovery studies were performed byspiking samples with a known concentration of C30ab-hopane;recoveries were in the range of 86–91%.

Steranes fingerprints were developed using a longer Agilent GCcolumn (19091 J-436, 60 m � 250 lm � 0.25 lm) in the SIM mode.The oven temperature was kept at 50 �C during injection (for1 min) then the temperature was increased at a rate of 70 �C/minto 150 �C. After 2 min, the temperature was raised to 310 �C at arate of 5 �C/min and was fixed at 310 �C for 15 min; the total runtime for the method was 51.4 min. The inlet and source tempera-tures were maintained at 300 �C. SIM chromatograms for steraneswere developed using the standard m/z 217 ion (Rosenbauer et al.,2010). All the qualitative and quantitative datasets were acquiredusing an Agilent data acquisition system and were analyzed usingAgilent Technologies Mass Hunter Workstation Qualitative (B4.0)and Quantitative Analysis (B 5.0) software.

3. Results and discussion

3.1. Comparison of hopane and sterane fingerprints

Hopane fingerprints for DH, EDH oil, mousse and tar-ball (TB1-TB7) samples are shown in Figs. 2 and 3. The structural assign-ments of various hopanes were achieved by pattern recognitionof mass spectra, comparison of GC-retention time data with refer-ence standards, and through available literature data for hopanefingerprints (Peters and Modowan, 1993; Wang et al., 1997). Thechromatographic profiles show that hopane distributions in DHoil, mousse and DH-related tar ball samples are dominated byC27–C35 pentacyclic hopanes, with high levels of C30ab-hopane(see Figs. 2 and 3). The concentrations of Ts, Tm and C31–C35 22S/22R homohopane epimers in these samples are relatively low com-pared to C30ab-hopane levels. The figures show that hopane distri-bution in DH crude oil, mousse and weathered-tar-ball sampleshave similar-looking fingerprints.

Although the above results have shown that the DH oil and itsrelated tar balls contained C30ab-hopane as the major biomarker,crude oils can vary in their hopane content from C27 to C35 andhave a unique source-specific finger print. To demonstrate source-specific hopane distribution patterns, three crude oil samples ofdifferent geological origin (AC, BC, and VHC samples) were analyzedand their hopane fingerprints were compared with those of DH oil

Fig. 4. Distribution of hopanes in Deepwater Horizon oil (DH), Arabian crude (AC), Bazra crude (BC), and Venezuelan heavy crude (VHC).

V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154 151

in Fig. 4. These data show that the intensity of C27, C29 and C30

hopanes are distinctly different in all four crude oils, thus demon-strating the source-specific nature of hopane fingerprints.

The chromatographic fingerprints of steranes in DH oil, mousseand tar balls (TB2 and TB7) were also developed using a GC/MSmethod acquired in SIM mode (m/z of 217). Comparisons of chro-matographic profiles of steranes in DH oil, mousse and tar balls areshown in Fig. 5. The results show that C27, C28 and C29 steraneswere abundant in DH oil, which is known to be the characteristicsignature of DH source crude (Rosenbauer et al., 2010). Chromato-graphic signatures of steranes in mousse sample are similar to DHoil (see Fig. 5). Similarly, the relative distribution of various steranepeaks in mousse and tar ball samples are almost identical indicat-ing that these samples must have originated from the same source.

3.2. Diagnostic ratios of source-specific hopanes

The relative ratios of source-specific hopanes often differ fromoil to oil. This variation depends on source rock, depositional envi-ronment, and maturity of the oil. Wang et al. (2004) reported thatthe diagnostic ratios of hopane biomarkers commonly used for oilspill fingerprinting include: Ts/Tm, Ts/(Ts + Tm), C29/C30, and homo-hopane. The ratios of Ts/Tm, Ts/(Ts + Tm) and C29/C30 were also re-ported as characteristic markers and can be used for sourceidentification (Wang et al., 1994c, 2011). We computed the follow-ing hopane characteristic ratios: Ts/Tm, C29ab/C30ab, C31(22S)/C31(22S + 22R), C32(22S)/C32(22S + 22R), C33(22S)/C33(22S + 22R),

