pubblication #1

Characterization of a crude oil weathering series by ultrahigh-resolutionmass spectrometry using multiple ionization modes

Anna Katarina Huba, Piero R. Gardinali ⁎Department of Chemistry & Biochemistry, (SERC), Florida International University, 3000 NE 151 Street, Biscayne Bay Campus, North Miami, Florida 33181, USASoutheast Environmental Research Center (SERC), Florida International University, 3000 NE 151 Street, Biscayne Bay Campus, North Miami, Florida 33181, USA

H I G H L I G H T S

• A crude oil weathering series is ana-lyzed using ultrahigh-resolution massspectrometry.

• An increase in oxidation products is ob-served with increased weathering.

• Formation of acidic, ketonic and quinoniccompounds is suggested.

• Preferential ionizations of specific com-pounds may significantly skew results.

G R A P H I C A L A B S T R A C T

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

Article history:Received 29 January 2016Received in revised form 16 March 2016Accepted 16 March 2016Available online 17 May 2016

Editor: D. Barcelo

Accidental crude oil releases, such as the Deepwater Horizon (DWH) accident, are always a potential threat to pris-tinemarine ecosystems. Since the toxicity of crude oil heavily depends on its variable composition, the comprehen-sive characterization of crude oil compounds as a function of weathering is an important area of research.Traditional gas chromatography-based characterization presents significant limitations, and the use of ultrahigh-resolution mass spectrometric (UHRMS) techniques (that allow for the assignment of molecular formulae) hasbeen shown to be better equipped to address the complex nature of crude oils. This study used an Orbitrap QExactive mass spectrometer operated at a resolving power of 140,000 FWHM with both electrospray ionization(ESI) and atmospheric pressure photoionization (APPI) sources, in order to characterize a crude oil weathering se-ries of the Macondo oil released during the DWH incident (the source oil, two differently weathered surface slicks,and a beached residue). Preliminary gas chromatography mass spectrometry (GC–MS) and gas chromatographyflame ionization detection (GC–FID) results suggested that the four oils comprised a trueweathering series (includ-ing biodegradation and photodegradation in addition to other well-known processes such as dissolution and evap-oration). UHRMS results showeda clear increase in oxygenated compoundswithweathering, and further suggesteda significant gain of acidic compounds, aswell as the transformation of phenols to ketonic and quinonic compoundswith weathering. A complementary study on a weathered oil sample amended with selected model compoundscontributed additional insight into the functional group types that are accessible in each ionization technique.

© 2016 Published by Elsevier B.V.

Keywords:PetroleumDeepwater HorizonWeathering seriesUltrahigh-resolution mass spectrometryOxygenated hydrocarbons

Science of the Total Environment 563–564 (2016) 600–610

⁎ Corresponding author at: Department of Chemistry & Biochemistry and Southeast Environmental Research Center (SERC) Florida International University Biscayne Bay Campus 3000NE 151st Street Marine Science Building, MSB-356 North Miami, Florida, 33181.

http://dx.doi.org/10.1016/j.scitotenv.2016.03.2330048-9697/© 2016 Published by Elsevier B.V.

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

1. Introduction

In a society that is highly dependent on energy derived from petro-leum, constant crude oil extraction and transportation lead to virtuallyunavoidable releases into the environment (Wang et al., 2013). OnApril 20, 2010, a turbulent mixture of gas and a light, sweet (Macondotype) crude oil was released into the Gulf of Mexico as a result of theDeepwater Horizon (DWH) drilling rig incident in theMississippi CanyonBlock 252 (MC252) (Camilli et al., 2012; McNutt et al., 2012; Aeppliet al., 2012). In the days following the accident, some of the escapedMacondo oil rose 1500 m to the surface of the ocean, creating oil slickssome of which ultimately reached the coast (Liu et al., 2012). In theevent of crude oil discharges into marine environments, weatheringprocesses such as dissolution, dispersion, emulsification, evaporation,biodegradation, and photo-oxidation (Wang et al., 2013; Aeppli et al.,2012) constantly modify the oil composition (Wang et al., 2013;Jordan and Payne, 1980). Weathering, thus, adds compositional com-plexity to what is already one of the most complex natural mixturesknown to mankind (Hsu et al., 2011). The thousands of compoundspresent in crude oil can be divided into four main classes: saturated hy-drocarbons (straight, branched, and cyclic alkanes), aromatic com-pounds (containing one or more rings), resins (relatively highmolecular weight and polar compounds, which are soluble in the oil,and contain heteroatoms such as nitrogen, sulfur, and oxygen), andasphaltenes (highest molecular weight and most polar compounds,also containingheteroatoms such as nitrogen, oxygen and sulfur but un-like resins insoluble in the oil) (Garrett et al., 1998; Speight, 2004).These different compound types are affected differently by oilweathering as each of the weathering processes is selective towardsspecific compounds. For example, evaporation and emulsification de-plete the oil of its volatile compounds, while water washing removeswater-soluble compounds (Mansuy et al., 1997), and biodegradation af-fects primarily n-alkanes followed by branched and cyclic hydrocar-bons, closely followed by naphthenic compounds (Wang et al., 2013).Photo-oxidation has been shown to primarily modify select polycyclicaromatic hydrocarbons (PAHs), and in addition to other oxidationmechanisms (including biodegradation) to be responsible for the in-crease of the oxygen content in the remaining oil (McKenna et al.,2013; Prince et al., 2003). Since several constituents in crude oil havebeen associated with some degree of toxicity, the characterization offresh and especially of the modified weathered crude oil is crucial inorder to understand the potential environmental effects.

