Assessment of photochemical processes in marine oil spill fingerprinting

Marine Pollution Bulletin 79 (2014) 268–277

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Assessment of photochemical processes in marine oil spill fingerprinting Jagoš R. Radovic´ a,b, Christoph Aeppli b,d, Robert K. Nelson b, Núria Jimenez c, Christopher M. Reddy b, Josep M. Bayona a, Joan Albaigés a,⇑ a

Department of Environmental Chemistry, IDAEA-CSIC, Barcelona 08034, Spain Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA c Federal Institute for Geosciences and Natural Resources (BGR), Stilleweg 2, Hannover D-30655, Germany d Bigelow Laboratory for Ocean Sciences, East Boothbay, ME 04544, USA b

a r t i c l e

i n f o

Keywords: Photooxidation Oil spills Oil fingerprinting Polycyclic aromatic hydrocarbons Triaromatic steranes

a b s t r a c t Understanding weathering processes plays a critical role in oil spill forensics, which is based on the comparison of the distributions of selected compounds assumed to be recalcitrant and/or have consistent weathering transformations. Yet, these assumptions are based on limited laboratory and oil-spill studies. With access to additional sites that have been oiled by different types of oils and exposures, there is a great opportunity to expand on our knowledge about these transformations. Here, we demonstrate the effects of photooxidation on the overall composition of spilled oils caused by natural and simulated sunlight, and particularly on the often used polycyclic aromatic hydrocarbons (PAHs) and the biomarker triaromatic steranes (TAS). Both laboratory and field data from oil released from the Macondo well oil following the Deepwater Horizon disaster (2010), and heavy fuel-oil from the Prestige tanker spill (2002) have been obtained to improve the data interpretation of the typical fingerprinting methodology. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Over one million metric tonnes of petroleum enter the marine environment annually from municipal and industrial sources, marine transport, natural oil seeps and accidental oil spills (GESAMP, 2007). Although the number of the latter has decreased significantly during the past decades, catastrophic accidents such as the sinking of the Prestige tanker near the coast of Galicia (Spain) and the Deepwater Horizon (DWH) platform blowout in the Gulf of Mexico, still pose an important threat to marine and coastal ecosystems, causing extensive environmental (Albaigés et al., 2006; White et al., 2012) and economical (Garza-Gil et al., 2006; McCreaStrub et al., 2011) damages. And while the past history has been promising, new environmental concerns arise as the oil industry is venturing to recover oil into more hostile, challenging and dangerous regions such as ultra deep-water, Arctic and along national boundaries that lack the infrastructure to respond effectively to any mishaps. Identifying the source of oil releases, acute or chronic, is the primary step in assessing their consequences and better defining the response strategies. Efficient and unambiguous analytical methods for the characterization of these spillages are also needed from the standpoint of the enforcement of the pollution-control laws, designed to protect the public health and the environment. The most mature methodology for oil-spill characterization is

⇑ Corresponding author. Tel.: +34 934006152; fax: +34 932045904. E-mail address: [email protected] (J. Albaigés). 0025-326X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.marpolbul.2013.11.029

based on the chemical fingerprinting approach, where a series of petroleum hydrocarbons can be profiled by gas chromatography fitted with a flame ionization detector or coupled to mass spectrometry (GC–FID and GC–MS) (Albaiges and Albrecht, 1979; Wang et al., 2007). Characteristic distributions and ratios of selected ’diagnostic’ compounds generate an oil ’fingerprint’ that can be used to identify the source of the spilled oil. This methodology has been extensively used (Wang and Fingas, 2003) and has recently been standardized (CEN, 2012). In past few years comprehensive two-dimensional gas chromatography (GC  GC) coupled to flame ionization detector (FID) or time-of flight mass spectrometry (TOFMS) has also demonstrated great potential for oil fingerprinting due to its unprecedented resolving capability and excellent sensitivity (Eiserbeck et al., 2012). It has been successfully applied for fingerprinting oil samples (Aeppli et al., 2012; Lemkau et al., 2010; Ventura et al., 2010). However, once in the marine environment, any spilled oil undergoes a variety of physical, chemical, and biological processes (weathering), including evaporation, dissolution, microbial degradation and photooxidation that modify the original oil compositional patterns (NRC, 2003). A fundamental understanding of these processes is essential to refine the diagnostic value of the source recognition indices and interpret the profiles in tracking oil sources. This is particularly important for identifying the sources in areas of heavy traffic (Diez et al., 2007) or with natural oil seeps, such as in some areas of the Gulf of Mexico (Anderson et al., 1983; MacDonald, 2002), where misinterpretations may arise by the concurrence of other sources.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

The physical and biological processes are well studied and have been reviewed in detail (Fingas, 1995; Prince, 1988; Yang and Wang, 1977), but photooxidation is still poorly characterized (Nicodem et al., 2001; Payne and Phillips, 1985; Plata et al., 2008). Oil beached above the high tide mark is more exposed and more affected by direct sunlight (Douglas et al., 2002; Prince et al., 2003). In the sea surface microlayer, the process is thought to be driven by a series of radical reactions due to the presence of photosensitizers (e.g., dissolved organic matter), which can be excited by light to form various reactive species (RO, HO2 , HO, etc.) (Liss et al., 2005; Schwarzenbach et al., 2005). Deeper in the euphotic zone, dissolved or particulate-associated oil molecules, such as polycyclic aromatic hydrocarbons (PAHs), can be photooxidized directly by absorbing UV radiation (Lee, 2003). A number of studies have been focused on the increase of toxicity of oil residues after photooxidation (Maki et al., 2001; Lee, 2003; Barron et al., 2005). The coupling of photooxidation with biodegradation has also been studied (Dutta and Harayama, 2000; Maki et al., 2001), but the information on the oil photochemical transformations at molecular level is limited (Jacquot et al., 1996; Bobinger and Andersson, 2009; Charrie-Duhaut et al., 2000). In this respect, it is well known that oil weathering leads to an increase of polar components, akin in basic characteristics to resins and asphaltenes found in crude oils (Albaiges and Cuberes, 1980; Garrett et al., 1998; Maki et al., 2001; Prince et al., 2003; Aeppli et al., 2012). Moreover, there is evidence that methyl-substituted aromatic molecules photochemically oxidize at a faster rate than the parent compounds (Garrett et al., 1998). Some previous reports have also suggested that particular biomarker molecules used in oil forensics, such as triaromatic steranes (TAS), can be unexpectedly depleted due to field or simulated weathering (Barakat et al., 2002; Charrie-Duhaut et al., 2000; Jacquot et al., 1996). However, a more comprehensive research on the effects of photooxidation on the reliability of the oil fingerprinting methodology is still lacking. The goal of this paper is to investigate compositional changes of marine oil spills that can be attributed to photooxidation and how they may affect the diagnostic ratios commonly used in oil spill fingerprinting. To this end, laboratory and field data have been obtained from two different oils, Macondo well (MW) oil from the 2010 Deepwater Horizon disaster in the Gulf of Mexico (SE USA), and heavy fuel-oil from the 2002 Prestige tanker spill off the Galicia coast (NW Spain) that were analyzed using GC  GC–FID and GC–MS fingerprinting methodologies. These included the analyses of PAHs and TAS diagnostic compounds in weathered field samples from the Gulf of Mexico and the Galicia coast, and irradiated samples in two laboratory-scale experiments, using natural sunlight and a Xe lamp. Target compounds in original and weathered oil samples were compared to calculate their losses, which in turn were used to calculate commonly used fingerprinting ratios to test if they can be reliably employed in environmental forensics in their current fashion.

