Aud

Ja

b

ARRAA

KAAPALS

1

m(baitatm

mT

0h

Journal of Chromatography A, 1294 (2013) 33– 40

Contents lists available at SciVerse ScienceDirect

Journal of Chromatography A

jou rn al hom epage: www.elsev ier .com/ locate /chroma

new liquid chromatography–tandem mass spectrometry methodsing atmospheric pressure photo ionization for the simultaneousetermination of azaarenes and azaarones in Dutch river sediments

an Brulika,b,∗, Zdenek Simekb, Pim de Voogta

Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The NetherlandsResearch Centre for Toxic Compounds in the Environment, Masaryk University, Kamenice 753/5, 625 00 Brno, Czech Republic

a r t i c l e i n f o

rticle history:eceived 1 February 2013eceived in revised form 27 March 2013ccepted 29 March 2013vailable online 17 April 2013

eywords:zaareneszaaronesAHPPIC–MS/MSedimets

a b s t r a c t

A new method for the analysis of azaarenes and their degradation products (azaarones) was devel-oped, optimized and validated using liquid chromatography coupled with atmospheric pressure photoionization tandem mass spectrometric detection (LC–APPI/MS/MS). Seventeen compounds including4 PAHs (naphthalene, anthracene, phenanthrene, benz[a]anthracene), 7 azaarenes (quinoline, acri-dine, phenanthridine, 5,6-benzoquinoline and 7,8-benzoquinoline, benzo[a]acridine, benzo[c]acridine),and 6 azaarones (2-OH-quinoline, 4-OH-quinoline, 5-OH-quinoline, 6-OH-quinoline, 9(10H)-acridone,6(5H)phenanthridinone) were analyzed in sediment samples from Dutch rivers. All compounds wereanalyzed simultaneously in multi reaction monitoring (MRM) mode. Soxhlet extraction was used for theextraction of analytes from sediments. The limits of quantification of azaarenes and azaarones variedfrom 0.21 to 1.12 �g/l and from 0.23 to 1.58 �g/l, respectively. The limits of quantification for PAHs var-ied from 32 to 769 �g/l. Matrix-independent recoveries of sediment samples were in the range 85–110%;matrix-dependent recoveries were in the range 73–148%, respectively. The method was tested on realsediment samples and the results were compared with a previous study in which GC/MS/MS was used

for the simultaneous measurement of azaarenes and azaarones. 4-, 5- and 6-OH-quinolines and naphtha-lene, anthracene and phenanthrene were not present or below detection limits in some samples. All otheranalytes were present in samples in the concentration range 0.2–1200 ng/g (dw). To our knowledge, thisis the first report showing the possibility of measurement non-polar polyaromatic hydrocarbons togetherwith polar azaarenes and their degradation products azaarones simultaneously with sufficient sensitivityand accuracy using LC/MS/MS.

. Introduction

Azaarenes and azaarones belong to a group of polycyclic aro-atic compounds related to polycyclic aromatic hydrocarbons

PAH). Compared with homocyclic analog PAHs, one or more car-on atoms in the aromatic structure of azaarenes are replaced with

nitrogen atom. Azaarenes are produced naturally by organismsn the form of mycotoxins, alkaloids and nucleotides [1,2], buthe largest amount occurring in the environment originates from

nthropogenic sources such as chemical industry including coalar and oil shale processing, wood preserving facilities, chemical

anufacturing plants, and from the combustion of hydrocarbons

∗ Corresponding author at: Research Centre for Toxic Compounds in the Environ-ent, Masaryk University, Kamenice 753/5, 625 00 Brno, Czech Republic.

el.: +420 31641173936.E-mail address: brulik@recetox.muni.cz (J. Brulik).

021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.chroma.2013.03.079

© 2013 Elsevier B.V. All rights reserved.