Conventional oil characterization by gas chromatography flameionization detection (GC–FID), or gas chromatography mass spectrom-etry (GC–MS) (Maki et al., 2001), has been able to elucidate weatheringrelated compositional changes of hydrocarbons and PAHs. However,coelution and the inability to analyze polar, nonvolatile, or thermallyunstable compounds (Wang et al., 2013; Garrett et al., 1998; McKennaet al., 2013; Burns, 1993; Charrie-Duhaut et al., 2000), limit the amountand type of analytes that can be detected (McKenna et al., 2013). Conse-quently, both techniques tend to overlook the analysis of oxygenatedhydrocarbons (Aeppli et al., 2012), which is particularly problematicfor weathered oil that has a lower amount of low-boiling and non-polar compounds that are GC amenable, and has a higher amount ofhigh-boiling and polar compounds (Aeppli et al., 2012). The character-istic oil “hump” or unresolved complex mixture (UCM), which is araised baseline due to the coelution of numerous compounds and is ob-served in GC analysis and is most prominent in weathered oils, is an in-dication of the limitations of these previously described techniques(Gough and Rowland, 1990). In order to resolve a larger fraction ofthe components present in oil, more advanced techniques need to beused, such as two-dimensional gas chromatographymass spectrometry(GC×GC–MS) or Fourier transform mass spectrometry (FT–MS).GC×GC–MS somewhat expands the accessible analytical window andallows for isomer differentiation (McKenna et al., 2013); however, it isstill limited to volatile compounds that are GC amenable (up to C45).

FT–MS, on the other hand, provides the possibility to extend thisrange up to C100 (McKenna et al., 2013), and to analyze nonvolatileand/or highly polar compounds (Qian et al., 2001a, 2001b; Mapoleloet al., 2009). Moreover, FT–MS analysis provides ultrahigh-resolution(Kaiser et al., 2011; Podgorski et al., 2013), and mass accuracy of lessthan 1 ppm (with internal calibration) (Savory et al., 2011), whichleads to the possibility of assigning elemental compositions (Rodgerset al., 2005). The ability to couple several different ionization techniquesto FT-MS is also crucial in the analysis of complex and diversified mix-tures such as crude oil. Common sources that have been used areelectrospray ionization (ESI) and atmospheric pressure photo ionization(APPI) (McKenna et al., 2013), but others such as atmospheric pressurechemical ionization (APCI) (Roussis and Fedora, 2002), and atmospher-ic pressure laser ionization (APLI) (Schrader et al., 2008) have also beenemployed. APPI is particularly useful to characterize nonpolar or slightlypolar species, while ESI offers the advantage of accessing more polarspecies while avoiding interferences of the hydrocarbonmatrix. As a re-sult, comprehensive characterizations of weathered oils must beachieved by a combination of multiple techniques, aimed at accessingthe maximum possible number of compounds.

A big portion of the previous studies on the Macondo oil releasedduring the DWH accident have focused mainly on common analytesthat are GC amenable, such as PAHs, alkanes, and hopane and steranebiomarkers (Aeppli et al., 2012; McKenna et al., 2013). Unsurprisingly,reports have shown that, with increasedweathering, the oil was deplet-ed ofmost of its saturated and aromatic compounds (Aeppli et al., 2012;Atlas and Hazen, 2011; Carmichael et al., 2012; Kostka et al., 2011; Limaet al., 2005; Liu et al., 2012). However, McKenna et al. (2013) estimatedthat in surface slicks only about 40% of the total mass of hydrocarbonscould be analyzed by conventional GC-based techniques, while Reddyet al. (2012) estimated that for weathered Macondo oil traditionalanalytes only account for less than 25% of the oil mass. Moreover, anincrease in oxygenated hydrocarbons with a concurrent decrease insaturated hydrocarbons and aromatics was reported in weathered oildeposited at the shoreline (Aeppli et al., 2012). Hall et al. (2013) furtherpredicted by GC×GC–MS analysis that this oxygenated fraction is large-ly due to the oxidation of saturates, which has only recently been shownto be a significant process during oil weathering (Hall et al., 2013). FT-ICR analysis of oiled sands has shown a similar trend, more specificallydetecting the possible formation of carboxylic acids, ketones, and alco-hols (Ruddy et al., 2014), all being consistent with photo-oxidationand biodegradation transformation products. However, a significantportion of the currently available knowledge originates from GC×GC–MS analysis, and thus only applies to a limited amount of compounds.Data currently available from ultrahigh-resolution mass spectrometry(UHRMS) is limited, which evidences a strong need to expand theknowledge on weathering products of the Macondo crude oil by FT–MS techniques. This study, therefore, aims to characterize and identifycompositional changes that occurred in a weathering series (freshcrude oil, two distinct oil slicks, and a beached oil mat) of the Macondocrude oil. UHRMS coupled with APPI and ESI in both positive and nega-tive ionization mode are used in order to expand the range and type ofcompounds that can be detected.

2. Materials and methods

2.1. Samples and preparation

Four different field-collected oils were characterized in this study.The unweathered Macondo oil (denoted as Massachusetts oil fromhereon)was collected by a production vessel on August 15, 2010 direct-ly at the MC-252 wellhead, and transferred to the Massachusetts oilbarge. Two weathered oils originating from two distinct surface slickswere skimmed from the Gulf of Mexico, and were collected by theUSCG Cutter Juniper and Barge No. CTC02404 on July 19, 2010 andJuly 29, 2010, respectively (referred to as Juniper and CTC oil from

601A.K. Huba, P.R. Gardinali / Science of the Total Environment 563–564 (2016) 600–610

now on). The last oil was buried in the shoreline of Elmer's Island (Lou-isiana), was exposed after hurricane Isaac, and was subsequently col-lected in August 2012 (denoted as Elmer's Island mat from hereon). A20,000 ppm oil stock solution was then created for the four oils by dis-solving approximately 1 g of crude oil in 50 mL of methylene chloride.The stock solutions were then diluted twofold to a final concentrationof 10,000 ppm for GC–FID and GC–MS analyses. GC–MS samples werespiked with 100 μL of a PAH surrogate standard mixture (naphtha-lene-d8, acenaphtene-d10, phenanthrene-d10, and perylene-d12), aswell as 100 μL of a PAH internal standard mixture (fluorene-d10 andbenzo(a)pyrene-d12). GC–FID samples, on the other hand, were spikedwith 100 μL of an aliphatic surrogate (n-dodecane-d26, n-eicosane-d42,n-triacontane-d62, p-terphenyl-d14) aswell as 100 μL of an aliphatic in-ternal standard (5α androstane and n-hexadecane-d34). For UHRMSanalysis 50 μL of oil stock was left to air dry, and was subsequentlyreconstituted into 50:50 toluene/methanol to a final concentration of2500, 5000, 5000, and 10,000 ppm for Massachusetts, CTC, Juniper,and Elmer's Island mat, respectively. The final solutions were spikedwith 1% formic acid, and 1% ammonium hydroxide for positive and neg-ative ionizationmode, respectively. An internal standard (tetradecanoic14,14,14-d3 acid, 11.6 ppm) was added to all UHRMS samples. Forthe model compound study, Elmer's Island mat was spiked withten standards covering a range of functional group types: phenol, 2-ethylphenol, 4-isopropylphenol, coprostane, coprostan-3-one, choles-terol, tetradecanoic 14,14,14,-d3 acid (all at approximately 10 ppm),coprostan-3-ol (1 ppm), tetracosanol (2 ppm), and tetracosanoic acid(5 ppm). All the solvents used were Optima LC/MS grade andwere pur-chased from Fisher Scientific (Fair Lawn, NJ, USA).