2. Materials and methods 2.1. Chemicals All solvents (dichloromethane, n-hexane, toluene and methanol) were obtained from Merck (SupraSolvÒ) (Darmstadt, Germany). Neutral alumina and anhydrous sodium sulfate were also obtained from Merck and activated at 400 °C overnight.

269

(IGT) sampler, in June 2010, deployed from a remotely operated vehicle (ROV) (Reddy et al., 2012). Field sampling of Prestige oil was focused not on the major oil paths or the recently oiled shorelines, but on the lumps appearing at sea or arriving at the coast from time to time during 2003 and 2004. Samples were collected with a metal spoon, placed in precleaned amber glass jars and stored in a portable refrigerator for transport to the laboratory. The MW oil was recovered from oil splashes on jetty rocks (‘‘rock scrapings’’) collected at two time points, first in April 2011 (350 days after the spill) at Buckaneer State Park (30°150 6500 , 89°240 2200 ) and Waveland (30°160 5500 , 89°220 0700 ), MS and second in August 2012 (750 days after the spill) at Fort Gaines, AL (30°140 ’4700 , 88°40 3200 ). They were found above the sea level, and were exposed to sunlight. Prestige samples (80–150 mg) were dissolved or extracted (in the case of field samples) with hexane and cleaned-up in an open glass-column, over anhydrous Na2SO4, and neutral alumina (5% water deactivated), eluting with hexane. The eluate was reduced in volume under a nitrogen stream to provide a concentration suitable for injection into the GC–MS. The samples of fresh MW oil and field rock scraping samples were prepared as described previously (Aeppli et al., 2012). Briefly, the MW oil sample was dissolved in dichloromethane (DCM) at concentrations of 10–50 mg mL1. The rock scrapings were extracted three times with DCM/methanol (80/20), by vigorously shaking and centrifuging (1600 rpm for 5 min), and the combined extracts were dried over anhydrous Na2SO4 and used for further analysis by TLC-FID and GC  GC–FID. 2.3. Irradiation experiments Both natural and simulated solar irradiation was used to investigate the photooxidation of the two oils. In the first experiment, approximately 80 mg of the Prestige oil was placed in an uniform thin layer in Petri plates and irradiated using a SUNTESTÒ CPS flatbed Xe-exposure system (Atlas, Chicago, USA), equipped with a 1500B NrB4 Xe lamp that was operated at the potential of 507.5 W/m2 to simulate natural irradiation. The system is equipped with a ventilator to maintain a constant temperature and prevent sample overheating. Control plates covered with Al foil were irradiated simultaneously. Samples were collected after 12 h, a time span found adequate for simulating moderate field conditions. The oil was washed off the plates with 3–5 mL of DCM, the solvent was then evaporated under gentle nitrogen stream, and the oil was finally cleaned-up for injection to GC–MS as described in Section 2.2. For the MW oil experiments, approximately 50 mg of oil was added to each quartz glass tube to form relatively uniform thin layer. Control tubes were wrapped in Al foil. The tubes were mounted horizontally on an Al foil wrapped surface and exposed to sunlight on the top of a 3-m high cargo container in the Woods Hole Oceanographic Institution Quisset Campus (Woods Hole, MA). The duplicate tubes were collected every day during the first week, then after 20 and 70 days when the experiment finished (18 samples in total). Oil was extracted with DCM/methanol (80/20) and dried over anhydrous Na2SO4 for further analysis. Experiment was performed from May to August 2012, and the average irradiation potential during this period was 750 W/m2 as recorded by the nearby solar station (41°420 3800 , 70°40 3100 ). 2.4. Instrumental analysis

2.2. Samples and sample preparation The fuel oil sample was obtained from the Prestige cargo tanks and the MW oil from the gushing well using an isobaric gas-tight

Samples were characterized by Fourier transform infrared spectroscopy (FT-IR) with a Nicolet Avatar 360 Thermo Scientific Spectrometer (Waltham, MA, USA) and by thin layer chromatography

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

270

(TLC) coupled to FID, using a MK-5 Iatroscan apparatus (Iatron Labs, Tokyo, Japan) according to previously reported procedures (Aeppli et al., 2012; Radovic et al., 2012). Briefly, it involved a three-step sequential elution of the sample extracts on silicacoated quartz rods (ChromaRodÒ-SIII) using solvents of increasing polarity [n-hexane, n-hexane:toluene and methanol:DCM, respectively] resulting into four different classes, namely saturates (S), aromatics (Ar), and two more polar fractions (PI) and (PII) (Kamin´ski et al., 2003). GC–MS analysis was carried out with a TRACE 2000 MS Thermo-Finnigan instrument (Manchester, UK) in the electron impact (EI) mode at 70 eV. A 20 m  0.18 mm i.d. capillary column coated with 0.18 lm of TRB-5MS stationary phase (Teknokroma, Sant Cugat del Vallés, Spain) was used. The carrier gas was He, with a constant flow rate of 0.6 mL min1. The sample (1 lL) was injected in the splitless mode, the injector temperature was held at 280 °C and the purge valve was activated 50 s after the injection. The column temperature was held at 60 °C for 1 min, then the temperature was increased to 200 at 14 °C min1 and finally to 320 at 7.5 °C min1, holding that temperature for 4 min. Transfer line and ion source temperatures were held at 250 and 200 °C, respectively. Data were acquired in the full-scan mode from 50 to 350 amu (10 scan s1) with 6 min of solvent delay and processed by the Xcalibur Thermo Finnigan software (San Jose, California, USA). The GC  GC–FID system employed a dual stage cryogenic modulator (Leco, Saint Joseph, Michigan) installed in an Agilent 7890A gas chromatograph configured with a 7683 series split/splitless auto-injector and two capillary columns. Each sample was injected in splitless mode and the purge vent was opened at 0.5 min. The

