Development and application of an analytical method using...

Research Article

Received: 18 October 2013 Revised: 4 February 2014 Accepted: 5 February 2014 Published online in Wiley Online Library

Rapid Commun. Mass Spectrom. 2014, 28, 948–956

948

Development and application of an analytical method using gaschromatography/triple quadrupole mass spectrometry forcharacterizing alkylated chrysenes in crude oil samples

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

RATIONALE: Recent advances in analytical techniques have led to the development of gas chromatography/triplequadrupole mass spectrometry methods that allow the identification of target analytes in complex environmentalsamples. We have employed this technology to develop a method for characterizing alkylated chrysenes, which areenvironmental toxins that are resistant to weathering.METHODS: An Agilent 7890 gas chromatograph coupled to an Agilent 7000B triple quadrupole mass spectrometer wasused. The mass spectral fragmentation of seven commercially available alkylated chrysene standards was studied underfull-scan and product-ion scan conditions. The calibration curves used were in the linear range with r2 values greaterthan 0.99. The recovery and limit of detection of target analytes in the samples were in the range of 80–120% and0.11–1.09 ng/mL, respectively.RESULTS: The information inferred from full-scan and product-ion scan data was combined with literature data todevelop a GC/MS/MS method for the identification and quantification of C1-, C2-, C3-, and C4-chrysene homologues.The method was employed to characterize MC252 crude oil which was released during the 2010 Deepwater Horizonaccident. The results showed that the chrysene concentrations estimated by the proposed method were well within therange of previously reported values.CONCLUSIONS: The proposed method is useful for analyzing chrysene and its alkylated homologues in crude oilsamples. Copyright © 2014 John Wiley & Sons, Ltd.

(wileyonlinelibrary.com) DOI: 10.1002/rcm.6868

Polycyclic aromatic hydrocarbons (PAHs) and their alkylatedhomologues are a class of common organic environmentalcontaminants that contain two or more aromatic rings.[1] Theyare naturally found in organic substances such as crude oil,coal and creosote. They are also formed during the incompletecombustion of coal, oil, gas, or other organic substances.[2,3] Itis well established that alkylated homologues of certain PAHs(known as alkylated PAHs) can be present in high abundancein environmental samples, and, in some cases, can be moretoxic than their parent PAHs.[4,5] In addition, certain groupsof alkylated PAHs in crude oil released into the marineenvironment (e.g., alkylated chrysenes) are highly resistantto natural weathering processes such as volatilization,photooxidation, and biodegradation.[6–8] Furthermore, dueto their low aqueous solubility, the relative concentrations ofsome of these PAHs can increase over time under certainenvironmental conditions.[9] These properties make chryseneand alkylated chrysenes excellent indicator compounds forassessing the impacts of large-scale environmental eventssuch as oil spills.In this study we focus on four homologues of alkylated

chrysenes (C1- to C4-chrysenes). In general, alkylated chrysenesare commonly estimated using gas chromatography/mass

* Correspondence to: T. P. Clement, Department of CivilEngineering, Auburn University, Auburn, AL 36849, USA.E-mail: [email protected]

Rapid Commun. Mass Spectrom. 2014, 28, 948–956

spectrometry (GC/MS).[7,10,11] The efficiency of GC/MSmethods depends on the level of interference from the samplematrix, and lower detection limits can be achieved only afterextensive sample clean-up procedures prior to analysis.[10]

Recent advances in mass spectrometry have led to thedevelopment of triple quadrupole mass spectrometry (or GC/MS/MS) that allows the identification of low concentrationtarget analytes in complex samples with greater certainty byminimizing or removing matrix interferences.[12] This isachieved by monitoring multiple reactions between precursorions from the electron ionization process and characteristicproduct ions from collision cell reactions of these precursorions. A comparison study carried out by Fernández-Gonzálezet al.[10] has shown that GC/MS/MS methods can avoidinterferences in biota samples, and can also yield lowerdetection limits. Pitarch et al.[12] quantified various semi-volatile organic compounds in water samples and concludedthat GC/MS/MS methods can lead to excellent selectivityand sensitivity.

Currently, there are no published GC/MS/MS methodsavailable for accurately identifying and quantifying differenttypes of alkylated chrysenes present in environmentalsamples. The objective of this project was to develop a robustanalytical method using gas chromatography/triplequadrupole mass spectrometry to characterize alkylatedchrysenes in crude oil and/or weathered oil spill samples. Anew GC/MS/MS instrument was first used to study the mass

Copyright © 2014 John Wiley & Sons, Ltd.

GC/MS/MS method for characterizing alkylated chrysenes in crude oil

spectral fragmentation patterns of several commerciallyavailable alkylated chrysenes standards using full-scan andproduct-ion scan modes. These experimental data were thenused to predict the fragmentation patterns of other isomers ofchrysene homologues for which standards were not available.The experimental and theoretical fragmentation data were thencombined to develop a GC/MS/MS method for identifyingand quantifying the total concentrations of C1-, C2-, C3-, andC4-chrysene homologues in crude oil samples. The developedmethodwas tested in the characterization of alkylated chrysenesin MC252 crude oil which was spilled into the Gulf of Mexicowaters during the 2010 Deepwater Horizon oil spill event.