2.2. GC–FID analysis

GC–FIDanalysiswas carriedout on a ThermoTrace 1310GC-FID,fittedwith an Rxi®-5Sil fused silica capillary column (30 m × 0.25 mm ×

0.25 μm). A sample volume of 2 μL was injected (in splitless mode) intothe instrument. The inlet temperature was held at 325 °C, and the carriergas was set at a constant flow rate of 2.4 mL/min. The starting oven tem-peraturewas 40 °C, followed by an initial 7.5 °C/min ramp to 215 °C, and asecond 10 °C/min ramp to 320 °C, and then a final hold of 13 min.

2.3. GC–MS analysis

GC–MS analysis was carried out in electron impact mode (70eV) ona Thermo Finnigan Ultra trace TSQ Quantum XLC GC–MS operated inselected ion monitoring (SIM) mode. The GC–MS was fitted with anRxi®-5Sil fused silica capillary column (30 m × 0.25 mm × 0.25 μm),and helium was used as the carrier gas and set at a constant flow of1.7 mL/min. A sample volume of 2 μL was introduced (in splitlessmode) into the injector which was held at 300 °C. The initial oven tem-perature was 40 °C, followed by a 7.5 °C/min ramp to 295 °C, and aneight minute hold.

2.4. Ultrahigh-resolution mass spectrometric analysis

Analysiswas carried out on a Q Exactive Orbitrap (Thermo Scientific,NJ, USA) by direct infusion through a 500 μL syringe (Thermo Scientific,NJ, USA) at a typical flow rate of 30 μL/min. In addition to the acquisitionof the sample, each infusion data file contained acquisitions of a mobilephase background, and a solvent background. Datawere acquired in fullscan mode over a mass range of 80–1200 m/z, and the instrument wasoperated at a resolution of 140,000 FWHM. The automatic gain control(AGC) target was set to 1 × e6, while the maximum injection time wasset to 50 ms. External mass calibration provided a mass accuracy of5 ppm. The APPI ionization source (Thermo Scientific, NJ, USA) wasequipped with a krypton UV gas discharge lamp (Syagen Technology,Inc, Tustin, CA) that produces 10–10.2 eV photons (120 nm). N2 sheathgas at 40 psi was used to facilitate the ionization, while the auxiliary

Fig. 1. GC–FID chromatograms of the weathering series (Mass, CTC, Juniper, and Elmer's Island mat), highlighting the decrease of overall signal, as well as the disappearance of lowmolecular weight compounds and the formation of the UCM.

602 A.K. Huba, P.R. Gardinali / Science of the Total Environment 563–564 (2016) 600–610

port remained closed. The heated vaporizer region was held at 350 °C,while the capillary temperature was set to 300 °C, for both positiveand negative mode. For the ESI analysis a heated electrospray (HESI)source (Thermo Scientific, NJ, USA) was used, and typical conditionsfor positive mode were a spray voltage of 5.20 V, a heated vaporizer re-gion at 300 °C, capillary temperature of 300 °C, and sheath and auxiliarygas at 40 and 5 psi, respectively. For negative mode, the typical condi-tions were a spray voltage of 4.50 V, a heated vaporizer region at300 °C, capillary temperature of 200 °C, and sheath and auxiliary gasat 35 and 30 psi, respectively.

2.4.1. Data analysisMass spectrawere obtained by averaging a selected range of consec-

utively acquired infusion spectra. A background spectrum acquired inthe same infusion run as the sample was subtracted to account for ex-ternal contamination. Data processing was performed by using theComposer 1.0.6 software (Sierra Analytics, CA, USA), which relies onpetroleum specific composition assignment algorithms. Criteria usedfor peak detection and molecular formula assignments included: am/z range of 80–1000 Da, a match tolerance of 5 ppm for formulaassignments, a DBE range from −0.5 to 65, and element ranges ofC ≤ 200, H ≤ 1000, O ≤ 5,N ≤ 4, S ≤ 2. The setting for theminimumrelativepeak abundance accepted was sample specific in order to adjust for thevariable nature of the samples.

3. Results and discussion

3.1. GC-FID and GC–MS analysis, and weathering studies

Initially, the characterization of the weathering series was per-formed by visual inspection of the chromatograms obtained by GC-FIDanalysis. The characteristic trend in depletion of volatile compounds(predominantly alkanes, and to a lesser amount aromatics) for theweathered oils was observed and is shown in Fig. 1. A clear loss of allthe front-end compounds (which represent the low molecular weightand thus volatile hydrocarbons and aromatics), and an appearanceand increase of the characteristic oil UCM (which is due to hydrocarbonspecies that coelute in chromatographic analysis) is clearly noticeablewhen going from Massachusetts to the weathered oils (CTC, Juniper,and Elmer’s Island mat). The specific order of the oils in the weatheringseries is shown by the disappearance of more andmore alkanes, and anoverall decrease in signal. Massachusetts is shown to be a relativelyfresh oil, while out of the two surface slick oils Juniper appears to bemore weathered than CTC (which agrees with previous studies ontotal PAH depletion(BP, 2014), and the Elmer's Island mat is shown tobe the most weathered. Further characterization studies of the sameoils (shown in Fig. 2) were conducted in order to perform a more in-depth characterization of the weathering series based on processesother than dissolution and evaporation. GC–FID and GC–MS analysesand subsequent quantifications of specific alkanes and PAHs,

Fig. 2. Ratios of the concentrations of chrysene/benz(a)anthracene (top) and n-C18/phytane (bottom) for the weathering series, showing an increase in photodegradation andbiodegradation, respectively. The top ratio was obtained from GC–MS data, while the bottom ratio used GC–FID data.


respectively, showed results that reinforce the previously determinedweathering order, showing an increase in both biodegradation as wellas photodegradation going from Massachusetts, to CTC, to Juniper, tothe Elmer's Island mat. Fig. 2 illustrates these results, and shows plotsof two degradation ratios (chrysene/benz(a)anthracene (Plata et al.,

2008; Lemkau et al., 2010; Yim et al., 2011; Behymer and Hites, 1988)and n-C18/phytane (Lemkau et al., 2010; Yim et al., 2011; Wang et al.,1995a, 1998)) that have previously been used as good indicators ofcrude oil photodegradation and biodegradation, respectively. The ratioof chrysene/benz(a)anthracene increases for the weathering series,

Fig. 4. (−) APPI spectra of the oil weathering series, with the zoomed in spectra shown on the right emphasizing the shift towards higher molecular weight.