3. Results and discussion 3.1. Overall compositional changes The fuel-oil cargo of the Prestige tanker was a heavy residue (d15°C = 0.993 kg/L; sulfur = 2.6%; nitrogen = 0.69%) that was composed of 22%, 49%, and 29% saturates, aromatic hydrocarbons, and resins and asphaltenes, respectively (Diez et al., 2005). The Macondo oil was a light crude oil (d15°C = 0.820 kg/L; sulfur = 0.4%; nitrogen = 0.38%) with a saturate, aromatic and resins and asphaltenes content of 74%, 16%, and 10% respectively (Reddy et al., 2012). Despite the different composition of the two oils, the most remarkable changes observed after irradiation, either by natural sunlight or with a UV-lamp, were a substantial decline of the

Fresh

50

Irradiated Trans.

(A)

inlet temperature was 300 °C. The 1st dimension column was a nonpolar Restek DB-1 (60 m  0.25 mm i.d., 0.25 lm film thickness) held at 60 °C for 10 min and then ramped to 325 °C at 1.25 °C min1. The thermal modulator cold jet gas was dry N2, chilled with liquid N2. The thermal modulator hot jet air was heated to 15 °C above the temperature of the secondary GC oven. The hot jet was pulsed for 1 s every 12.0 s with a 5.0 s cooling period between stages. The 2nd dimension separations were performed with a 50% phenyl polysilphenylene–siloxane column (SGE BPX50, 1.25 m  0.10 mm i.d., 0.1 lm film thickness) held at 65 °C for 10 min and then ramped to 330 °C at 1.25 °C min1. The carrier gas was H2 at a constant flow rate of 1 mL min1. The FID detector signal was sampled at a rate of 100 Hz.

Composition%

40

S=O S(=O)2

30

C=O

1000

1200

1400

1600

1800

2000

cm -1

20 10 0

S

Ar

PI

Trans.

(B)

PII

C=O

60 1000

Composition %

1500

2000

cm-1

50

Fresh

day7

day20

day70

40 30 20 10 0

S

Ar

PI

PII

Fig. 1. Compositional changes after photooxidation determined by TLC-FID. (A) Prestige fuel-oil irradiated 12 h using a Xe lamp. (B) Macondo well oil irradiated by natural sunlight. Insets show FT-IR spectra.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

or by the introduction of oxygenated groups such as hydroxyl or carbonyl. In this respect, a photooxidation study performed in the laboratory on an Arabian light crude oil showed an increase in oxygen content of resins and asphaltenes along with a relative decrease of the saturate and aromatic fractions (Albaiges and Cuberes, 1980). The formation of oxygenated residues was also observed during oil weathering of the Macondo oil after the DWH disaster (McKenna et al., 2013). The nature of the oxidation products in the resins was rather varied with an appearance in the FTIR spectra of a broad band at 3400 cm1 (hydroxyl group) and a suite of bands in the 1700–1000 cm1 region (see insets in Fig. 1), namely 1760– 1740 cm1 (esters and anhydrides), 1730 cm1 (carbonyl group), 1720–1700 cm1 (carboxylic carbonyl group), 1695 cm1 (alkylarylketones), 1665 cm1 (diarylketones), 1150 cm1 (sulfone) and 1060–1030 cm1 (sulfoxide) (Boukir et al., 2001). Similar patterns were observed in field samples of the Prestige oil (Fernandez-Varela et al., 2006) and the Macondo oil collected after the DWH disaster (Aeppli et al., 2012). Significant changes in the molecular composition of oil were also observed in the analysis of the aromatic fraction by GC–MS, depending on the molecular structure and alkylation of the aromatic backbone. As shown in Fig. 2, in each family of aromatic

100

% depletion

80 60 40 20

B aA

Fl Py

C C 1C C 2C D C BT 1D C BT 2D B T C C B 1C C B 2C B

P C 1P C 2P

N C 1N C 2N

0

Fig. 2. Compositional changes of the PAHs distribution in the Prestige fuel-oil after photooxidation (12 h) using a Xe lamp. Effect of chemical structure and alkyl substitution.

aromatic fraction and a significant increase of the polar fractions relative to the controls, as evidenced by TLC-FID (Fig. 1). This trend, already observed in previous studies (Garrett et al., 1998; Maki et al., 2001; Nicodem et al., 1997; Prince et al., 2003), could be explained by the conversion of the aromatic compounds to more polar derivatives, either by oxidative cleavage of the aromatic rings

100

Laboratory test

80

271

MA

(a)

2-MP

(b)

60 40 20 0 100

3-MP

80 60

9/4-MP 1-MP

40 20 0 15,0

16,0

17,0

18,0

19,0

Time (min) 100

Field sample

(a)

80 60 40 20 0 100

(b)

80 60 40 20 0 16,0

17,0

18,0

19,0

20,0

Time (min) Fig. 3. Compositional changes of methyl-phenanthrenes (MP) and methyl-anthracene (MA) (m/z 192) in the Prestige fuel-oil, before (a) and after photooxidation (b) in the laboratory and in the field.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

272

4-MPy 1-MPy 2-MPy

Laboratory test

100 80 60

(a)

B[b+c]F

40

B[a]F

2-MFl

20 0 100

(b)

80 60 40 20 0

20,0

21,0

22,0

23,0

24,0

25,0

Time (min) 100

-

Field sample

(a)

80 60 40 20 0 100

(b)