EXPERIMENTAL

Materials

High-purity standards of chrysene (purity >98%), C1-chrysenes(1-methylchrysene, 2-methylchrysene, 3-methylchrysene), C2-chrysene (6-ethylchrysene), C3-chrysenes (6-n-propylchrysene,1,3,6-trimethylchrysene) and C4-chrysene (6-n-butylchrysene)were purchased from Chiron AS (Trondheim, Norway). Forstructures, see Fig. 1. All alkylated chrysene homologuestandards were of purity greater than 99%. Chrysene-d12 (purity>99.9%) was purchased from Supelco (Bellefonte, PA, USA). p-Terphenyl-d14 (purity >98.5%) and pyrene-d10 (purity >98.6%)were purchased from AccuStandard (New Haven, CT, USA).Hexane and dichloromethane were purchased from VWRInternational (Suwanee, GA, USA). All organic solvents used inthis study were of reagent grade. GC capillary columns (J&WDB-EUPAH, 20 m×0.18 mm×0.14 μm, p/n 121-9627) anddeactivated GC liners (splitless tapered glasswool) werepurchased from Agilent Technologies (Wilmington, DE, USA).MC252 crude oil was provided by British Petroleum (BP).

Preparation of weathered oil under laboratory conditions

MC252 crude oil (1 mL) was transferred into an aluminum pan(5 cmdiameter and 1.5 cm deep) andwas evaporated for 7 dayswithin a laboratory fume hood (Mott Manufacturing Ltd,Brantford, Ontario, Canada) at a face velocity of 200 feet perminute (fpm). The weathering process was completed at roomtemperature (22 °C) and the resulting evaporated crude oil wasdesignated as ’Weathered MC252 oil’.

Column chromatographic fractionation

Fractionation of oil samples was conducted using methodssimilar to those discussed in earlier research.[13,14] Activatedsilica gel was prepared using the methods outlined in Wang

Figure 1. Structures of chrysene and alkylated-chrysenestandards used in this study.

Copyright © 2014 JoRapid Commun. Mass Spectrom. 2014, 28, 948–956

94

et al.[14] Anhydrous sodium sulfate was purified by heatingat 400 °C. A glass column (10.1 mm diameter and 200 mmlength) was plugged with glass wool, and about 3 g of activatedsilica gel was added to the column, followed by 1 g of anhydroussodium sulfate. Then 20mL of hexanewas added to the column.About 25 mg of an oil sample was weighed in a 12 mL glass vialand spiked with surrogates p-terphenyl-d14 and pyrene-d10. Thecontents were then extracted with 1 mL of hexane andtransferred to the column. The vialwas then sequentiallywashedwith 2 mL of hexane, with 1 mL in each step, and the contentswere transferred to the column. All solvents eluted from thecolumn prior to this step were discarded. Then 12 ml of hexanewas used to elute aliphatic hydrocarbon fractions from thesample. The hexane-eluted fraction was labeled as F1. Next, a50% hexane and 50% dichloromethane solvent mixture (15 mL)was used to elute all the aromatic hydrocarbons, and this fractionwas labeled as F2. The F1 and F2 fractions were concentratedunder a gentle stream of nitrogen gas and then reconstituted in10mLof a hexane and hexane/dichloromethane solventmixture(ratio of 1:1), respectively. The samples were filtered through0.2 μm PTFE filters, spiked with internal standard (IS)chrysene-d12, and analyzed for alkylated chrysenes. All thesamples were prepared in duplicate and analyzed in triplicate.

Instrumentation

An Agilent 7890 gas chromatograph coupled with an Agilent7000B triple quadrupole (QqQ) mass spectrometer, fitted withan electron ionization (EI) source and a collision cell was used.Various analytical experiments were performed using thisinstrument running under different conditions including full-scan, product-ion scan, and multiple reaction monitoring(MRM)modes. TheGCandMS conditions used are summarizedin Table 1.

Full-scan analysis

To identify the appropriate precursor ions, high concentrations oftarget analytes (5 mg/L) were prepared by diluting thestandards in a solvent mixture of hexane and dichloromethane(ratio of 1:1). These samples were then run in the full-scanmode,scanning from m/z 50 to 300. Results of the full-scan analysiswere used to select the precursor ion for each homologue.

Product ion scan analysis

The product-ion (PI) spectra of selected precursor ions wereacquired at various collision energies (ranging from 5 to40 eV). To facilitate the collision-induced dissociation (CID)of the precursor ion, ultra high purity nitrogen was deliveredto the collision cell at a flow rate of 1.5 mL/min, and heliumwas delivered at a flow rate of 2.3 mL/min to quench thereactions. From the product-ion spectrum of a precursor ion,the most abundant product ions and the optimal collisionenergy required to obtain these ions were selected.