C34(22S)/C34(22S + 22R), and C35(22S)/C35(22S + 22R). The ratioswere calculated for all the field samples (mousse and seven tar ballsamples) and were compared against the values estimated for theDH source crude oil and the results are summarized in Table 2.These estimates were derived from the estimated values of peakareas of different characteristic hopanes. The computed valuesfor Ts/Tm for DH oil and mousse samples were 0.91 and 0.92,respectively; the values of C29/C30 for DH oil and mousse sampleswere 0.38 and 0.37, respectively. The ratios of Ts/Tm for tar ballsamples ranged from 0.92 to 0.95, and the values of C29/C30 rangedfrom 0.37 to 0.38 (see Table 2). The agreement of diagnostic ho-pane ratios for the tar ball samples and their match with those val-ues estimated for the DH source oil provide compelling evidencethat all these tar balls originated from the DH oil spill. The datashow that the characteristic hopane pair ratios in DH related oiland tar ball samples, especially Ts/Tm and C29/C30, were relativelyunaffected by weathering; similar results have been shown by oth-ers (Shen, 1984). Shen (1984) also pointed out that samples thatdiffer in their Ts/Tm ratios by over 20% most likely originated fromdifferent sources. We compared the diagnostic hopane ratios of thethree reference crude oils (AC, BC, and VHC) with DH oil. The ratiosof Ts/Tm in AC, BC, VHC, and DH oils are 0.82, 0.19, 0.35 and 0.91,respectivly; and C29/C30 ratios are 1.04, 1.19, 0.63 and 0.38, respec-tively (Table 2). As expected, the diagnostic ratios of both Ts/Tm andC29/C30 measured in AC, BC, and VH coils are considerably differentfrom the values measured in DH oil indicating that all four oil sam-ples originated from distinctly different reservoirs.

Fig. 5. Chromatographic signatures of steranes (m/z = 217) in DH oil, mousse collected on June 2010, tar ball collected on September 2011 (TB2), and tar ball collected onFebruary 2012 (TB7).

Table 2Diagnostic ratios of characteristic hopanes in crude oil, mousse, and tar ball samples.

Sample Diagnostic ratios of hopanes

Ts/Tm Ts/(Ts + Tm) C29/C30 C31S/(S + R) C32S/(S + R) C33S/(S + R) C34S/(S + R)

DH oil 0.91 ± 0.06 0.47 ± 0.02 0.38 ± 0.02 0.61 ± 0.01 0.65 ± 0.01 0.60 ± 0.02 0.67 ± 0.04EDH oil 0.91 ± 0.04 0.47 ± 0.01 0.37 ± 0.02 0.63 ± 0.02 0.62 ± 0.01 0.63 ± 0.01 0.63 ± 0.01Mousse 0.92 ± 0.05 0.48 ± 0.01 0.37 ± 0.01 0.63 ± 0.01 0.65 ± 0.01 0.61 ± 0.01 0.63 ± 0.01TB1 0.92 ± 0.06 0.48 ± 0.02 0.38 ± 0.01 0.63 ± 0.03 0.64 ± 0.02 0.63 ± 0.03 0.65 ± 0.03TB2 0.93 ± 0.03 0.49 ± 0.01 0.37 ± 0.01 0.64 ± 0.01 0.64 ± 0.01 0.61 ± 0.02 0.59 ± 0.01TB3 0.95 ± 0.03 0.49 ± 0.01 0.38 ± 0.01 0.64 ± 0.03 0.63 ± 0.01 0.60 ± 0.03 0.62 ± 0.01TB4 0.94 ± 0.03 0.48 ± 0.02 0.37 ± 0.02 0.62 ± 0.01 0.63 ± 0.01 0.65 ± 0.01 0.63 ± 0.02TB5 0.93 ± 0.06 0.48 ± 0.02 0.37 ± 0.02 0.61 ± 0.01 0.64 ± 0.02 0.63 ± 0.02 0.63 ± 0.01TB6 0.93 ± 0.03 0.48 ± 0.01 0.37 ± 0.02 0.62 ± 0.00 0.63 ± 0.02 0.63 ± 0.03 0.65 ± 0.02TB7 0.93 ± 0.03 0.49 ± 0.01 0.37 ± 0.02 0.62 ± 0.00 0.63 ± 0.00 0.63 ± 0.01 0.66 ± 0.01AC 0.82 ± 0.01 0.45 ± 0.01 1.04 ± 0.05 0.62 ± 0.02 0.64 ± 0.01 0.59 ± 0.01 0.62 ± 0.03BC 0.19 ± 0.01 0.16 ± 0.01 1.19 ± 0.00 0.60 ± 0.00 0.65 ± 0.02 0.55 ± 0.06 0.61 ± 0.01VHC 0.35 ± 0.03 0.26 ± 0.01 0.63 ± 0.02 0.60 ± 0.00 0.63 ± 0.00 0.60 ± 0.03 0.58 ± 0.04