[1]. The first paper concerning the determination of azaarenes inmarine sediments was published in 1977 [3]. Since then, manypapers focused on the determination of azaarenes in soil, sediment,water, air and biota have been published [4–8]. The concentrationsof azaarenes in environmental matrices are usually between 1 and10% of the concentration of PAHs [9]. However, due to nitrogenin their aromatic structure, azaarenes are more polar compoundswith higher water solubility than PAHs. Therefore, they are morebioavailable. Additionally, an unknown portion of azaarenes is bio-logically or photochemically oxidized to hydroxylated azaarenes,also known as azaarones in the aquatic environment [10,11]. Theconcentrations of hydroxylated N-heterocycles measured in sed-iments have been reported as higher than the concentrations ofrelated PANHs [4,12]. Moreover, previous studies have shown

that acute toxicity decreases from PANHs to azaarones, but someof the azaarones have a much higher chronic toxic effect thanazaarenes, comparable, in fact, with PAHs [13]. For the general eval-uation of toxic pressure as a result of the presence of polycyclic

3 atogr

aiatTehtMcs[(tddoFo�tonhctimamttutwwpiew

iaPta[Aadio[

Wcaat

2

2

f

4 J. Brulik et al. / J. Chrom

romatic compounds in water or sediment, it is necessary to takento account not only the content of PAHs, but also that of azaarenesnd azaarones in the environment. For this reason, the simul-aneous analysis of PAHs, azaarenes and azaarones is necessary.here are many methods for analysing PAHs and PANHs in thenvironment, but, as far as the authors know, only one methodas been published which allows the simultaneous determina-ion of these three groups of polycyclic aromatic compounds [14].

ethods used for the determination of PAHs by means of gashromatography (GC), usually in conjunction with tandem masspectrometry (MS/MS), provide detection limits in the range of ng/L15,16]. Also the use of high performance liquid chromatographyHPLC) with diode array detection (DAD) [17] or fluorescence detec-ion [18] has been published. Several methods were used for theetermination of PAHs by means of liquid chromatography–massetection [19]. Methods for the determination of azaarenes basedn LC with MS/MS detection [6,8] were published recently, butLD or DAD detectors are very often still being used. In the casef fluorescence detection, limits of detection are in the range ofg/L [18,20] which is comparable with mass spectrometry detec-

ion [8], can be achieved. In comparison with PAHs and PANHs,nly limited information can be found on the analytical determi-ation and occurrence of azaarones in the environment. Studiesave been conducted focusing on single hydroxylated azaareneompounds (mainly 8-hydroxyquinoline) [21,22] but only some ofhese focused on the complex analysis of mixtures of azaaronesn the environment [4,8,18]. LC–MS [12,23,24] and/or GC–MS [4]

ethods are used for the simultaneous determination of azaarenesnd azaarones. Of all of these methods, LC–MS/MS appears to be theost advantageous for the environmental analysis of PANHs and

heir hydroxylated derivatives. In comparison with FLD or DAD,andem mass spectrometry provides both structural informationseful for the identification of target compounds and lower detec-ion limits. Moreover, substances with similar retention times butith different transitions can easily be measured and identifiedithout pre-fractionation procedures or clean up during samplereparation. Sample preparation for LC is much less time consum-

ng and simpler when compared with GC. In our study crude organicxtracts can be filtered and injected directly into the LC systemhich is demonstrated in this paper.

The main ionization techniques used for LC are electro-sprayonization (ESI), atmospheric pressure chemical ionization (APCI),nd atmospheric pressure photo ionization (APPI). Methods forANHs based on ESI and APPI have been described in the litera-ure [6,8] as well as methods for the analyses of PAHs using APCInd APPI ionization [19,25] and the analyses of azaarones using ESI12]. However, a method for the measurement of azaarones usingPPI is still missing. The present study describes the developmentnd optimization of a LC–APPI/MS/MS method for the simultaneousetermination of 7 PANHs, 6 hydroxy PANHs and 4 PAHs. The work-

ng range of the method for sediments was set up to cover the rangef concentrations previously measured in the Dutch coastal zone4].

Samples of sediments and waters from the River Rhine and theestern Scheldt estuary were collected for verification of the appli-

ability of the method. The extraction method for water samplesppeared to be non suitable for measurement of some compoundsnd details of this method are described in Electronic Supplemen-ary Material.

. Materials and methods

.1. Chemicals

Water, methanol and acetonitrile (all in ULC/MS quality) usedor the standard solutions and mobile phase preparation were