Multiple reaction monitoring (MRM) analysis

The combination of precursor ion, the most intense product ionsand their corresponding optimal collision energies wereemployed to form the MRM transitions for each analyte; twodifferent MRM transitions were monitored – one as aquantitative transition and the other as a confirmatory transition.

wileyonlinelibrary.com/journal/rcmhn Wiley & Sons, Ltd.

9

Figure 2. Full-scan electron ionization mass spectra for theseven alkylated chrysene standards.

Table 1. Gas chromatograph and mass spectrometerparameters

GC conditions

Column DB-5MS (J&WAgilent Technologies,20 m×180 μm×0.14 μm)

Inlet temperature 300 °CInlet pressure 22.151 psiCarrier gas HeliumFlow rate 0.8 mL/minInjection mode Pulsed splitlessOven program 50 °C (0.8 min hold); 40 °C/min to

250 °C (2 min hold); 0.5 °C/min to260 °C (2 min hold); 30 °C/minto 325 °C(1 min hold)

Total run time 32.97 minInjection volume 1 μLTransfer linetemperature

320 °C

MS conditionsDelta EMV –70 eVAcquisitionparameters

Electron ionization (EI)

Solvent delay 2.5 minMS sourcetemperature

350 °C

Quadrupoletemperatures

Q1=Q2= 180 °C

G. F. John et al.

950

RESULTS AND DISCUSSION

Full-scan analysis

The electron ionization (EI) mass spectra of the alkylatedchrysene standards are shown in Fig. 2. For C1-chrysenes, thebase peak is the molecular ion [M]+ at m/z 242, as cleavage ofthe aromatic ring requires much higher energy.[15] Fragmentions were observed at m/z 241 and 239. All three C1-chryseneisomers exhibit similar EI mass spectra.For the C2-chrysene, 6-ethylchrysene, the base peak is the

[M–CH3]+ ion at m/z 241 (Fig. 3), formed by cleavage of the

benzylic bond of the ethyl group.[15] Other abundant ions werethe molecular ion at m/z 256 and a fragment ion at m/z 239.For the C3-chrysene, 1,3,6-trimethylchrysene, the base peak is

the molecular ion at m/z 270 with major fragment ions at m/z255 and 239. The base peak for 6-n-propylchrysene, however,is the [M–C2H5]

+ ion at m/z 241 formed by cleavage of thebenzylic bond of the propyl group[15]; other abundant ionswere themolecular ion atm/z 270 and a fragment ion atm/z 239.For 6-n-butylchrysene, (a C4-chrysene) the base peak is the

[M–C3H7]+ ion at m/z 241, formed by cleavage of the benzylic

bond of the butyl group[15]; other abundant ions were themolecular ion at m/z 284 and a fragment ion at m/z 239.

Analysis of quantitative MRM transitions fromexperimental data

The parent chrysene molecule, chrysene-d12, and thesurrogates, p-terphenyl-d14 and pyrene-d10, have been studiedextensively by others, and the expected MRM transitions canbe inferred from the literature data.[16,17] However, product-

wileyonlinelibrary.com/journal/rcm Copyright © 2014 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2014, 28, 948–956

Figure 3. Product-ion scan mass spectra for the sevenalkylated chrysene standards at a collision energy of 20 eV(precursor ions selected are reported in the respective figures).



95

ion (PI) information is not available for any of the alkylatedchrysenes. In all our PI scan experiments, we selected themolecular ions as the precursor ions since abundant molecularion peaks were obtained for all four alkylated chrysenehomologues (Fig. 2). The range in the PI scan experiment foreach pure alkylated chrysene standard was from m/z 50 to themass of the molecular ion.

Typical PI scans collected at 20 eV for all pure alkylatedchrysene standards are shown in Fig. 3. For all three C1-chrysenes, the predominant product ion is the [M–H]+ ionatm/z 241, with its maximum intensity occurring at a collisionenergy of 20 eV (Fig. 4). Therefore, the optimal quantitativeMRM transition for C1-chrysenes is m/z 242→m/z 241 at acollision energy of 20 eV.

For 6-ethylchrysene, a C2-chrysene, the predominantproduct ion was the [M–CH3]

+ ion at m/z 241, with itsmaximum intensity occurring at a collision energy of 15 eV(Fig. 4). Therefore, the optimal MRM transition for this typeof C2-chrysene was m/z 256→m/z 241 at a collision energyof 15 eV.

The most significant product ion from the molecular ion of1,3,6-trimethychrysene is the [M–H]+ ion at m/z 255 (Fig. 3),while the most significant product ion for 6-n-propylchryseneis the [M–C2H5]

+ ion at m/z 241. Based on these results, weselected m/z 270→m/z 255 as the quantitative MRMtransition for monitoring C3-chrysenes that have three methylgroups and m/z 270→m/z 241 as the quantitative MRMtransition for monitoring C3-chrysenes with a propyl group.Figure 4 shows that the optimal collision energies formonitoring these transitions to m/z 255 and 241 are 20 and15 eV, respectively. Therefore, the transitions of m/z 270→m/z 255 and m/z 270→m/z 241 were monitored at collisionenergies of 20 and 15 eV, respectively.