152 V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154

3.3. Quantitation of primary hopane (C30ab-hopane) in crude andweathered oil samples

The concentration of C30ab-hopane (the primary hopane com-pound) was quantified in all crude-oil and tar-ball samples. A cal-ibration curve was first developed using Chiron C30ab-hopanestandard (>98% by GC/MS) with concentrations ranging from 50to 400 ng/mL and spiked with an internal standard (C30bb-hopane,100 ng/mL, >98% by GC/MS). The calibration response was linearacross the selected analytical range, yielding a correlation coeffi-cient (r2) value of 0.999. Quantitative results indicate that DH oilcontained 53 ± 3 mg/kg of C30ab-hopane, which is consistent withthe values of 44 ± 21 mg/kg reported in a multi-laboratory studycompleted by National Institute of Standards and Technology

(NIST) (Schantz and Kucklick, 2011). The average values of C30ab-hopane in EDH and mousse samples were 103 and 91 mg/kg,respectively. The amount of C30ab-hopane in the tar ball samplesTB1, TB2, TB3, TB4, TB5, TB6 and TB7 were 102, 104, 111, 123,103, 120 and 119 mg/kg oil, respectively. The amount of C30ab-ho-pane in AC, BC, and VHC crude oil samples was 94, 131 and 85 mg/kg, respectively.

3.4. Assessment of weathering levels using C30ab-hopaneconcentrations

We used C30ab-hopane as an internal conservative biomarker toestimate the degree of weathering of EDH, mousse and the seventar ball samples using the following equation (Wang et al., 2001):

C30 αα

ββ -hopane concentration (mg/kg)

Wea

ther

ing

Days since accident

Weathering % Hopane concentration

Fig. 6. Changes in C30ab-hopane concentration and percentage weathering level(data for DH oil, mousse and tar-ball samples are arranged based the time they wereexposed to environmental weathering processes).

V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154 153

Pð%Þ ¼ 1� Cs

Cw

� �� 100;

where P is the percentage weathering level with respect to thesource crude, and Cs and Cw are the concentrations of C30ab-hopanein the source oil (DH) and weathered sample, respectively. Theweathering levels of EDH and mousse samples were found to be49% and 42%, respectively. The level of weathering estimated forthe tar ball samples TB1–TB7 were: 48%, 49%, 52%, 57%, 48%, 56%,and 55%, respectively. The EDH sample was prepared by evaporat-ing a known amount DH source crude oil for about 6 months. Thesample was periodically weighed and gravimetric analysis of thedata indicated that the crude oil evaporated by 25% within a day,30% within 2 days, 40% within a month, and about 47% in 6 months.The hopane-based procedure estimated that the 6-month old EDHsample should have weathered by 49%, which compares well withthe gravimetric-measurement based weathering level of 47%. Thisdata indicates that the accumulation level of C30ab-hopane inweathered samples can be used to track the overall weatheringlevel.

Fig. 6 summarizes all the C30ab-hopane data and the estimatedvalues of percentage weathering levels for all the field samples. Thedata points shown in Fig. 6 are organized based on the approxi-mate time (in days) taken for the sample to weather in the envi-ronment, with zero indicating the beginning of the accident (notethe data for fresh oil collected at the well head plotted at timezero). Floating mousse material was collected from Alabamabeaches on June 11th, 2010 (assumed to have weathered about50 days over the ocean). The hopane concentration levels showthat due to various ocean-scale transport processes, includingevaporation and dissolution, oil in the form of mousse that arrivedalong the beaches had weathered by about 42%. However, the aver-age weathering level of the tar ball samples collected after about2 years was about 50%. This indicates that very little weatheringhas occurred since DH oil was buried in Alabama’s nearshoreenvironment.