. A 1294 (2013) 33– 40

purchased from Biosolve, The Netherlands. Ethyl acetate p.a. andn-hexane p.a. used for Soxhlet extraction, and toluene as dopantfor APPI ionization were obtained from Acros, Belgium. KieselguhrGR was used as received from Merck, Germany. Nitrogen used forextract evaporation and mass spectrometry was provided by Linde,The Netherlands. Deuterated standards D7-quinoline and D9-acridine were purchased from Cambridge Isotopes, U.S. Chemicalstandards 9(10H)-acridone, 6(5H)-phenanthridinone, phenanthri-dine, naphthalene, anthracene, acridine, benzo(c)acridine, quino-line, 2-OH-quinoline, 4-OH-quinoline, 5-OH-quinoline, 6-OH-quinoline and 8-OH-quinoline were obtained from Sigma–Aldrich,The Netherlands. Benzo(a)acridine was obtained from LGC Stan-dards, The Netherlands. Phenanthrene was obtained from J.T.Baker, The Netherlands. Benzo(a) anthracene was obtained fromDr. Ehrensdorfer, The Netherlands, and 7,8-Benzoquinoline, 5,6-benzoquinoline were obtained from Acros, Belgium. All compoundswere of highest purity available (invariably higher than 98%).

2.2. Preparation of standard solutions

Single stock solutions of all chemical standards were preparedin acetonitrile. Amounts of 5 ± 0.25 mg of azaarenes and azaaronesand 50 ± 0.25 mg of PAHs were individually diluted with acetoni-trile in 100 ml volumetric glass to concentrations of approx. 50 mg/land 500 mg/l respectively.

A mixed standard solution was prepared from 17 single stocksolutions by mixing 1 ml of each individual stock solution (exclud-ing D-standards) in a volumetric flask and diluting it to 25 mlwith acetonitrile to produce final concentrations of about 2 mg/lof azaarenes and azaarones and 20 mg/l of PAHs.

Stock solutions of D-quinoline and D-acridine used as inter-nal standards were prepared in acetonitrile in concentrations ofapprox. 5 mg/l. 1 mL of each of the D-quinoline and D-acridinestock solutions were mixed together and diluted with acetonitrileto 10 ml. 100 �l of this solution was added to each sample to achievea final concentration of 50 �g/l of internal standards in 1 ml of sam-ple. D-quinoline was used as internal standard for all azaarones andquinoline and D-acridine for all other analyzed compounds.

Nine calibration levels contained from 0.2 to 60 �g/l of all PANHsand hydroxy-PANHs and 2–600 �g/l of PAHs were prepared. Tocover the whole range of PAHs concentrations present in sediment5, extra calibration solutions containing only PAHs in the range1000–5000 �g/l were prepared.

2.3. Sampling

Sediment samples were taken by grab sampling and a box corerfrom 7 different locations at Rotterdam-Botlek Harbor (river Rhine)in April 2010 (5 samples) and October 2010 (2 samples). 3 sampleswere collected from the shore (mudflat) at the Hansweert location(Western Scheldt) in April 2011. At each location, 200 ml of wet sed-iment samples were taken and immediately transferred into glassjars and stored at −20 ◦C until analysis.

All sampling locations are shown in Fig. 1.

2.4. Sample preparation

One day before analysis, sediment samples were removed fromthe freezer (−20 ◦C) and thawed to room temperature. Stones,non-decomposed organic material, and shells were removedmechanically. All parts of equipment used for sample extractionwere cleaned with acetonitrile and methanol. Subsequently, 5 g

of wet sediment were mixed with kieselguhr in order to dry, andthen 100 �l of the internal standard was added. Soxhlet extrac-tion was used as the extraction method for sediment samples.This method was developed and validated previously for the

J. Brulik et al. / J. Chromatogr. A 1294 (2013) 33– 40 35

Fig. 1. Map of sediment sampling locations in Rotterdam-Botlek and Hansweert. Sediments samples 1–5 were taken in April 2010, 6–7 in October 2010 and 8–10 in April2

eosesTeTfi

2

popSua

2

o5twMac

011.

xtraction of azaarenes and azaarones in our laboratory [4]. 150 mlf a mixture of EtAc:Hexane (2:8, v:v) was used as the extractionolvent. The extraction time was 18 h. Extracts thus obtained werevaporated using a glass vigreux column to approx. 10 ml and sub-equently evaporated under a gentle stream of nitrogen to 0.5 ml.he evaporated sample was transferred into 5 ml of acetonitrile andvaporated again under a nitrogen stream to a final volume of 1 ml.he final evaporated solution was filtered through a 0.2 �m syringelter (GHP Acrodisc 13) and stored at 4 ◦C until LC/MS analysis.