For 6-n-butylchrysene, a C4-chrysene, the most significantproduct ion (Fig. 3) is the [M–C3H7]

+ ion at m/z 241 whosemaximum intensity occurred at a collision energy of 15 eV (Fig. 4).Therefore, the optimal quantitative MRM transition for this typeof C4-chrysene ism/z 284→m/z 241 at a collision energy of 15 eV.

Analysis of quantitative MRM transitions from literature data

For C2-chrysene, the PI scan data (of the 6-ethylchrysenestandard) indicated that them/z 256→m/z 241 transitionwouldcapture all ethyl-type C2-chrysenes. However, C2-chrysenescan also have two methyl groups, and McLafferty andTureček[15] showed that an abundant fragment ion would beyielded by removing one of the methyl groups. Therefore, them/z 256→m/z 241 transition, originally intended to captureethyl-type C2-chrysenes, would also naturally identify thepresence of C2-chrysenes with two methyl groups.

From the C3-chrysene dataset, we know that multipleMRM transitions are needed to capture compounds havingthree different types of alkylations. Based on the PI scan dataof the C3-chrysene standards, we have already assigned them/z 270→m/z 255 transition for monitoring C3-chryseneshaving three methyl groups, and the m/z 270→m/z 241transition for monitoring C3-chrysenes having a propylgroup. However, C3-chrysene compounds can also have anethyl and methyl group. Based on our PI scan analysis ofthe other chrysenes, we expect the most significant production for these chrysenes to be the [M–CH3]

+ ion at m/z 255.For i-propylchrysenes we again expect the predominant


1

Figure 4. Collision energy optimization data for the sevenalkylated chrysene standards (precursor and product ionsselected are reported in the respective figures).

G. F. John et al.

wileyonlinelibrary.com/journal/rcm Copyright © 2014 John Wi

952

product ion to be the [M–CH3]+ ion at m/z 255.[18] Therefore,

the m/z 270→m/z 255 transition, originally intended tomonitor C3-chrysenes having three methyl groups, wouldalso capture C3-chrysenes having an ethyl group and amethyl group and the i-propylchrysenes.

For the C4-chrysenes, the PI scan data of 6-n-butylchryseneindicated that the m/z 284→m/z 241 transition would captureall the butyl-type C4-chrysenes. However, C4-chrysenes canalso have other types of alkyl substitution with compoundsinvolving: (a) four methyl groups, (b) two ethyl groups, (c) amethyl and a propyl group, (d) i-butylchrysenes, and (e)t-butylchrysenes. For C4-chrysenes containing four methylgroups, we expect there to be an abundant [M–CH3]

+ production at m/z 269, as observed for 1,3,6-trimethychrysene.Therefore, the m/z 284→m/z 269 transition can be used tocapture all the C4-chrysenes containing four methyl groups.This transition to the [M–CH3]

+ product ion at m/z 269 wouldalso be suitable for the detection of C4-chrysenes containingtwo ethyl groups and of t-butylchrysene.[18] Therefore, the m/z284→m/z 269 transition, used to capture C4-chrysenes havingfour methyl groups, would also detect the C4-chrysenescontaining two ethyl groups and t-butylchrysene. We selecteda collision energy of 20 eV to monitor this transition.

For C4-chrysenes containing a methyl and a propyl groupthe most abundant product ion would be expected to be the[M–C2H5]

+ ion at m/z 255, as found for 6-n-propylchrysene.For i-butylchrysene, the predominant product ion would bethe [M–C3H7]

+ ion at m/z 241, as found for i-butylbenzene.[18]

Therefore, the m/z 284→m/z 255 transition is used to captureall the C4-chrysenes having a methyl and a propyl group, andt-butylchrysene, with the selected collision energy for thistransition being 15 eV.

Selection of confirmatory MRM transitions fromexperimental data

Confirmatory MRM transitions are chosen to improve theselectivity and sensitivity of GC/MS/MS MRM methods.[19]

Full-scan EI mass spectra of all seven alkylated chrysenestandards yielded a fragment ion at m/z 239, although theintensities differed. Furthermore, PI spectra of the molecularions [M]+ of all the compounds yielded a product ion at m/z239, whose abundance increased as the collision energy wasincreased. Therefore, we used the following four confirmatoryMRM transitions: m/z 242→m/z 239, m/z 256→m/z 239, m/z270→m/z 239 and m/z 284→m/z 239 to verify the C1-, C2-, C3-and C4-chrysene homologues, respectively. The collision energyfor all these confirmatory transitions was set at 40 eV. A detailedsummary of the proposed MRM method is presented inTable 2. The relative intensities of the quantitative transitionto that of the confirmatory transition for p-terphenyl-d14,pyrene-d10, chrysene, 3-methylchrysene, 2-methylchrysene,1-methylchrysene, 6-ethylchrysene, 6-n-propylchrysene,1,3,6-trimethylchrysene and 6-n-butylchrysene are 1.30, 1.14,76.44, 1.48, 1.49, 1.53, 2.67, 3.85, 0.89 and 4.19, respectively.