80 60 40 20 0

20,0

21,0

22,0

23,0

24,0

25,0

Time (min) Fig. 4. Compositional changes of methyl-fluoranthenes/pyrenes and benzofluorenes (m/z 216) in the Prestige fuel-oil before (a) and after photooxidation (b) in the laboratory and in the field.

hydrocarbons the extent of photodegradation increases concurrently with the alkylation degree. This could be explained by the alkyl groups enriching the electron density of the p system of the molecule, facilitating electron excitation and its resulting photodegradation. This effect is also enhanced by the number of aromatic rings in the molecule, so that chrysenes are relatively more depleted than phenanthrenes (Garrett et al., 1998). On the other hand, peri-condensed structures (e.g., pyrene (Py) and benzo[a]anthracene (BaA)) are more sensitive to photooxidation than cata-condensed ones (e.g., fluoranthene (Fl) and chrysene (C)). The large cross-sectional area of peri-condensed compounds makes them more efficient absorbents of the UV-radiation (Plata et al., 2008). The presence of S in the molecule (e.g., dibenzothiophenes, DBT) reduces relatively the photodegradation. Although the S atom would be expected to be the site of preferred oxidation, this seems not to be the case in aqueous media (Bobinger and Andersson, 2009). Conversely, N heteroatom (e.g., carbazoles, CB) produces a relative enhancement. In this case, the alkyl substitution did not affect the degradation rate. In fact, all carbazole derivatives exhibit an extensive decline (80%) (Fig. 2). This could be attributed to the strong effect of the N atom, which contributes its unshared pair of electrons into the p aromatic ring system, facilitating the reactivity of its host molecule and the subsequent photooxidation (Schwarzenbach et al., 2005). The high depletion

of C1–C2 naphthalenes (>60%) is mainly due to evaporation during the test as it is confirmed by the control samples in the dark. These patterns can be conveniently used in assessing the weathering of oil spills by chemical fingerprinting, so that a concurrent increase of the ratios of C2/C3 components or chrysene/ pyrene and DBT/CB derivatives in weathered samples can be indicative of photodegradation. 3.2. Compositional changes at molecular level 3.2.1. Polycyclic aromatic hydrocarbons (PAHs) Considering that the aromatic fraction is the most sensitive to photooxidation (Payne and Phillips, 1985; Prince et al., 2003), the analysis of individual compounds by GC–MS gives a better insight into the observed compositional changes and may provide new diagnostic ratios. Relevant molecular profiles of the Prestige oil before and after photooxidation are shown in Figs. 3–6. In the methylphenanthrenes/anthracene (MPs/MA) profile (m/z 192) (Fig. 3), the most significant change is the preferential degradation of MA over MPs, conversely to what is observed during biodegradation where MA is more persistent than MPs (Bernabeu et al., 2013). It also appears that the 2-MP is slightly more refractory than the other MP isomers, whereas in the case of biodegradation the 9-MP is more

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

%Remaining ¼ 100  %Depleted   X weathered =Hweathered  100 ¼ 100  1  X original =Horiginal

resistant to degradation. The profiles of methyl fluoranthenes/pyrenes (MFl/MPys) and benzofluorenes (BFs) (m/z 216), widely used in oil spill fingerprinting (CEN, 2012), are also altered by photodegradation. The 1- and 2-methylpyrenes as well as the benzofluorenes are significantly reduced with respect to the 4-methylpyrene (Fig. 4). Isomeric preferential photooxidation is also observed within the methyl chrysenes (MCs) (m/z 242), the 2-methyl isomer being the more resistant (Fig. 5). The higher resistance of the S-compounds to photodegradation is reflected in the compositional changes of the C4-phenanthrenes (C4-Ps) and benzo[b]naphto[1,2-d]thiophene (BNT) profile (m/z 234) (Fig. 6), where the former were more depleted around 30% after irradiation relative to BNT. As shown in Figs. 3–6, the results found in the laboratory tests are consistent with those exhibited by the samples collected in the field. Similar trends were observed in the irradiation experiments of the Macondo oil, thus confirming the consistency of the compositional change patterns. A clear way of illustrating quantitatively these compositional changes in weathered samples is shown in Fig. 7 (PW-plot: ‘‘percentage weathering’’ plot) (CEN, 2012). In this case, the individual components of the original and laboratory irradiated Prestige samples are normalized to the 17a(H),21b(H)-hopane (C30-hopane), highly refractory to weathering (Prince et al., 1994), and compared. The percentage remaining of any compound in the weathered sample is calculated by comparison to the original sample, according to Eq. (1): 100

273

where Xweathered and Xoriginal represent the measure (peak area or height) of a compound or compound class in the weathered and original samples, respectively, and Hweathered and Horiginal the peak area or height of the C30-hopane peak in the weathered and original samples, respectively. This percentage is plotted against the retention times of the compounds, which are related to their boiling points and molecular weights. The PW-plots are simple yet effective tools for finding differences caused by the weathering of the spilled oil. This can be seen in Fig. 7, which reveals significant depletion of MA (80%), M-BFls (40%) and M-Pys (20–40%) and C4-Ps (30%) in the photooxidized sample. Evaporation has affected the ratios of the lower-boiling compounds such as C1-naphthalene (C1-N) and sesquiterpanes. 3.2.2. Triaromatic sterane (TAS) biomarkers This biomarker family includes eight compounds (C20-TAS, C21-TAS, C26-TAS(R/S), C27-TAS(R/S), C28-TAS(R/S)) that are pregnane and cholestane derivatives, sharing a common aromatic backbone structure but differing in the structure of the alkyl chain (Fig. 8). TAS have been recognized as highly recalcitrant compounds, unaffected by biodegradation (Diez et al., 2005), and

3-MC

Laboratory test

80

(a)

2-MC 60

1-MC

40

6-MC

20 0 100

(b)

80 60 40 20 0 25,0

26,0

27,0

28,0

29,0

Time (min) 100

Field sample

80

(a)

60 40 20 0 100

(b)

80 60 40 20 0 26,0

ð1Þ

27,0

28,0

29,0

Time (min) Fig. 5. Compositional changes of methyl-chrysenes (m/z 242) in the Prestige fuel-oil before (a) and after photooxidation (b) in the laboratory and in the field.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

274

100

C4-P

Laboratory test

80

(a)

BNT

60 40 20 0 100

(b)

80 60 40 20 0 18

19

20

21

22

23

24

25

26

27

28

Time (min) 100

Field sample

(a)

80 60 40 20 0 100

(b)

80 60 40 20 0 18

19

20

21

22

23

24

25

26

27

28

Time (min) Fig. 6. Compositional changes of C4-phenanthrenes and benzo[b]naphto[1,2-d]thiophene (BNT) (m/z 234) in the Prestige fuel-oil before (a) and after photooxidation (b) in the laboratory and in the field.