Fig. 3. (−) ESI spectra of the oil weathering series, highlighting the three distinct areas of major changes, and a relative increase of higher molecular weight compounds with weathering.


indicating an increase in photodegradation of the oils. On the otherhand, the n-C18/phytane ratio decreases, which is an indication ofincreased biodegradation. This strongly suggests that this is a trueweathering series (going from Massachusetts, to CTC, to Juniper, tothe Elmer's Island mat), and likely includes both photodegradationand biodegradation, in addition to other weathering processes such asdissolution and evaporation. An interesting fact that is noticeable isthat, based on the results here obtained, the degree of weatheringdoes not appear to be strictly time or location dependent, butmainly as-sociated to the oil’s path and the environmental factors related withit (such as temperature, nutrients, salinity, pH, sun incidence, andcurrents). Having a series of weathered oils is, therefore, essential inorder to achieve a more comprehensive understanding of the dynamicchanges that affect crude oil in amarine environment. This understand-ing is the fundamental basis to any oil toxicity estimations.

3.2. Ultrahigh-resolution mass spectrometric analysis

Since GC–FID and GC-MS analyses pose significant limitations in acomprehensive oil characterization, in order to expand the analyticalwindow of compounds detected, ultrahigh-resolution mass spectro-metric analysis of the four oils was performed by means of an OrbitrapQ Exactive instrument. The four oils were analyzed in both (±) ESI aswell as (±) APPI, in order to target a broader range of compounds

(polar and nonpolar). The resulting mass spectra show significantchanges, which can be seen particularly well in the negativemode spec-tra provided in Figs. 3 and 4. Overall, a common trend independent ofthe ionization source or mode, is the relative increase in highermolecu-larweight compounds. The (−) ESI spectra shown in Fig. 3, for example,clearly show this trend as the first section of the spectrum significantlydecreases, while the second section increases with weathering. More-over, a completely new series of compounds appears in the third sectionof the Elmer's Island mat spectrum, which illustrates the additionalcompositional complexity of beached oils (that may have incorporatedexogenous materials). Compared to the ESI data shown in Fig. 3, the(−) APPI data shown in Fig. 4 show less dramatic changes, but also ap-pear to present a slight shift towards higher molecular weight com-pounds, which is especially visible in the magnified spectra shown inthe inserts. Furthermore, it can be seen that the APPI spectra appear tocontain a much larger amount of individual masses. This evidences themore selective nature of the ESI ionization source towards more polarcompounds, as it eliminates the background hydrocarbon interferencethat dominates the APPI spectrum. This is a good illustration of howthese two ionization techniques are complementary andhowa compre-hensive crude oil characterization must use a combination of both.

In order to confirm these visually apparent changes, class distribu-tion plots were created by categorizing all the assigned molecular for-mulae (CvHwNxOySz) into specific heteroatom classes (O1, O2, NO, NO2,

Fig. 5. Class distributions for (+) APPI (top) and (+) ESI (bottom) for samples of the weathering series (Massachusetts, CTC, Juniper, Elmer's Islandmat). Full arrows depict a consistenttrend over the whole weathering series, while dashed arrows depict trends in which the Elmer’s Island mat is an exception. Compounds showing as protonated ions are denoted by the(H), others are radical ions.


etc.), and by plotting the relative abundance of each of these classes.These graphs are shown in Figs. 5 and 6, and show the presence oflarge compositional diversity among the four oils and the two ionizationmethods. Fig. 5 depicts the differences between ESI and APPI in positiveionization mode. The most dominant class detected in (+) APPI is thehydrocarbon class (protonated molecules are denoted by the (H)).(+) ESI, on the other hand, is dominated by nitrogen containing com-pounds, while the hydrocarbon portion is much smaller. Another inter-esting concept shown in these two plots is that while ESI ionizationrequires protonation, APPI provides the possibility to ionize otherspecies by charge transfer owing to the presence of dopant molecules(toluene). This creates radical compound classes for the hydrocarbon,nitrogen, and oxygen classes that become fairly prominent. Fig. 5 alsoshows a clear increase in oxygenated species (O1, O2, O3), with a con-current reduction in hydrocarbons as the weathering degree of theoils increases. Some of the classes of the Elmer's Island mat are an ex-ception to this trend, which could be due to the differing nature ofbeached oils. The increase in oxidationwithweathering that is observedcorroborates data from other studies (Aeppli et al., 2012; Hall et al.,2013; Ruddy et al., 2014). Negativemode data (shown in Fig. 6) extendsthe compositional coverage to highly oxygenated species (O4 and O5).However, unlike for positive mode where all the oxygenated classes in-creasedwithweathering, in negativemode there seems to be a decreasein lower oxygenated species (O1 for APPI, and O1 and O2 for ESI) with a

concurrent increase in higher oxygenated species (O2–O5 for APPI, andO3–O5 for ESI).

In order to more clearly depict the changes in hydrocarbons and ox-ygenated compounds, Kendrick mass defect plots were created(Kendrick mass = IUPAC mass×(14.00000/14.01565)). Such graphsplot the Kendrick Mass Defect vs. the Nominal Kendrick Mass (differ-ence between the nominal and the exact Kendrick masses), and canbe used in order to simplify the visualization of data originating fromcomplex matrices, and to better visualize compositional changes. Anexample of such plots is given in Fig. 7 for (−) APPI data, and shows asignificant increase in oxygenated hydrocarbons (O1 to O5), with anespecially large increase in compounds in the mid-mass range (m/z300–500) for weathered oils. This correlated to a reduction of the hy-drocarbon component, which mainly lost its higher molecular weight(m/z 350 and up) compounds.