Application of the MRMmethod for the characterization ofMC252 crude oil

The developed MRM method was used to characterize acomplex environmental sample, MC252 crude oil which wasspilled during the 2010 Deepwater Horizon oil spill event.[20,21]

ley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2014, 28, 948–956

Table 2. Summary of optimized MRM transitions used for quantifying various target compounds

Homologue/Compound

Quantitative transitions Confirmatory transitions

Ions (m/z) Collision energy (eV) Ions (m/z) Collision energy (eV)

†p-terphenyl-d14 212 ➔ 208 40 212 ➔ 210 40†Pyrene-d10 244 ➔ 212 40 244 ➔ 160 40†Chrysene-d12 240 ➔ 236 40 – –†Chrysene 228 ➔ 226 38 228 ➔ 224 38C1-chrysenes 242 ➔ 241 20 242 ➔ 239 40C2-chrysenes 256 ➔ 241 15 256 ➔ 239 40C3-chrysenes 270 ➔ 241 15 270 ➔ 239 40

270 ➔ 255 20C4-chrysenes 284 ➔ 241 15 284 ➔ 239 40

284 ➔ 255 15284 ➔ 269 20

†These MRM transitions are selected from published literature.

Figure 5. The total ion MRM chromatogram of the standardsolution used for calibration.


95

One of the major unknowns of this oil spill is the fate of toxicPAHs such as chrysene and its alkylated homologues.Therefore, detailed characterization of various componentsof chrysene in MC252 crude oil would help to provide thepreliminary data required to assess the potential human andecological risks resulting from the oil spill. Furthermore,Mulabagal et al.[21] reported hopane levels in MC252 oil spillresidue recovered along the Alabama shoreline and showedthat the relatively fresh looking, sticky brownish oil spillsamples are not degrading and hence have the potential toremain in the environment for an extended period of time.[22]

These studies support the importance of understanding thefate of various forms of chrysenes in oil spill samples.To accurately quantitate the concentration of alkylated

chrysenes in environmental samples, pure standards of variousforms of alkylated chrysenes are required; however, several ofthese compounds might not be available commercially.Therefore, United States Environmental Protection Agency(USEPA) methods typically use relative response factors (RRFs),which are estimated based on the response of the parentchrysene, to make semi-quantitative estimates of the alkylatedchrysenes.[23] It has been reported that methods that employa parent PAH to quantify alkylated homologues can introducelarge quantitation errors.[24] Other simpler methods havealso been proposed to make highly approximate estimatesfor alkylated compounds. For example, during the DeepwaterHorizon oil spill event, the USEPA recommended usinga constant (factor of 5 with respect to parent chrysene) toestimate the total amount of the alkylated chrysenes in oilspill samples.[25]

In this study, we used seven different alkylated chrysenestandards. The MRM responses of these standards (at theconcentration level of 100 ng/mL) are given in Fig. 5. Theirfragmentation patterns in full-scan and PI scan spectraare presented in Figs. 2 and 3, respectively. The data showthat all three C1-chrysene standards (3-methylchrysene, 2-methylchrysene and 1-methylchrysene) have similarfragmentation patterns, and they also show similar levels ofresponse. Therefore, any one of these pure compounds can beused as the standard to quantify C1-chrysenes. In this study, weselected 3-methylchrysene, which is a standard compound usedin previous research.[11]


Sinceonlyone isomer is commercially available for theC2- andC4-chrysene homologues (6-ethylchrysene for C2 and 6-n-butylchrysenefor C4), these two compounds were used as representativecalibration standards for these two sets of homologues.

In the case of C3-chrysenes, two standards were availablethat represent two distinct types of alkylation with twodifferent MRM transitions. We used 6-n-propylchrysene toquantify those C3-chrysenes that had the MRM transition m/z270→m/z 241, and 1,3,6-trimethylchrysene to quantify thosethat had the MRM transition m/z 270→m/z 255.

The quantitation approach used here (which uses a distinctcalibration standard for each homologue) is a robust approachfor quantifying different types of chrysene homologues. Acalibration stock solution consisting of 10 mg/mL of the parentchrysene, 3-methylchrysene, 6-ethylchrysene, 6-n-propylchrysene,1,3,6-trimethylchrysene, 6-n-butylchrysene, and two surrogatestandards p-terphenyl-d14 andpyrene-d10was prepared in a solventmixture of hexane and dichloromethane in a ratio of 1:1. Thesolution was then diluted to prepare calibration standards havingconcentrations of 1, 2, 5, 10, 20, 50, 100, 200 and 500 ng/mL. Thecalibration curves were constructed using the standard solutions.The linearity of the calibration curve was confirmed by ensuringthat the r2 values were greater than 0.99.[26]

LOD/LOQ and recovery data

The limit of detection (LOD) and limit of quantitation (LOQ)for the representative compounds were determined bymeasuring a series of blanks with no analyte. The blank meanvalue and the standard deviation (SD) were calculated for each