4. Conclusions

We used GC/MS methods for fingerprinting hopanes and ster-anes in source crude oil and weathered tar ball samples. The meth-ods use simplified extraction and cleanup protocols that arerelatively easy to implement. This study is the first to quantifyhopane chemical profiles of multiple DH oil spill related tar ballscollected along the Alabama shoreline. In this work we havepresented hopane fingerprints, diagnostic ratios of characteristic

hopane pairs, and C30ab-hopane concentration levels for DHsource crude oil, three other reference crude oils, emulsifiedmousse collected on Alabama’s beaches in June 2010, and sevendifferent tar balls collected from Alabama’s beaches from Septem-ber 2011 to February 2012. We have also provided the sterane fin-gerprints of several field samples and compared them against DHoil fingerprints. The C30ab-hopane was found to be the most abun-dant hopane in DH oil and C30ab-hopane concentration was foundto be 53 ± 3 mg/kg, which is well within the range of 44 ± 21 mg/kgreported in a multi-laboratory study completed by NIST (Schantzand Kucklick, 2011). Comparison of hopane fingerprints, selectivesterane fingerprints, and characteristic hopane ratios, especiallythe ratios of Ts/Tm and C29/C30, show that the mousse and tar-ballsamples analyzed in this study originated from DH oil. Based onmeasured C30ab-hopane concentrations, the average weatheringlevel of the tar ball samples collected after about 2 years (betweenSeptember 2010 and February 2012) is 50%, which is quite close tothe weathering level of 42% estimated for the first-arrival moussecollected in June 2010. This result indicates that DH oil submergedalong the Alabama coast has not weathered significantly over thepast 2 years. The lab-synthesized tar ball sample (EDH sample)showed that evaporation alone can weather DH oil by about 40%within a month, suggesting evaporation during transport acrossthe GOM (from the MC252 well head to Alabama’s shoreline)was likely the most important weathering process.

Results of a field survey summarized in Unified Command’sshoreline cleanup completion plan (SCCP, 2011) pointed out thatMC252-related tar balls have unique physical characteristics: theyare fragile, soft, sticky brownish material containing considerableamount of sand and have a noticeable petroleum odor. All the fieldsamples recovered in this study matched these physical character-istics. The shoreline cleanup completion plan also stated that it ispossible to occasionally encounter some highly weathered non-MC252 tar balls on Alabama’s beaches; however, these rare non-MC252 tar balls are comprised primarily of high molecular weightasphaltenes, contain very little to no sand, are usually very hard,and have a very weak petroleum odor (and often no odor). Duringa background survey completed by our team on May 8th and 9th,2010, about three weeks prior to the arrival of DH oil on Alabama’sbeaches, we saw no sign of soft crude-oil residues along the Ala-bama shoreline. Also, to the best of our knowledge, there is no re-cord of anyone ever recovering other types of fragile, soft, stickybrownish tar balls from Alabama’s beaches, tainted with SouthLouisiana crude oil other than MC252 oil, at least within the past3–4 year period. Our post oil-spill sampling surveys, particularlythe observations made after major storm events such as TropicalStorm Lee and Hurricane Isaac, indicated that literally thousandsof fragile, sticky, brownish tar balls continue to be deposited alongour study region, a 28-mile long sandy white beach region that ex-tends from Orange Beach to Fort Morgan (Clement et al., 2012).Based on the hopane and sterane datasets presented in this study,and based on the fact that there is no record of fragile, sticky tarball deposition in our study region prior to the DH oil spill, onecould safely conclude that virtually all fragile, sticky, brownishtar balls currently found on Alabama’s beaches are from the Deep-water Horizon oil spill.

Acknowledgments

This work was supported, in part, by funding received from theCity of Orange Beach, Alabama. The GC/MS/MS facility and charac-terization methods were developed using grant funds receivedfrom the National Science Foundation (NSF), Samuel Ginn Collegeof Engineering, Auburn University, and Marine Environmental Sci-ences Consortium (MESC).