The results so far have shown an overall increase in oxygenated hy-drocarbons (mostly in the mid to high molecular weight range), buthave given little insight into the changes occurring within the specificoxygen classes. Double bond equivalent (DBE) plots were created byplotting the DBEs (number of rings and double bonds) versus thecarbon number, in order to visualize changes happening in individualclasses and get a better understanding on the saturation level of thecompounds involved. The (+) APPI plots are shown in Fig. 8 as an ex-ample of the results that were obtained. The appearance of oxygenated

Fig. 6. Heteroatom class distributions for (−) APPI (top) and (−) ESI (bottom) of the weathering series (Massachusetts, CTC, Juniper, Elmer's Islandmat). Full arrows depict a consistenttrend over the whole weathering series, while dashed arrows depict trends in which the Elmer's Island mat is an exception. Compounds showing as protonated ions are denoted by the(H), others are radical ions.


species that was previously observed with weathering is mostly con-firmed by these plots; moreover, it becomes evident that for (+) APPImode the newly formed or enriched oxygenated compounds aremostlyunsaturated or aromatic compounds (DBE 5-15) with 15–40 carbons.ESI and negative ionization mode results (not shown) have shown sim-ilar ranges, and since PAHs are compounds that fall in that range, these

results may suggest that PAHs and their derivatives could be a signifi-cant portion of the compounds that undergo oxidation during theweathering process.

This possibility is confirmed by the results of Fig. 9, which shows aVan Krevelen diagram (that plots H/C versus O/C and indicatesunsaturation and oxidation, respectively) for the unweathered

Fig. 8. DBE vs. carbon number for the O1–O3 containing hydrocarbons detected in the weathering series in (+) APPI mode.

Fig. 7. Kendrick Mass Defect (KMD) vs. Nominal Kendrick Mass contour plots for oxygen containing hydrocarbons (left) and hydrocarbons (right) obtained using (−) APPI conditions.


Massachusetts oil (top) and the weathered Juniper oil (bottom) in(+) APPI mode. This plot depicts all the peaks that were assigneda molecular formula containing at least one oxygen, and evidencesa drastic increase in the number and in the relative intensity of oxygenat-ed compounds present in the weathered oil. The areas of major changes(shown by the red rectangles) contain compounds with a H/C in therange of 0.5–1.5. Completely saturated hydrocarbons would have a H/Cratio of 2, while completely aromatic species would have a H/C of b1(with benzene starting at 1, and the H/C decreasingwith increasing num-ber of rings, so that chrysenewould have a H/C of about 0.67). The natureof the compounds whose detected ions are enhanced is therefore eithercompletely aromatic in nature (when H/C b1), or contain some kind ofunsaturation and aromaticity if they fall in the H/C between 1 and 2.

All these different types of plots are essential in visualizing gen-eral trends in data sets containing thousands of assigned compounds.However, they do not provide unequivocal information on the functionalgroup types of the molecules. Some information regarding what specific

types ofmolecules are present can be inferred from ionization studies elu-cidating selective ionization mechanisms or preferential ionization of in-dividual heteroatoms. Figs. 5 and 6, show how the relative abundancesof the different oxygen classes vary based on the ionization source andmode, and suggest that different functional group types may be involved.For example, in negative ionizationmode (Fig. 6) forweathered oils thereis a substantial prevalence of O2 species with respect to O1 species. Thisagrees with previous reports (Mapolelo et al., 2009, 2011; Ruddy et al.,2014), and has been attributed to a preferential ionization of carboxylicacid species that can be easily deprotonated. For the unweathered oilthis is not always true as in (−) APPI the O1 species represent a larger rel-ative fraction compared to the O2 species. This could eithermean that theO1 fraction (such as alcohols, phenols, ketones) is large enough to domi-nate a preferential ionization of acidic O2 species, or that there are singlyoxygenated compounds mostly present in the unweathered oil that aresimilarly well ionized as the carboxylic acids. In positive ionizationmode, on the other hand, there is a prevalence of O1 over O2 compounds

Fig. 10. (−) APPI spectrum of the Elmer's Islandmat fortified with several individual model compounds used to test ionization efficiency, and Kendrick mass defect plot of the O1 and O2

classes of the same sample showing the corresponding detection and correct assignment of the model compounds.

Fig. 9. Van Krevelen plots of source (Massachusetts) and weathered (Juniper) oils obtained in (+) APPI mode, showing the relative increase of aromatic and unsaturated oxidationproducts.


(shown in Fig. 5), which ismost likely due to the preferential ionization ofsomeO1 species (such as alcohols, phenols, or ketones)with respect to O2