3

G. F. John et al.

954

representative compound.TheLOD is estimated as themeanblankvalue plus three times the SD. The LOQ is estimated as the meanblank value plus ten times the SD.[27] The LOD and LOQ valuesfor the representative compounds are presented in Table 3.The MC252 crude oil extract was spiked to increase the

concentration by 10 and 20 ng/mL (using the following purestandards: chrysene, 3-methylchrysene, 2-methylchrysene, 1-methylchrysene, 6-ethylchrysene, 6-n-propylchrysene, 1,3,6-trimethylchrysene and 6-n-butylchrysene) prior to GC/MS/MS analysis. The recovery levels were computed using thefollowing equation[28]:

%R ¼ CS � CU

Cn

� �X 100

where %R is the measured percentage recovery level, CS andCU are the concentration levels of the target analyte in thespiked and non-spiked samples, and Cn is the spikedconcentration level (10 and 20 ng/mL). The recovery levelsare summarized in Table 3. The overall spiked target recoverylevels observed ranged from 85 to 120% and these values arewell within typical spiked compounds recovery levels.[28]

Results for MC252 crude oil sample

The crude oil fractions F1 and F2 (discussed in the Experimentalsection) were analyzed and alkylated chrysenes were foundonly in the F2 fraction. In order to enhance sensitivity, withinthe MRM method, we set distinct time segments to monitoreach homologue. The total ion chromatogram for the crudeoil and the extracted ion chromatograms for each MRMtransition are given in Fig. 6.To quantify the concentrations, the extracted chromatographic

peaks that satisfied both the quantitative and the confirmatoryMRM transitions of a homologue were selected and quantifiedusing the respective calibration standard. The estimatedconcentration values of chrysene and its alkylated homologuesinMC252 crude oil are summarized in Table 4. The two surrogatecompounds, p-terphenyl-d14 andpyrene-d10,were also quantified;the relative recovery of the surrogate compounds ranged between80% and 120%, which is well within the USEPA recommendedlevels for surrogate recovery in environmental samples.[28]

The National Institute of Standards and Testing (NIST,Gaithersburg, MD, USA) compiled the analytical results ofthe MC252 oil characterization studies completed by 26

Table 3. LOD and LOQ values, and percentage recoveryvalues from spiking studies

CompoundLOD

(ng/mL)LOQ

(ng/mL)

%R

Mean %RSD

Chrysene 0.57 0.62 113.0 5.83-methylchrysene 0.11 0.12 101.8 12.62-methylchrysene 0.12 0.14 84.8 11.01-methylchrysene 0.12 0.14 84.1 9.96-ethylchrysene 0.92 0.93 89.8 7.06-n-propylchrysene 1.09 1.10 88.8 7.61,3,6-trimethylchrysene 0.77 0.80 118.3 6.66-n-butylchrysene 1.00 1.05 109.0 6.0

Figure 6. Total ion chromatogram and extracted ionchromatograms for variousMRMtransitionsused for quantifyingchrysene and alkylated-chrysene homologues in the MC252crude oil sample.


different laboratories and released a summary report.[11]

Table 4 compares the values of the different types ofchrysenes estimated in our study against the reference datasetpresented in the NIST report. The data show that theconcentration estimates for chrysene and its homologuesevaluated in our study are well within the range reportedby the NIST.


Table 4. Comparison of estimated values of variouschrysenes using the proposed GC/MS/MS method withthe NIST data reported for MC252 crude oil

Homologue

Concentration (mg/kg)

Currentstudy

NIST report

Mean Median Range

Chrysene 57± 3 44.8 47.5 20.7 – 64.4C1-chrysenes 128 ± 7 108 108 72.4 – 134C2-chrysenes 91 ± 4 130 128 92.5 – 163C3-chrysenes 166 ± 5 91.3 92.7 39.4 – 140C4-chrysenes 31 ± 1 65.4 70.9 39.1 – 87.8

Table 5. Method validation data acquired using lab-weathered MC252 crude oil sample

Homologue

Concentration inlab weathered MC252crude oil (mg/kg)

%weathering

of oil

Chrysene 92± 2 39C1-chrysenes 221± 5 42C2-chrysenes 170± 3 46C3-chrysenes 341± 10 51C4-chrysenes 67± 2 54Total 891 47


We also prepared a synthetic weathered oil sample byevaporating the original MC252 crude oil under a fume hoodfor 7 days, resulting in a loss of bulk oil mass of 44% (estimatedby the gravimetricmethod). Since chrysene and its homologuesare semi-volatile organic compounds (relatively resistant tovolatilization), ideally, they would have concentrated in theweathered sample by about 44%. The weathered MC252 oilsamplewas analyzed using theMRMmethod discussed above.The concentration levels of various chrysenes measured in thesynthetic sample are summarized in Table 5, and these valueswere also used to estimate the net weathering level of the oilusing the following equation:[21]

P ¼ 1� CS

CW

� �x 100

where P is the percentage weathering level with respect to thesource crude oil, and CS and CW are the concentrations ofchrysenes in MC252 and the weathered MC252 crude oilsamples, respectively. The average level of weatheringestimated using the total amount of chrysene was 47%, whichis close to the gravimetric estimate of 44%. Other estimatescomputed based on individual chrysenes (Table 5) showedless than 20% variability from the gravimetric estimate.