154 V. Mulabagal et al. / Marine Pollution Bulletin 70 (2013) 147–154

References

Barakat, A.O., Mostafa, A.R., Rullkotter, J., Hegazi, A.R., 1999. Application of amultimolecular marker approach to fingerprint petroleum pollution in themarine environment. Mar. Pollut. Bull. 38, 535–544.

Clement, T.P., Hayworth, J.S., Mulabagal, V., Yin, F., John, G.F., 2012. Research Brief II:Impact of Hurricane Isaac on Mobilizing Deepwater Horizon Oil Spill Residuesalong Alabama’s Coastline – A Physicochemical Characterization Study. SamuelGinn College of Engineering, Auburn University, <http://www.eng.auburn.edu/files/acad_depts/civil/oil-research-hurricane-isaac.pdf> (accessed 09.02.13).

Escobar, M., Márquez, G., Inciarte, S., Rojas, J., Esteves, I., Malandrino, G., 2011. Theorganic geochemistry of oil seeps from the Sierra de Perijá eastern foothills,Lake Maracaibo Basin, Venezuela. Org. Geochem. 42, 727–738.

Hauser, A., Dashti, H., Khan, Z.H., 1999. Identification of biomarker compounds inselected Kuwait crude oils. Fuel 78, 1483–1488.

Hayworth, J.S., Clement, T.P., 2011. BP’s operation deep clean-could dilution be thesolution to beach pollution? Environ. Sci. Technol. 45, 4201–4202.

Hayworth, J.S., Clement, T.P., 2012. Provenance of Corexit-related chemicalconstituents found in near shore and inland Gulf Coast waters. Mar. Pollut.Bull. 64, 2005–2014.

Hayworth, J.S., Clement, T.P., Valentine, J.F., 2011. Deepwater Horizon oil spillimpacts on Alabama beaches. Hydrol. Earth Syst. Sci. 15, 639–3649.

Idris, H.K., Salihu, A., Abdulkadir, I., Almustapha, M.N., 2008. Application ofgeochemical parameters for characterization of oil samples using GC–MStechnique. Int. J. Phys. Sci. 3, 152–155.

Peters, K.E., Walters, C.C., Moldowan, J.M., 2005. The Biomarker Guide. Biomarkersand Isotopes in Petroleum Systems and Earth History, vols. 1 and 2, CambridgeUniversity Press, USA, p. 1155.

McNutt, M.K., Camilli, R., Crone, T.J., Guthrie, G.D., Hsieh, P.A., Ryerson, T.B., Savas,O., Shaffer, F., 2011. Review of flow rate estimates of the Deepwater Horizon oilspill. Proc Natl Acad Sci USA.

Mostafa, A.R., Hegazi, A.H., El-Gayar, M.S., Andersson, J.T., 2009. Sourcecharacterization and the environmental impact of urban street dusts fromEgypt based on hydrocarbon distributions. Fuel 88, 95–104.

Pauzi Zakaria, M., Okuda, T., Takada, H., 2001. Polycyclic aromatic hydrocarbon(PAHs) and hopanes in stranded tar-balls on the coasts of Peninsular Malaysia:applications of biomarkers for identifying sources of oil pollution. Mar. Pollut.Bull. 42, 1357–1366.

Peters, K.E., Modowan, J.M., 1993. The Biomarker Guide. Interpreting MolecularFossils in Petroleum and Ancient Sediments. Prentice-Hall+, Englewood Cliffs,NJ.

Peters, K.E., Walters, C.C., Moldowan, J.M., 2007. The Biomarker Guide, Biomarkersand Isotopes in Petroleum Exploration and earth history, second ed., vol. 2,Cambridge.

Prince, R.C., Clark, J.R., Lindstrom, J.E., Butler, E.L., Brown, E.J., Winter, G., Grossman,M.J., Parrish, P.R., Bare, R.E., Braddock, J.F., Steinhauer, W.G., Douglas, G.S.,Kennedy, J.M., Barter, P.J., Brag, J.R., Harner, E.J., Atlas, R.M., 1994.Bioremediation of the Exxon Valdez Oil Spill: Monitoring Safety and Efficacy.Hydrocarbon Bioremediation. Lewis Publishers, Ann Arbor, MI, pp. 107–124.