120

BNT

% remaining

100

Steranes Triterpanes

C1-F C1-C

C1-P

80

M-Pys

Sesquiterpanes

60

TAS

C4-P

M-BFls

40

C2-N

20

MA

C1 -N

0 7

10

13

16

19

22

25

28

retention time (min) Fig. 7. Percentage of PAHs (circles) and biomarkers (squares) of the Prestige sample after photooxidation (12 h), calculated according to Eq. (1).

successfully used to discriminate hydrocarbon sources in the Prince William Sound (Alaska) after the Exxon Valdez oil spill (Hostettler et al., 1999). However, we observed that irradiated samples of Prestige oil showed a noticeable depletion (20%) of these compounds (Fig. 7) that cannot be explained by biodegradation. Furthermore, GC  GC analysis of the rock scraping samples containing MW oil, collected in the Gulf of Mexico 350 and 750 days

after the DWH blowout, showed respectively, on average, a TAS depletion (calculated according to Eq. (1)) of 87.5% and 96.7% (Fig. 8). These samples were found as oily patches adhered on the rocks, high above the sealine and heavily exposed to solar irradiation for most of the daily sunlight. These observations prompted the study of the effects of photooxidation on TAS biomarkers in MW oil under controlled conditions in the laboratory. Fig. 9 shows that in the first 24 h TAS content already begin to decrease, relative to the original oil, and this trend continued throughout the first week of irradiation and until the end of the experiment (70 days) when no TAS compounds could be detected using the applied methodology. Evaporation can be discarded as the cause of TAS loss since the mass of the oil in the quartz tubes remained stable after an initial loss (30%) in the first 24 h. Moreover, the dark control taken after 20 and 70 days did not show significant change in TAS content. Biodegradation neither can be the cause as the evolution of TAS compounds depletion showed a converging trend not expected in the case of biodegradation, which would preferentially remove individual congeners depending on their physical properties or structure as reported elsewhere (Barakat et al., 2002; Frontera-Suau et al., 2002). Moreover, other early indicators of biodegradation (e.g., n-alkanes) were unchanged. Since no aqueous phase was used in the irradiation experiments, the dissolution could also be

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

275

100 90 80

% depletion

70 60

-CH2CH3 (C20-TAS) -CH(CH3)2 (C21-TAS)

50

H 3C

samples after 350 days samples after 750 days

40

H

CH3

R

CH3

C26-TAS H 3C H 3C

30

H

CH3

Overall depletion 350 days

CH3

C27-TAS

20

CH3

Overall depletion 750 days H 3C

H

CH3

10

CH3

C28-TAS

AS C 28 R -T

AS C 27 R -T

26 STA C 26 S R +C 27 STA S C 28 STA S

C

C 21 -T

C 20 -T

AS

AS

0

Fig. 8. Compositional changes of triaromatic steranes (m/z 231) in Macondo well oil 350 and 750 days after the spill. Inset shows the common aromatic backbone of TAS and alkyl substituents of individual compounds.

100

% depletion

80 C20-TAS C21-TAS C26-TAS C26R+C27S-TAS C28S-TAS C27R-TAS C28R-TAS dark controls (avg. TAS loss)

60

40

20

0 0

1

2

3

4

5

6

7

20

30

40

50

60

70

time (days) Fig. 9. Compositional changes of triaromatic steranes in the Macondo well oil during the photooxidation experiment.

discarded as the cause of the observed transformations. Therefore, photooxidation should be the main driver of the observed TAS depletion in the reported experiment. We hypothesize that such photooxidation driven process of TAS removal is also the cause of observed TAS loss in rock scraping samples of weathered MW oil from the Gulf of Mexico. The rocks where samples were found were not submerged in the moment of sampling, and their location suggests that the oil was probably in limited contact with sea water (e.g., only during high tide). Moreover, estimated water solubility of TAS (1.2  106 to 3.5  103 mg L1, estimated using SPARC on-line calculator v4.6), is similar to the values reported for four and five ring PAHs (Schwarzenbach et al., 2005), whose solubility is generally known to decrease with increasing molecular weight and alkylation

(Moore and Ramamoorthy, 1984; Neff, 1979). Comparable weights of investigated TAS and their alkyl-substituted structure corroborate the estimated low solubility and the conclusion that dissolution was not an important driver of TAS removal in the field samples. Overall based on the non-preferential TAS depletion observed in the field and confirmed in the irradiation experiment and previous studies that report their recalcitrance to biodegradation (Diez et al., 2005), we conclude that the loss of TAS in MW oil samples from DWH blowout was the consequence of photooxidation. Furthermore, the rate and magnitude of TAS depletion observed during the irradiation experiment (average loss of 11% per day during the first week), suggest that under field conditions photooxidation is more important weathering process than usually considered (NOAA, 1992). Possible reasons for the pronounced photosensitivity of TAS could be found in their structure, namely the aromatic (phenanthrene) backbone, and the alkyl substitution. As seen in the case of irradiated Prestige oil, photoreactivity was highly influenced by increasing aromaticity and alkylation. Note the depletion of alkylated phenanthrenes, particularly C4-Ps (Fig. 6), which are structurally related to TAS. Based on the obtained FT-IR spectra, we hypothesize that potential mechanisms could include the cleavage of the alkyl substituents and oxidation of the aromatic rings to form epoxy (Dowty et al., 1974) or quinone structures (Payne and Phillips, 1985). Moreover, benzylic positions are particularly sensitive to autoxidation (Ehrhardt and Petrick, 1984) and triaromatic steroid benzylic ketones, together with a depletion of TAS, have been found in oxidized petroleum bitumens under natural conditions (Charrie-Duhaut et al., 2000). Since the oil photooxidation of PAHs in the oil films depends on the layer thickness (Plata et al., 2008), the kinetics of TAS depletion in the case of DWH oil slicks were probably very similar to the one observed in the irradiation experiment, possibly even higher due to the presence of photosensitizers in the sea water; while in the case of oil found on the rocks, degradation speed probably would be lower, depending on the thickness of the film.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277

276

Table 1 Relative difference between PAH and TAS ratios in the original and laboratory irradiated Prestige oil after 12 h. Ratio

Original

Irradiated

Rel. diff.