compounds.In order to gain a better understanding of the nature of specific oxy-

genated classes and the significance of these results, a systematic ioniza-tion studywas conducted by spiking several compounds into an Elmer'sIsland mat sample. These model compounds spanned a wide range offunctional group types, including: hydrocarbons, phenols, alcohols, ke-tones, and acids. The sample infused in APPI positive ionizationmode fa-vored the formation of the ketone functional group (coprostan-3-one),while phenols and hydrocarbonswere onlyweakly ionized, and straightchain alcohols and acids showed no ionization. Negative ionizationmode, on the other hand, preferentially ionized the acids and phenols,while none of the other compounds were detected. The mass spectrumand Kendrick Mass Defect plot of the (−) APPI data for the Elmer's Is-land mat are presented in Fig. 10, and show the phenols (in blue) andthe acid (in pink) that were ionized and correctly assigned. ESI resultsfrom the same spiked sample (results not shown) mostly corroboratedthe same preferential ionizations (ketones and to a lesser extent phe-nols in positive mode, and carboxylic acids and phenols in negativemode). These results clearly show that both the APPI and ESI sourcesionize only certain functional group types (and out of those somemuch better than others) depending on the ionization mode, and thishas to be accounted for when interpreting heteroatom class assign-ments plots from high-resolution mass spectrometric analyses. Thepreferential ionization of singly oxygenated ketones in positive ioniza-tion mode with respect to doubly oxygenated carboxylic acids suggeststhat ketones could be a significant portion of theO1 class that dominates(+) ESI and (+) APPI generated spectra, and this seems to corroborateprevious reports that have suggested ketones as oxidation products incrude oil weathering (Ruddy et al., 2014). Phenols have shown to besomewhat ionized and could therefore also contribute to the O1 class,while the lack of ionization of straight chain alcohols makes them anunlikely contributor. The O2 class (and other higher oxygen classes)may be combinations of functional group types, andmost likely containwell ionizable groups such as ketones. In negativemode, acids are beingpreferentially ionized by deprotonation, and are thus the most likelycontributor to the large relative fraction of the O2 classes, and theacidification of crude oil compounds with weathering and especiallybiodegradation has been previously reported (Charrie-Duhaut et al.,2000; Ruddy et al., 2014; Watson et al., 2002). Moreover, based on theH/C and DBE values detected, it is clear that the possible acidic fractionconsist of both an unsaturated and aromatic hydrocarbon backbone.The model compound study has also shown that phenols, in additionto acids, are a class that is well ionized in negative ionization mode,and the abundant O1 class in the source oil could be largely due to phe-nolic compounds, which have been proposed as intermediates inphotodegradation of PAHs (McConkey et al., 2002; Wang et al., 1995b;Chen et al., 2006; Kong and Ferry, 2003). These compounds have beenshown to undergo further photo-oxidation, being converted to ketonicand quinonic compounds (McConkey et al., 2002; Wang et al., 1995b;Chen et al., 2006; Kong and Ferry, 2003). This could explain why theO1 class in negative ionization mode decreases (oppositely to all otheroxygenated classes), as singly oxygenated phenols are converted todoubly oxygenated quinones (see Fig. 6). GC×GC-TOF data on thesame weathering series show the enrichment of straight chain ketonesand acids and the depletion of phenols, strengthening the conclusionson possible functional group types that are being transformed duringthe weathering process (Ding and Gardinali, 2015). The higher oxygen-ated fractions (O2 and up) are likely combinations of functional grouptypes and contain the well ionized carboxylic acid and phenolic groups.The ionization study has also illustrated that since straight chain alco-hols were not (or very poorly) ionized in any of the ionization tech-niques, they are not likely to give a significant contribution to anyobserved O1 class. This part of the study, overall, emphasized the needto gain more in-depth knowledge on the ionization of crude oil

compounds, and offered valuable insight into the possible functionalgroup types that are making up the O1, O2 and higher oxygenated hy-drocarbon classes that exhibit significant changes with weathering.

4. Conclusion

This study presented thefirst ultrahigh-resolutionmass spectromet-ric characterization of an oil weathering series, including the freshMacondo oil, two differently weathered surface slick oils, and a beachedoil tar. Preliminary GC–MS and GC–FID studies have confirmed the fouroils to be a true weathering series, and ratios of nC18/phytane andchrysene/benz(a)anthracene have further shown the oils to be likelybiodegradation and photodegradation series. Studying a completeweathering sequence provides the opportunity to achieve a better un-derstanding of the type of weathering processes that were most signif-icant in the DWH oil release, and consequently how these mechanismsaffected the composition of the oil. Ultrahigh-resolution results fromthis study have shown an increase in oxygenated compounds as theMacondo oil weathered, additionally suggesting a gain of ketones,quinones, and acidic compounds, with a concurrent decrease in pheno-lic compounds. The separate ionization study that was conducted byspikingmodel compounds into an oil sample also helped put the resultsinto a new perspective and further point out serious defects in currentinterpretations, as results from this study clearly showed how out ofthe nine spiked compounds some were not ionized while others werefully ionized. This proves that compound class assignment plots mighthave relative intensities largely skewed by preferential ionization,while some compounds could be abundant but poorly ionized andthus be underestimated. Future work will include a more extensivestudy of ionization mechanisms of several crude oil model compounds,spanning a wide size range, and including more functional groups,as well as heteroatom containing compounds. Results of such a studywill be necessary in order to make more conclusive compound assign-ments. This will ultimately help to expand the understanding of thetype of weathering processes that have played a significant role in theDeepwater Horizon accident, which will be fundamental in evaluatingthe long-term fate and toxicity of the oil that was released.

Acknowledgments

This work was supported by BP Exploration & Production Inc. andthe BP Gulf Coast Restoration Organization through FIU project800001596. This is contribution number 999 from the Southeast Envi-ronmental Research Center (SERC) at Florida International University.

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 horizondisaster led to the formation of oxygenated residues. Environ. Sci. Technol. 46 (16),8799–8807. http://dx.doi.org/10.1021/es3015138.

Atlas, R.M., Hazen, T.C., 2011. Oil biodegradation and bioremediation: a tale of the twoworst spills in US history. Environ. Sci. Technol. 45 (16), 6709–6715. http://dx.doi.org/10.1021/es2013227.

Behymer, T.D., Hites, R.A., 1988. Photolysis of polycyclic aromatic-hydrocarbons adsorbedon fly-ash. Environ. Sci. Technol. 22 (11), 1311–1319. http://dx.doi.org/10.1021/es00176a011.

BP, 2014. Gulf Science Data Reference Oil Characterization Data In Website: http://gulfsciencedata.bp.com/, directory: Oil; subdirectory: Oil Characteristics – additionalreference oils; filename: OilChemistry_O-04v01-03.zip (Last modified January 22,2014).

Burns, K.A., 1993. Evidence for the importance of including hydrocarbon oxidation-products in environmental assessment studies. Mar. Pollut. Bull. 26 (2), 77–85.http://dx.doi.org/10.1016/0025-326x(93)90095-2.

Camilli, R., Di Iorio, D., Bowen, A., Reddy, C.M., Techet, A.H., Yoerger, D.R., Whitcomb, L.L.,Seewald, J.S., Sylva, S.P., Fenwick, J., 2012. Acoustic measurement of the DeepwaterHorizon Macondo well flow rate. Proc. Natl. Acad. Sci. U. S. A. 109 (50),20235–20239. http://dx.doi.org/10.1073/pnas.1100385108.

Carmichael, C.A., Arey, J.S., Graham,W.M., Linn, L.J., Lemkau, K.L., Nelson, R.K., Reddy, C.M.,2012. Floating oil-covered debris from Deepwater Horizon: identification and appli-cation. Environ. Res. Lett. 7 (1). http://dx.doi.org/10.1088/1748-9326/7/1/015301.