Rosenbauer, R.J.C., P.L., L., Angela, Lorenson, T.D., Hostettler, F.D., Thomas, B., Wong,F.L., 2010. Reconnaissance of Macondo-1 Well Oil in Sediment and Tar ballsfrom the Northern Gulf of Mexico Shoreline, Texas to Florida: US GeologicalSurvey Open-File Report 2010–1290, p. 22.

SCCP, 2011. Shoreline clean up completing plan. Published by the Unified CommandSCCP Core Group, November 2011.

Schantz, M.M., Kucklick, J.R., 2011. Inter-laboratory analytical comparison study tosupport Deepwater Horizon natural resource damage assessment: descriptionand results for crude oil QA10OIL01. National Institute of Standards andTechnology (NIST) Report, Gaithersburg, MD 20899.

Shen, J., 1984. Minimization of Interferences from weathering effects and use ofbiomarkers in identification of spilled crude oils by gas-chromatography mass-spectrometry. Anal. Chem. 56, 214–217.

Stout, S.A., 2003. Applications of petroleum fingerprinting in known and suspectedpipeline releases—two case studies. Appl. Geochem. 18, 915–926.

Wang, Z.D., Fingas, M.F., 2003. Development of oil hydrocarbon fingerprinting andidentification techniques. Mar. Pollut. Bull. 47, 423–452.

Wang, Z.D., Fingas, M., Sergy, G., 1994c. Study of 22-year-old arrow oil samplesusing biomarker compounds by Gc/Ms. Environ. Sci. Technol. 28, 1733–1746.

Wang, Z.D., Fingas, M., Li, K., 1994a. Fractionation of a light crude-oil andidentification and quantitation of aliphatic, aromatic, and biomarkercompounds by Gc-Fid and Gc–Ms 1. J. Chromatogr. Sci. 32, 361–366.

Wang, Z.D., Fingas, M., Li, K., 1994b. Fractionation of a light crude-oil andidentification and quantitation of aliphatic, aromatic, and biomarkercompounds by Gc-Fid and Gc–Ms 2. J. Chromatogr. Sci. 32, 367–382.

Wang, Z.D., Fingas, M., Landriault, M., Sigouin, L., Feng, Y.P., Mullin, J., 1997. Usingsystematic and comparative analytical data to identify the source of anunknown oil on contaminated birds. J. Chromatogr. A 775, 251–265.

Wang, Z.D., Fingas, M., Page, D.S., 1999. Oil spill identification. J. Chromatogr. A 843,369–411.

Wang, Z.D., Fingas, M., Owens, E.H., Sigouin, L., Brown, C.E., 2001. Long-term fateand persistence of the spilled Metula oil in a marine salt marsh environment –degradation of petroleum biomarkers. J. Chromatogr. A 926, 275–290.

Wang, Z.D., Fingas, M., Lambert, P., Zeng, G., Yang, C., Hollebone, B., 2004.Characterization and identification of the Detroit River mystery oil spill(2002). J. Chromatogr. A 1038, 201–214.

Wang, Z.D., Stout, S.A., Fingas, M., 2006. Forensic fingerprinting of biomarkers for oilspill characterization and source identification. Environ. Forensics 7, 105–146.

Wang, Z., Yang, C., AL, E., 2007. Petroleum biomarker for oil spill characterizationand source identification. In: Wang, Z., Stout, S. (Eds.), Oil Spill EnvironmentalForensics. Academic Press, Oxford.

Wang, Z., Yang, C., Yang, Z., Sun, J., Hollebone, B., Brown, C., Landriault, M., 2011.Forensic fingerprinting and source identification of the 2009 Sarnia (Ontario) oilspill. J. Environ. Monit. 13, 3004–3017.

Yim, U.H., Ha, S.Y., An, J.G., Won, J.H., Han, G.M., Hong, S.H., Kim, M., Jung, J.H., Shim,W.J., 2011. Fingerprint and weathering characteristics of stranded oils after theHebei Spirit oil spill. J. Hazard. Mater. 197, 60–69.