2MP/1MP 9 + 4MP/1MP MA/1MP 4MDBT/1MDBT B[a]F/4MPy B[b + c]F/4MPy 2MPy/4MPy 1MPy/4MPy BNT/C4-P 2MC/1MC C2-C/C2-Py C26S/C26R + C27S C28S/C26R + C27S

1.95 1.18 0.37 2.57 0.30 0.25 0.86 0.84 0.75 1.57 0.12 0.48 0.88

1.80 1.00 0.03 2.13 0.20 0.20 0.60 0.69 1.77 2.17 0.39 0.40 0.86

8.45 16.51 170.0 18.72 40.00 22.22 35.62 19.61 80.46 32.09 105.9 18.18 2.30

Table 2 Relative difference between TAS ratios in the original and experimentally irradiated MW oil. Ratio

Original

Irradiated Day 7

C20-TAS/C21-TAS C21-TAS/C26R + C27S-TAS C26S-TAS/C26R + C27S-TAS C28S-TAS/C26R + C27S-TAS C27R-TAS/C26R + C27S-TAS C28R-TAS/C26R + C27S-TAS

0.87 1.55 0.38 0.65 0.58 0.55

1.14 0.99 0.21 0.74 0.38 0.46

20 Rel. diff. 27.23 43.89 57.12 13.05 41.80 17.66

1.02 0.83 0.16 0.58 0.32 0.38

Rel. diff. 16.17 60.23 80.74 12.23 57.79 37.41

3.3. The effect of photooxidation on the diagnostic ratios In order to improve the reliability of the assessment of the differences between the profiles of the spill and suspected source samples, a number of diagnostic ratios between single compounds or groups of compounds selected for their diversity in the chemical composition in petroleum products and their reported response to weathering processes have been selected (Hansen et al., 2007). The ratio comparison is applied to make the method more robust than the qualitative visual comparison of chromatograms and to make it more independent from the experience of the analyst. In this respect, it is important to know the effect of weathering processes on the diagnostic ratios for an adequate interpretation of the results in real cases. A number of ratios that are often used for oil spill forensics and are particularly affected by oxidation, as mentioned in the previous section, are shown in Table 1. It is evident that after photooxidation, all selected PAH ratios except 2MP/ 1MP showed significant differences, which can affect their reliability when used to confirm the source of the spill. However, the knowledge of the effects of irradiation at molecular level can help to interpret the observed differences and attribute them to photooxidation. For example, change of ratios, such as MA/1MP, 2MPy/ 4MPy, 1MPy/4MPy and BNT/C4-P, among others, could serve as early indicators of sunlight exposure of samples in the field. As it can be seen, the 1-methyl isomers of P, DBT and C, as well as the 4-MPy, are the more resistant components within each family. A concurrent increase of the ratios of the chrysene/pyrene components (e.g., C2-C/C2-Py) in weathered samples can also be indicative of photodegradation. Unexpectedly, changes in ratios of TAS biomarkers, which are traditionally considered to be robust at medium- to long-term environmental exposure (Peters et al., 2005), were observed after only 7 days. The relative differences of the majority of the

commonly used TAS ratios between the original and irradiated MW oil samples were above the limit of 14% (Table 2) which is the accepted threshold for a positive match with the source oil (CEN, 2012). The C26S/C26R + C27S ratio is particularly affected and although the C28S/C26R + C27S seems to be more robust (Table 1), based on the present observations, caution should be taken when using TAS ratios for fingerprinting spill samples heavily exposed to sunlight, even in a relatively short timeframe (weeks). In such cases, the sterane and triterpane ratios should be used for the confirmation of the source, while PAH and TAS could rather be more useful to estimate the weathering processes, in particular photooxidation. 4. Conclusions This work estimated the effects of photooxidation on both bulk and molecular oil composition, particularly PAHs and TAS biomarkers, commonly used for fingerprinting oil spill samples. At the bulk level, we observed the transformation of saturate and aromatic fractions to more polar, oxygen containing species. At molecular level, we observed that the photosensitivity increases with increasing aromaticity and alkylation and we demonstrated significant effects on a series of PAH compounds, i.e., methyl anthracene, methyl pyrenes, C4-phenanthrenes and methyl chrysenes. Unexpected and pronounced photodegradation of TAS biomarkers was also demonstrated at laboratory scale, which was comparable to the observations from the field samples. We consider that this effect is due to the aromatic structure found in their common backbone and alkylation, comparable to the effects observed in structurally analogous PAHs. In conclusion, we have provided evidence of the role of photooxidation on oil spills that will contribute to better interpret the oil compositional changes and enhance the fidelity of the fingerprinting techniques. Acknowledgments This research was made possible in part by grants from the NSF (OCE-0960841, RAPID OCE-1043976, RAPID OCE-1042097, EAR0950600, OCE-0961725 and OCE-1333148), and in part by a grant from BP/the Gulf of Mexico Research Initiative (GoMRI-015) and the DEEP-C consortium. Funding was also obtained from the MICINN of Spain (Projects Ref. CTM2008-02718-E/MAR and CTM200802721-E/MAR) through the project ‘‘European concerted action to foster prevention and best response to Accidental marine Pollution ‘‘AMPERA’’ (ERAC-CT2005-016165) within the framework of the EU ERA-Net Initiative (6th Framework Program). J.R.R. kindly acknowledges a pre-doctoral fellowship (JAE Predoc) from the Spanish National Council of Research (CSIC) and European Social Fund (ESF). C.A. acknowledges a Swiss National Science Foundation Postdoctoral Fellowship. Authors wish to thank to Dr. Helen K. White and Patrick Williams of Haverford College for their contribution to FT-IR analysis, as well as to Catherine A. Carmichael, Dr. Karin L. Lemkau, and Ellen Murphy for their assistance in sampling. Constructive comments of one reviewer are greatly appreciated. References Aeppli, C., Carmichael, C.A., Nelson, R.K., Lemkau, K.L., Graham, W.M., Redmond, M.C., Valentine, D.L., Reddy, C.M., 2012. Oil weathering after the Deepwater Horizon disaster led to the formation of oxygenated residues. Environ. Sci. Technol. 46, 8799–8807. Albaiges, J., Albrecht, P., 1979. Fingerprinting marine pollutant hydrocarbons by computerized gas chromatography–mass spectrometry. Int. J. Environ. Anal. Chem. 6, 171–190. Albaiges, J., Cuberes, M.R., 1980. On the degradation of petroleum residues in the marine environment. Chemosphere 9, 539–545.