Charrie-Duhaut, A., Lemoine, S., Adam, P., Connan, J., Albrecht, P., 2000. Abiotic oxidationof petroleum bitumens under natural conditions. Org. Geochem. 31 (10), 977–1003.http://dx.doi.org/10.1016/s0146-6380(00)00109-1.

Chen, J., Ehrenhauser, F.S., Valsaraj, K.T., Wornat, M.J., 2006. Uptake and UV-photooxidation of gas-phase PAHs on the surface of atmospheric water films. 1.Naphthalene. J. Phys. Chem. A 110 (29), 9161–9168. http://dx.doi.org/10.1021/jp062860b.

Ding, Y., Gardinali, P.R., 2015. Application of silica gel chromatography coupled withGCxGC-TOF in identification of oxygenated hydrocarbons in complex environmentalsamples. 39th International Symposium on Capillary Chromatography & 12th GCxGCSymposium, Fort Worth, TX, USA http://www.isccgcxgc2015.com/.

Garrett, R.M., Pickering, I.J., Haith, C.E., Prince, R.C., 1998. Photooxidation of crude oils. En-viron. Sci. Technol. 32 (23), 3719–3723. http://dx.doi.org/10.1021/es980201r.

Gough, M.A., Rowland, S.J., 1990. Characterization of unresolved complex-mixtures of hy-drocarbons in petroleum. Nature 344 (6267), 648–650. http://dx.doi.org/10.1038/344648a0.

Hall, G.J., Frysinger, G.S., Aeppli, C., Carmichael, C.A., Gros, J., Lemkau, K.L., Nelson, R.K.,Reddy, C.M., 2013. Oxygenated weathering products of Deepwater Horizon oilcome from surprising precursors. Mar. Pollut. Bull. 75 (1-2), 140–149. http://dx.doi.org/10.1016/j.marpolbul.2013.07.048.

Hsu, C.S., Hendrickson, C.L., Rodgers, R.P., McKenna, A.M., Marshall, A.G., 2011.Petroleomics: advanced molecular probe for petroleum heavy ends. J. MassSpectrom. 46 (4), 337–343. http://dx.doi.org/10.1002/jms.1893.

Jordan, R.E., Payne, J.R., 1980. Fate and Weathering of Petroleum Spills in the Marine En-vironment: A Literature Review and Synopsis. Ann Arbor Science. Publ.

Kaiser, N.K., Savory, J.J., McKenna, A.M., Quinn, J.P., Hendrickson, C.L., Marshall, A.G., 2011.Electrically compensated Fourier transform ion cyclotron resonance cell for complexmixture mass analysis. Anal. Chem. 83 (17), 6907–6910. http://dx.doi.org/10.1021/ac201546d.

Kong, L., Ferry, J.L., 2003. Effect of salinity on the photolysis of chrysene adsorbed to asmectite clay. Environ. Sci. Technol. 37 (21), 4894–4900. http://dx.doi.org/10.1021/es026124o.

Kostka, J.E., Prakash, O., Overholt, W.A., Green, S.J., Freyer, G., Canion, A., Delgardio, J.,Norton, N., Hazen, T.C., Huettel, M., 2011. Hydrocarbon-degrading bacteria and thebacterial community response in Gulf of Mexico beach sands impacted by the Deep-water Horizon oil spill. Appl. Environ. Microbiol. 77 (22), 7962–7974. http://dx.doi.org/10.1128/aem.05402-11.

Lemkau, K.L., Peacock, E.E., Nelson, R.K., Ventura, G.T., Kovecses, J.L., Reddy, C.M., 2010.The M/V Cosco Busan spill: source identification and short-term fate. Mar. Pollut.Bull. 60 (11), 2123–2129. http://dx.doi.org/10.1016/j.marpolbul.2010.09.001.

Lima, A.L.C., Farrington, J.W., Reddy, C.M., 2005. Combustion-derived polycyclic aromatichydrocarbons in the environment — a review. Environ. Forensic 6 (2), 109–131.http://dx.doi.org/10.1080/15275920590952739.

Liu, Z.F., Liu, J.Q., Zhu, Q.Z., Wu, W., 2012. The weathering of oil after the DeepwaterHorizon oil spill: insights from the chemical composition of the oil from the sea sur-face, salt marshes and sediments. Environ. Res. Lett. 7 (3). http://dx.doi.org/10.1088/1748-9326/7/3/035302.

Maki, H., Sasaki, T., Harayama, S., 2001. Photo-oxidation of biodegraded crude oil and tox-icity of the photo-oxidized products. Chemosphere 44 (5), 1145–1151. http://dx.doi.org/10.1016/s0045-6535(00)00292-7.

Mansuy, L., Philp, R.P., Allen, J., 1997. Source identification of oil spills based on the isotopiccomposition of individual components in weathered oil samples. Environ. Sci. Technol.31 (12), 3417–3425. http://dx.doi.org/10.1021/es970068n.

Mapolelo, M.M., Stanford, L.A., Rodgers, R.P., Yen, A.T., Debord, J.D., Asomaning, S.,Marshall, A.G., 2009. Chemical speciation of calcium and sodium naphthenatedeposits by electrospray ionization FT-ICR mass spectrometry. Energy Fuel 23 (1),349–355. http://dx.doi.org/10.1021/ef800642b.

Mapolelo, M.M., Rodgers, R.P., Blakney, G.T., Yen, A.T., Asomaning, S., Marshall, A.G., 2011.Characterization of naphthenic acids in crude oils and naphthenates by electrasprayionization FT-ICR mass spectrometry. Int. J. Mass Spectrom. 300 (2-3), 149–157.http://dx.doi.org/10.1016/j.ijms.2010.06.005.

McConkey, B.J., Hewitt, L.M., Dixon, D.G., Greenberg, B.M., 2002. Natural sunlight inducedphotooxidation of naphthalene in aqueous solution. Water Air Soil Pollut. 136 (1-4),347–359. http://dx.doi.org/10.1023/a:1015223806182.

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 spillcharacterization by ultrahigh resolution mass spectrometry: beyond gas chromatog-raphy. Environ. Sci. Technol. 47 (13), 7530–7539. http://dx.doi.org/10.1021/es305284t.