95

CONCLUSIONS

The objective of this study was to develop a GC/MS/MSMRM method for identifying and quantifying chrysene andits alkylated homologues in crude oil samples. The method


was developed based on full-scan and product-ion scan datacollected for seven different alkylated standards using a novelMS/MS instrument (Agilent 7000B MS/MS system). The full-scan data, product-ion scan data, and the collision cell energyoptimization data for all seven standards are provided. Basedon these experimental data, coupled with literature-derivedinformation, the fragmentation patterns of other isomersof chrysene homologues for which standards are notcommercially available were proposed. This informationwas used to develop a GC/MS/MS method for identifyingand quantifying the total concentrations of C1-, C2-, C3-, andC4-chrysenes in crude oil samples. The developed methodwas employed to characterize MC252 crude oil, released intothe Gulf of Mexico during the Deepwater Horizon oil spillevent. The results show that the concentrations of chryseneand its alkylated homologue estimated by the proposedmethod are mostly well within the range of previouslyreported values. The proposed method is a promisingapproach for quantifying alkylated chrysenes in morecomplex environmental samples such as crude oils andweathered oil-spill-waste samples.

AcknowledgementsThanks are extended to Professor Randall Clark, Department ofDrug Discovery and Development at Auburn University, forproviding various GC/MS ideas to the lead authors GFJ andFY; his kindness and generosity are much appreciated. RCMreviewers are also thanked for their detailed review comments.Thisworkwas supported, in part, by funding received from theCity of Orange Beach, Alabama, USA. The GC/MS/MS facilityand the characterization methods were developed using grantfunds received from the National Science Foundation (NSF-MRI grant #G00006697), Samuel Ginn College of Engineering,Auburn University, and Marine Environmental SciencesConsortium (MESC). All the authors have contributed to thiseffort and their relative responsibilities, computed using theClement[29] approach, are: GFJ (40%), FY (23%), VM (10%),JSH (7%), and TPC (18%).

REFERENCES

[1] J. M. Neff. Polycyclic Aromatic Hydrocarbons in the AquaticEnvironment: Sources, Fates and Biological Effects. AppliedScience Publishers, London, 1979.

[2] ATSDR. Chemical and physical information, in ToxicologicalProfile for Polycyclic Aromatic Hydrocarbons (PAHs), ATSDR,Atlanta, Georgia, USA, 1995.

[3] G.-W. Lien, C.-Y. Chen, C.-F. Wu. Analysis of polycyclicaromatic hydrocarbons by liquid chromatography/tandemmass spectrometry using atmospheric pressure chemicalionization or electrospray ionization with tropyliumpost-column derivatization. Rapid Commun. Mass Spectrom.2007, 21, 3694.

[4] Z. Wang, K. Li, P. Lambert, C. Yang. Identification,characterization and quantitation of pyrogenic polycylicaromatic hydrocarbons and other organic compounds in tirefire products. J. Chromatogr. A 2007, 1139, 14.

[5] D. Turcotte, J. V. Headley, N. L. Abudulai, P. V. Hodson,R. S. Brown. Identification of phase II in vivo metabolites ofalkyl-anthracenes in rainbow trout (Oncorhynchus mykiss).Chemosphere 2011, 85, 1585.


5

G. F. John et al.

956

[6] Z. Wang, M. Fingas, S. Blenkinsopp, G. Sergy, M. Landriault,L. Sigouin, J. Foght, K. Semple, D. W. S. Westlake. Comparisonof oil composition changes due to biodegradation and physicalweathering in different oils. J. Chromatogr. A 1998, 809, 89.

[7] Z. Wang, M. F. Fingas. Development of oil hydrocarbonfingerprinting and identification techniques. Mar. Pollut.Bull. 2003, 47, 423.

[8] F. Chaillan, M. Gugger, A. Saliot, A. Couté, J. Oudot. Role ofcyanobacteria in the biodegradation of crude oil by atropical cyanobacterial mat. Chemosphere 2006, 62, 1574.

[9] Z. Wang, M. Fingas, G. Sergy. Study of 22-year-old arrow oilsamples using biomarker compounds by GC/MS. Environ.Sci. Technol. 1994, 28, 1733.

[10] V. Fernández-González, S. Muniategui-Lorenzo, P. López-Mahía, D. Prada-Rodríguez. Development of a programmedtemperature vaporization-gas chromatography–tandemmass spectrometry method for polycyclic aromatichydrocarbons analysis in biota samples at ultratrace levels.J. Chromatogr. A 2008, 1207, 136.

[11] M. M. Schantz, J. R. Kucklick. National Institute of Standardsand Technology, Gaithersburg, MD 20899, USA, 2011.