J.R. Radovic´ et al. / Marine Pollution Bulletin 79 (2014) 268–277 Albaigés, J., Morales-Nin, B., Vilas, F., 2006. The Prestige oil spill: a scientific response. Mar. Pollut. Bull. 53, 205–207. Anderson, R.K., Scalan, R.S., Parker, P.L., Behrens, E.W., 1983. Seep oil and gas in Gulf of Mexico slope sediment. Science 222, 619–621. Barakat, A.O., Qian, Y., Kim, M., Kennicutt Ii, M.C., 2002. Compositional changes of aromatic steroid hydrocarbons in naturally weathered oil residues in the Egyptian western desert. Environ. Forensics 3, 219–225. Barron, M.G., Carls, M.G., Short, J.W., Rice, S.D., Heintz, R.A., Rau, M., Di Giulio, R., 2005. Assessment of the phototoxicity of weathered Alaska North Slope crude oil to juvenile pink salmon. Chemosphere 60, 105–110. Bernabeu, A.M., Fernández-Fernández, S., Bouchette, F., Rey, D., Arcos, A., Bayona, J.M., Albaiges, J., 2013. Recurrent arrival of oil to Galician coast: the final step of the Prestige deep oil spill. J. Hazard. Mater. 250–251, 82–90. Bobinger, S., Andersson, J.T., 2009. Photooxidation products of polycyclic aromatic compounds containing sulfur. Environ. Sci. Technol. 43, 8119–8125. Boukir, A., Aries, E., Guiliano, M., Asia, L., Doumenq, P., Mille, G., 2001. Subfractionation, characterization and photooxidation of crude oil resins. Chemosphere 43, 279–286. CEN, 2012. CEN/TR 15522-2:2012, Oil Spill Identification – Waterborne Petroleum and Petroleum Products – Part 2: Analytical Methodology and Interpretation of Results. Charrie-Duhaut, A., Lemoine, S., Adam, P., Connan, J., Albrecht, P., 2000. Abiotic oxidation of petroleum bitumens under natural conditions. Org. Geochem. 31, 977–1003. Diez, S., Sabate, J., Viñas, M., Bayona, J.M., Solanas, A.M., Albaiges, J., 2005. The Prestige oil spill. I. Biodegradation of a heavy fuel oil under simulated conditions. Environ. Toxicol. Chem. 24, 2203–2217. Diez, S., Jover, E., Bayona, J.M., Albaiges, J., 2007. Prestige oil spill. III. Fate of a heavy oil in the marine environment. Environ. Sci. Technol. 41, 3075–3082. Douglas, G.S., Owens, E.H., Hardenstine, J., Prince, R.C., 2002. The OSSA II pipeline oil spill: the character and weathering of the spilled oil. Spill Sci. Technol. Bull. 7, 135–148. Dowty, B.J., Brightwell, N.E., Laseter, J.L., Griffin, G.W., 1974. Dye sensitized photooxidation of phenanthrene. Biochem. Biophys. Res. Commun. 57, 452– 456. Dutta, T.K., Harayama, S., 2000. Fate of crude oil by the combination of photooxidation and biodegradation. Environ. Sci. Technol. 34, 1500–1505. Ehrhardt, M., Petrick, G., 1984. On the sensitized photo-oxidation of alkylbenzenes in seawater. Mar. Chem. 15, 47–58. Eiserbeck, C., Nelson, R.K., Grice, K., Curiale, J., Reddy, C.M., 2012. Comparison of GC– MS, GC–MRM–MS, and GC  GC to characterise higher plant biomarkers in Tertiary oils and rock extracts. Geochim. Cosmochim. Acta 87, 299–322. Fernandez-Varela, R., Gomez-Carracedo, M.P., Fresco-Rivera, P., Andrade, J.M., Muniategui, S., Prada, D., 2006. Monitoring photooxidation of the Prestige’s oil spill by attenuated total reflectance infrared spectroscopy. Talanta 69, 409–417. Fingas, M.F., 1995. A literature review of the physics and predictive modelling of oil spill evaporation. J. Hazard. Mater. 42, 157–175. Frontera-Suau, R., Bost, F.D., McDonald, T.J., Morris, P.J., 2002. Aerobic biodegradation of hopanes and other biomarkers by crude oil-degrading enrichment cultures. Environ. Sci. Technol. 36, 4585–4592. Garrett, R.M., Pickering, I.J., Haith, C.E., Prince, R.C., 1998. Photooxidation of crude oils. Environ. Sci. Technol. 32, 3719–3723. Garza-Gil, M.D., Prada-Blanco, A., Vazquez-Rodriguez, M.X., 2006. Estimating the short-term economic damages from the Prestige oil spill in the Galician fisheries and tourism. Ecol. Econ. 58, 842–849. GESAMP, 2007. Estimates of Oil Entering the Marine Environment from Sea-Based Activities, GESAMP Reports and Studies. IMO. Hansen, A.B., Daling, P.S., Faksness, L.-G., Sörheim, K.R., Kienhuis, P., Duus, R., 2007. 7 – Emerging CEN Methodology for Oil Spill Identification, Oil Spill Environmental Forensics. Academic Press, Burlington, pp. 229–256. Hostettler, F.D., Rosenbauer, R.J., Kvenvolden, K.A., 1999. PAH refractory index as a source discriminant of hydrocarbon input from crude oil and coal in Prince William Sound, Alaska. Org. Geochem. 30, 873–879. Jacquot, F., Guiliano, M., Doumenq, P., Munoz, D., Mille, G., 1996. In vitro photooxidation of crude oil maltenic fractions: evolution of fossil biomarkers and polycyclic aromatic hydrocarbons. Chemosphere 33, 671–681. Kamin´ski, M., Gudebska, J., Górecki, T., Kartanowicz, R.X., 2003. Optimized conditions for hydrocarbon group type analysis of base oils by thin-layer chromatography-flame ionisation detection. J. Chromatogr. A 991, 255–266. Lee, R.F., 2003. Photo-oxidation and photo-toxicity of crude and refined oils. Spill Sci. Technol. Bull. 8, 157–162. 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, 2123–2129. Liss, P.S., Liss, P.S., Duce, R.A., 2005. The Sea Surface and Global Change. Cambridge University Press.