.5. Instrumentation

A modular HPLC system (comprising LC-20AD mobile phaseumps, SIL-20AC autosampler with thermostat, CTO-20AC columnven, CBM20 communication module, and LC-10AD as a dopantump; Shimadzu, Kyoto, Japan) coupled with MS/MS system ABciex APPI/4000QtrapTM (Applied Biosystems, California, USA) wassed for LC/MS analyses. Instrument control, data acquisition, andnalysis were performed using Analyst software v 1.4.1.

.6. Chromatographic conditions

The new separation method employed in this study was devel-ped and optimized using an ACE 5 C18 column (250 × 4.6 mm i.d.,

�m Advanced Chromatography Technologies, Scotland, UK) pro-ected with the same type of guard column. Methanol and water

ere used for gradient elution as follows: 0–5 min – 10–35% ofeOH; 5–75 min – 35–90% MeOH; 75–85 min – 95% MeOH. An

dditional 15 min cleaning process (95% MeOH) was used after thehromatographic run. The mobile phase flow was 0.6 ml/min and

the column temperature was adjusted to 50 ◦C. 10 �l of sampleswas injected.

2.7. MS/MS conditions

The AB Sciex 4000QtrapTM operated in positive mode wascoupled with an APPI source. The MS/MS conditions for MRManalysis were evaluated in direct infusion mode. Single com-pound standard solutions of azaarenes and azaarones prepared inmethanol at a concentration of 500 �g/l and PAHs at 5000 �g/lwere directly infused into the photoionization source with aHarvard 11 syringe pump (Harvard apparatus, Holliston, USA).The flow of the syringe pump was set to 50 �l/min. The ion-ization conditions of precursor ions, i.e. declustering potential(DP), entrance potential (EP), and collision cell exit potential(CXP), were optimized in the standard optimization procedure.The focusing parameters were optimized by flow injection analy-sis (FIA) using 10 �l injection volume of single standard solutioninjected repeatedly into the system, bypassing the chromato-graphic column. Subsequently all parameters were verified on theanalytical column. The optimized parameters were as follows:source temperature (TEM) 500 ◦C, ion spray voltage (IS) 600 V,curtain gas (CUR) 10 AU (Arbitary Units), collision gas (CAD) 10AU, ion source gas 1 (GS1) 45 AU, and ion source gas 2 (GS2) 40AU.

Toluene was used as a dopant at 20% of the flow of the

mobile phase rate (120 �l/min). An example of the obtainedchromatograms can be found in Electronic SupplementaryMaterial.

Analyses were performed in scheduled MRM mode.

3 atogr. A 1294 (2013) 33– 40

3

3

ifPsc

rmwp[

qE

ottAocalhCorspcmaoesdim9

wcdwbpgtccfi

L1ObfhpM

Table 1Matrix depend and independent recovery for sediment (n = 3) samples.

Sedimentmatrixindependentrecovery (%)

CV (%) Sedimentmatrix dependrecovery (%)

CV (%)

4-OH-quinoline 103 2 97 36-OH-quinoline 105 3 127 65-OH-quinoline 99 5 136 72-OH-quinoline 102 3 91 4Quinoline 110 3 103 39(10H)acridon 96 4 100 46(5H)phenanthridinon 98 6 107 5Acridine 107 4 98 35,6-Benzoquinoline 108 5 112 5Phenanthridine 102 3 113 67,8-Benzoquinoline 104 3 108 3Naphthalene 85 5 73 8Benzo[a]acridine 99 2 100 2Anthracene 108 4 97 4Phenanthrene 95 4 109 5

6 J. Brulik et al. / J. Chrom

. Results and discussion

.1. LC–APPI-MS/MS

The use of APPI ionization enables the efficient ionization of allnvestigated compounds [26]. The main precursor ions observedor PAHs (naphthalene, anthracene and phenanthrene) were M+•.rotonated molecules [M+H]+ were observed as the main precur-or ions for benzo(a)anthracene and for all azaarene and azaaroneompounds as a result of APPI ionization.

These results are in accordance with previously publishedesults for the photo ionization of these compounds [6,27,28]. Theechanism of formation M+• and/or [M+H]+ precursor ion for PAHsas described earlier as being strongly influenced by the mobilehase used and the reaction time in the photoionization source27].

The most intensive MRM transitions were chosen for theuantification and identification of analytes in MS/MS mode (seelectronic Supplementary Material).