[12] E. Pitarch, C. Medina, T. Portolés, F. J. López, F. Hernández.Determination of priority organic micro-pollutants in waterby gas chromatography coupled to triple quadrupole massspectrometry. Anal. Chim. Acta 2007, 583, 246.

[13] D. R. Luellen, D. Shea. Calibration and field verification ofsemipermeable membrane devices for measuring polycyclicaromatic hydrocarbons in water. Environ. Sci. Technol. 2002,36, 1791.

[14] Z. Wang, M. Fingas, K. Li. Fractionation of a light crude oiland identification and quantitation of aliphatic, aromatic,and biomarker compounds by GC-FID and GC-MS, Part I.J. Chromatogr. Sci. 1994, 32, 361.

[15] F. W. McLafferty, F. Tureček. Interpretation of Mass Spectra,(4th edn.). University Science Books, Sausalito, CA, 1993.

[16] K. Xia, G. Hagood, C. Childers, J. Atkins, B. Rogers, L. Ware,K. Armbrust, J. Jewell, D. Diaz, N. Gatian, H. Folmer.Polycyclic aromatic hydrocarbons (PAHs) in Mississippiseafood from areas affected by the Deepwater Horizon oilspill. Environ. Sci. Technol. 2012, 46, 5310.

[17] Y. Sapozhnikova, S. J. Lehotay. Multi-class, multi-residueanalysis of pesticides, polychlorinated biphenyls, polycyclicaromatic hydrocarbons, polybrominated diphenyl ethersand novel flame retardants in fish using fast, low-pressuregas chromatography–tandem mass spectrometry. Anal.Chim. Acta 2013, 758, 80.

[18] P. J. Linstrom, W. G. E. Mallard. NIST Chemistry WebBook,NIST Standard Reference Database Number 69, NationalInstitute of Standards and Technology, Gaithersburg MD,20899. Available: http://webbook.nist.gov (accessedDecember 16, 2013).


[19] J. Nácher-Mestre, R. Serrano, T. Portolés-Nicolau,F. Hernández, L. Benedito-Palos, J. Pérez-Sánchez. A reliableanalytical approach based on gas chromatography coupledto triple quadrupole and time-of-flight mass analyzers forthe determination and confirmation of polycyclic aromatichydrocarbons in complex matrices from aquacultureactivities. Rapid Commun. Mass Spectrom. 2009, 23, 2075.

[20] J. S. Hayworth, T. P. Clement. BP’s Operation Deep Clean –Could dilution be the solution to beach pollution? Environ.Sci. Technol. 2011, 45, 4201.

[21] V. Mulabagal, F. Yin, G. F. John, J. S. Hayworth,T. P. Clement. Chemical fingerprinting of petroleumbiomarkers in Deepwater Horizon oil spill samples collectedfrom Alabama shoreline. Mar. Pollut. Bull. 2013, 70, 147.

[22] M. Pauzi Zakaria, T. Okuda, H. Takada. Polycyclic aromatichydrocarbons (PAHs) and hopanes in stranded tar-balls onthe coasts of peninsular Malaysia: Applications ofbiomarkers for identifying sources of oil pollution. Mar.Pollut. Bull. 2001, 42, 1357.

[23] USEPA. Method 8270D: Semivolatile Organic Compounds byGas Chromatography/Mass Spectrometry (GC/MS), 2007.Available: http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/8270d.pdf. (accessed April 10, 2013).

[24] M. R. Burkhardt, S. D. Zaugg, T. L. Burbank, M. C. Olson,J. L. Iverson. Pressurized liquid extraction using water/isopropanol coupled with solid-phase extraction cleanupfor semivolatile organic compounds, polycyclic aromatichydrocarbons (PAH), and alkylated PAH homolog groupsin sediment. Anal. Chim. Acta 2005, 549, 104.

[25] USEPA. Explanation of PAH benchmark calculations usingEPA PAH ESB approach originally developed by DaveMount, ORD Duluth, 2010. Available: http://www.epa.gov/bpspill/water/explanation-of-pah-benchmark-calculations-20100622.pdf (accessed August 24, 2013).

[26] USEPA. Method 8272 – Parent and alkyl polycyclic aromaticsin sediment pore water by solid-phase microextractionand gas chromatography/mass spectrometry in selectedion monitoring mode, 2007. Available: http://www.epa.gov/osw/hazard/testmethods/pdfs/8272.pdf (accessedSeptember 20, 2013).

[27] D. MacDougall, W. B. Crummett. Guidelines for dataacquisition and data quality evaluation in environmentalchemistry. Anal. Chem. 1980, 52, 2242.

[28] USEPA. Method 8000B – Determinative chromatographicseparations, 1996. Available: http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/8000b.pdf (accessedSeptember 20, 2013).

[29] T. P. Clement. Authorship matrix: A rational approach toquantify individual contributions and responsibilities inmulti-author scientific articles. Sci. Eng. Ethics 2013. DOI:10.1007/s11948-013-9454-3.


http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/8000b.pdf

http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/8000b.pdf

Development and application of an analytical method using...

Documents

Transcript of Development and application of an analytical method using...