277

MacDonald, I.R., 2002. Constraining rates of carbon flux from natural seeps on northern Gulf of Mexico Slope. In: Abstr. Vol. 7th Int. Conf. Gas in Marine Sediments, p. 119. Maki, H., Sasaki, T., Harayama, S., 2001. Photo-oxidation of biodegraded crude oil and toxicity of the photo-oxidized products. Chemosphere 44, 1145–1151. McCrea-Strub, A., Kleisner, K., Sumaila, U.R., Swartz, W., Watson, R., Zeller, D., Pauly, D., 2011. Potential impact of the Deepwater Horizon oil spill on commercial fisheries in the Gulf of Mexico. Fisheries 36, 332–336. McKenna, A.M., Nelson, R.K., Reddy, C.M., Savory, J.J., Kaiser, N.K., Fitzsimmons, J.E., Marshall, A.G., Rodgers, R.P., 2013. Expansion of the analytical window for oil spill characterization by ultrahigh resolution mass spectrometry: beyond gas chromatography. Environ. Sci. Technol. 47, 7530–7539. Moore, J.W., Ramamoorthy, S., 1984. Aromatic Hydrocarbons–Polycyclics, Organic Chemicals in Natural Waters: Applied Monitoring and Impact Assessment. Springer-Verlag, New York, NY, pp. 67–87. Neff, J.M., 1979. Polycyclic Aromatic Hydrocarbons in the Aquatic Environment: Sources, Fates and Biological Effects. Applied Science Publishers Ltd., Essex, England. Nicodem, D.E., Fernandes, M.C.Z., Guedes, C.L.B., Correa, R.J., 1997. Photochemical processes and the environmental impact of petroleum spills. Biogeochemistry 39, 121–138. Nicodem, D.E., Guedes, C.L.B., Fernandes, M.C.Z., Severino, D., Correa, R.J., Coutinho, M.C., Silva, J., 2001. Photochemistry of petroleum. Prog. React. Kinet. Mech. 26, 219–238. NOAA, 1992. An Introduction to Coastal Habitats and Biological Resources for Oil Spill Response. NOAA Hazardous Materials Response and Assessment Division, Seattle, WA. NRC, 2003. Oil in the Sea III: Inputs, Fates, and Effects. The National Academies Press, Washington DC. Payne, J.R., Phillips, C.R., 1985. Photochemistry of petroleum in water. Environ. Sci. Technol. 19, 569–579. Peters, K.E., Walters, C.C.C., Moldowan, J.J.M., 2005. The Biomarker Guide: I. Biomakers and Isotopes in the Environment and Human History. Cambridge University Press. Plata, D.L., Sharpless, C.M., Reddy, C.M., 2008. Photochemical degradation of polycyclic aromatic hydrocarbons in oil films. Environ. Sci. Technol. 42, 2432– 2438. Prince, R.C., Eimendorf, D.L., Lute, J.R., Hsu, C.S., Haith, C.E., Senius, J.D., Dechert, G.J., Douglas, G.S., Butler, E.L., 1994. 17a,21b(H)-hopane as a conserved internal marker for estimating the biodegradation of crude oil. Environ. Sci. Technol. 28, 142–145. Prince, R.C., 1988. Crude oil biodegradation. Encyclopedia Environ. Anal. Rem., 1327–1342. Prince, R.C., Garrett, R.M., Bare, R.E., Grossman, M.J., Townsend, T., Suflita, J.M., Lee, K., Owens, E.H., Sergy, G.A., Braddock, J.F., Lindstrom, J.E., Lessard, R.R., 2003. The roles of photooxidation and biodegradation in long-term weathering of crude and heavy fuel oils. Spill Sci. Technol. Bull. 8, 145–156. Radovic, J.R., Dominguez, C., Laffont, K., Diez, S., Readman, J.W., Albaiges, J., Bayona, J.M., 2012. Compositional properties characterizing commonly transported oils and controlling their fate in the marine environment. J. Environ. Monit. 14, 3220–3229. Reddy, C.M., Arey, J.S., Seewald, J.S., Sylva, S.P., Lemkau, K.L., Nelson, R.K., Carmichael, C.A., McIntyre, C.P., Fenwick, J., Ventura, G.T., Van Mooy, B.A.S., Camilli, R., 2012. Composition and fate of gas and oil released to the water column during the Deepwater Horizon oil spill. Proc. Nat. Acad. Sci. USA 109, 20229–20234. Schwarzenbach, R.P., Gschcwend, P.M., Imboden, D.M., 2005. Environmental Organic Chemistry, second ed. Wiley. Ventura, G.T., Raghuraman, B., Nelson, R.K., Mullins, O.C., Reddy, C.M., 2010. Compound class oil fingerprinting techniques using comprehensive twodimensional gas chromatography (GCxGC). Org. Geochem. 41, 1026–1035. Wang, Z., Fingas, M.F., 2003. Development of oil hydrocarbon fingerprinting and identification techniques. Mar. Pollut. Bull. 47, 423–452. Wang, Z., Yang, C., Fingas, M., Hollebone, B., Hyuk Yim, U., Ryoung Oh, J., 2007. 3 – Petroleum Biomarker Fingerprinting for Oil Spill Characterization and Source Identification, Oil Spill Environmental Forensics. Academic Press, Burlington, pp. 73–146. White, H.K., Hsing, P.Y., Cho, W., Shank, T.M., Cordes, E.E., Quattrini, A.M., Nelson, R.K., Camilli, R., Demopoulos, A.W.J., German, C.R., Brooks, J.M., Roberts, H.H., Shedd, W., Reddy, C.M., Fisher, C.R., 2012. Impact of the Deepwater Horizon oil spill on a deep-water coral community in the Gulf of Mexico. Proc. Nat. Acad. Sci. USA 109, 20303–20308. Yang, W.C., Wang, H., 1977. Modeling of oil evaporation in aqueous environment. Water Res. 11, 879–887.