McNutt, M.K., Camilli, R., Crone, T.J., Guthrie, G.D., Hsieh, P.A., Ryerson, T.B., Savas, O.,Shaffer, F., 2012. Review of flow rate estimates of the Deepwater Horizon oil spill.Proc. Natl. Acad. Sci. U. S. A. 109 (50), 20260–20267. http://dx.doi.org/10.1073/pnas.1112139108.

Plata, D.L., Sharpless, C.M., Reddy, C.M., 2008. Photochemical degradation of polycyclic ar-omatic hydrocarbons in oil films. Environ. Sci. Technol. 42 (7), 2432–2438. http://dx.doi.org/10.1021/es702384f.

Podgorski, D.C., Corilo, Y.E., Nyadong, L., Lobodin, V.V., Bythell, B.J., Robbins, W.K.,McKenna, A.M., Marshall, A.G., Rodgers, R.P., 2013. Heavy petroleum composition.5. Compositional and structural continuum of petroleum revealed. Energy Fuel 27(3), 1268–1276. http://dx.doi.org/10.1021/ef301737f.

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 rolesof photooxidation and biodegradation in long-term weathering of crude and heavyfuel oils. Spill Sci. Technol. Bull. 8 (2), 145–156. http://dx.doi.org/10.1016/s1353-2561(03)00017-3.

Qian, K., Rodgers, R.P., Hendrickson, C.L., Emmett, M.R., Marshall, A.G., 2001a. Readingchemical fine print: resolution and identification of 3000 nitrogen-containingaromatic compounds from a single electrospray ionization Fourier transform ion cy-clotron resonance mass spectrum of heavy petroleum crude oil. Energy Fuel 15 (2),492–498. http://dx.doi.org/10.1021/ef000255y.

Qian, K.N., Robbins, W.K., Hughey, C.A., Cooper, H.J., Rodgers, R.P., Marshall, A.G., 2001b.Resolution and identification of elemental compositions for more than 3000 crudeacids in heavy petroleum by negative-ion microelectrospray high-field Fourier trans-form ion cyclotron resonance mass spectrometry. Energy Fuel 15 (6), 1505–1511.http://dx.doi.org/10.1021/ef010111z.

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. Composi-tion and fate of gas and oil released to the water column during the DeepwaterHorizon oil spill. Proc. Natl. Acad. Sci. U. S. A. 109 (50), 20229–20234. http://dx.doi.org/10.1073/pnas.1101242108.

Rodgers, R.P., Schaub, T.M., Marshall, A.G., 2005. Petroleomics: MS returns to its roots.Anal. Chem. 77 (1), 20A–27A. http://dx.doi.org/10.1021/ac053302y.

Roussis, S.G., Fedora, J.W., 2002. Quantitative determination of polar and ionic compoundsin petroleum fractions by atmospheric pressure chemical ionization and electrosprayionization mass spectrometry. Rapid Commun. Mass Spectrom. 16 (13), 1295–1303.http://dx.doi.org/10.1002/rcm.714.

Ruddy, B.M., Huettel, M., Kostka, J.E., Lobodin, V.V., Bythell, B.J., McKenna, A.M., Aeppli, C.,Reddy, C.M., Nelson, R.K., Marshall, A.G., Rodgers, R.P., 2014. Targeted petroleomics:analytical investigation of Macondo well oil oxidation products from PensacolaBeach. Energy Fuel 28 (6), 4043–4050. http://dx.doi.org/10.1021/ef500427n.

Savory, J.J., Kaiser, N.K., McKenna, A.M., Xian, F., Blakney, G.T., Rodgers, R.P., Hendrickson,C.L., Marshall, A.G., 2011. Parts-per-billion Fourier transform ion cyclotron resonancemass measurement accuracy with a “walking” calibration equation. Anal. Chem. 83(5), 1732–1736. http://dx.doi.org/10.1021/ac102943z.

Schrader, W., Panda, S.K., Brockmann, K.J., Benter, T., 2008. Characterization of non-polararomatic hydrocarbons in crude oil using atmospheric pressure laser ionization andFourier transform ion cyclotron resonance mass spectrometry (APLI FT-ICR MS).Analyst 133 (7), 867–869. http://dx.doi.org/10.1039/b801031e.

Speight, J.G., 2004. Petroleum asphaltenes - Part 1— asphaltenes, resins and the structureof petroleum. Oil Gas Sci. Technol. Rev. Ifp Energ. Nouvelles 59 (5), 467–477. http://dx.doi.org/10.2516/ogst:2004032.

Wang, Z.D., Fingas, M., Sergy, G., 1995a. Chemical characterization of crude-oil residuesfrom an arctic beach by GC/MS and GC/FID. Environ. Sci. Technol. 29 (10),2622–2631. http://dx.doi.org/10.1021/es00010a025.

Wang, C.X., Yediler, A., Peng, A., Kettrup, A., 1995b. Photodegradation of phenanthrene inthe presence of humic substances and hydrogen-peroxide. Chemosphere 30 (3),501–510. http://dx.doi.org/10.1016/0045-6535(94)00413-o.

Wang, Z.D., Fingas, M., Blenkinsopp, S., Sergy, G., Landriault, M., Sigouin, L., Foght, J.,Semple, K., Westlake, D.W.S., 1998. Comparison of oil composition changes dueto biodegradation and physical weathering in different oils. J. Chromatogr. A 809(1-2), 89–107. http://dx.doi.org/10.1016/s0021-9673(98)00166-6.

Wang, C.Y., Chen, B., Zhang, B.Y., He, S.J., Zhao, M.M., 2013. Fingerprint and weatheringcharacteristics of crude oils after Dalian oil spill, China. Mar. Pollut. Bull. 71 (1-2),64–68. http://dx.doi.org/10.1016/j.marpolbul.2013.03.034.

Watson, J.S., Jones, D.M., Swannell, R.P.J., van Duin, A.C.T., 2002. Formation of carboxylicacids during aerobic biodegradation of crude oil and evidence of microbial oxidationof hopanes. Org. Geochem. 33 (10), 1153–1169. http://dx.doi.org/10.1016/s0146-6380(02)00086-4.

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 the HebeiSpirit oil spill. J. Hazard. Mater. 197, 60–69. http://dx.doi.org/10.1016/j.jhazmat.2011.09.055.


pubblication #1

Documents