Only limited information concerning fragmentation patternsf the investigated compounds in APPI-MS/MS is available fromhe literature. Typical fragmentation characteristics for PAHs werehe presumable loss of C2H2 [M•−26]+ and C6H4 [M•−76]+ [19].dditionally, benzo(a)anthracene showed a very intensive signalf transition 229/228. For azaarenes, the main observed fragmentsorresponded to a loss of CH2N [M+H−28]+, C3H3N [M+H−53]+

nd C7H5N [M+H−103]+; however, a fragment resulting from theoss of CHN [M+H−27]+ was also observed. The fragmentation ofydroxyquinolines was caused mainly by losses of CO [M+H−28]+,HO [M+H−29]+, C3H3NO [M+H−69]+, and H2O [M+H−18]+. Ionsf m/z 77 correspond to C6H5

+ and m/z 91 the tropilium ion,espectively. They are typically present in the product ion masspectrum of hydroxyquinolines. For 9(10H)-acridone and 5(6H)-henanthridone, mass losses were less specific and differentompared to other investigated compounds. A neutral molecularass loss of CHxO was observed in the case of 9(10H)-acridone

nd in the case of CHxNO with 5(6H)-phenanthridone. Examplesf product ion spectra are shown in Fig. 2. Due to the pres-nce of isomers with equal MS/MS transitions, chromatographiceparation was carefully optimized. Different methanol/water gra-ients were tested (see Electronic Supplementary Material). The

nitial mobile phase composition was in the range from 10 to 40%ethanol increasing to a final composition with 100% methanol in

5 min.The majority of the compounds were sufficiently separated

ith minimal peak overlap except for 8-OH-quinoline although theonditions for 8-OH-quinoline were sufficiently optimized usingirect infusion method. No peak corresponding to this compoundas observed in the chromatogram. This effect was most proba-

ly caused by immobilization of 8-OH-quinoline in the stationaryhase of the chromatographic column in combination with theradient used. Apparently this gradient did not allow rapid elu-ion of 8-OH-quinoline resulting in an extremely broad peak in thehromatogram which could not be detected in a selected range ofoncentrations. Therefore it was decided to omit 8-OH-quinolinerom further experiments. Examples of chromatograms obtainedn scheduled MRM mode are presented in Fig. 3.

The maximum volume of sample that could be injected into theC system without unacceptable effects on the peak shape was0 �l of sample in acetonitrile and 20 �l of sample in methanol.ptimization of the dopant flow of toluene or benzene, which coulde between 5 and 50% of total HPLC flow, was tested and 20% was

ound to be an optimal value for toluene taking into account theighest signal to noise ratio for azaarones and azaarenes. Exam-les of chromatograms can be found in Electronic Supplementaryaterial.

Benzo[c]acridine 100 4 106 6Benzo[a]anthracene 104 3 148 13

3.2. Validation of measurement

The matrix-dependent recovery of the extraction of analytesfrom sediments was tested by spiking the sediment samples col-lected at the Rotterdam 1, 3 and 5 locations. 200 �l of standardmixture solution (final concentration in sample of 400 or 4000 �g/l)was added to each sample 2 h before the extraction. The aver-age matrix-dependent recovery of extraction (n = 3) was in therange of 91% (2-hydroxyquinoline) to 148% (benzo[a]anthracene).The matrix-independent recovery was tested by spiking approx 2 gof kieselguhr with 200 �l of standard mixture solution 2 h beforethe extraction. The analytes were then extracted from kieselguhrand the extracts were treated using the same procedures as forsediment samples. The average recovery from matrix-unaffectedsamples was in the range of 95% (phenanthrene) to 110% (quino-line).

Table 1.The relative matrix interferences using deuterated internal stan-

dards (MIs) were tested by comparing the ratio of analyte peak areato internal standard peak area for samples prepared in pure solvent(n = 3) and the ratio for spiked sediment extracts (n = 3) accordingto Eq. (1).

MIs = CAACN/IAACN

(CAS − BS)/IAS(1)

where CAACN is the peak area of an analyte in sample prepared inacetonitrile, IAACN is the peak area of an internal standard in thesame sample prepared in acetonitrile, CAS is the peak area of ananalyte in a spiked sediment sample, IAS is the peak area of an inter-nal standard in the same spiked sediment sample and BS is the peakarea in a non-spiked sediment sample.

The matrix interference values MIs were in the range 0.82–1.15(Table 2) for all compounds, which indicates effective coveringof matrix effects with use of selected two internal standards.The compounds 4-OH-quinoline, 2-OH-quinoline, naphthalene andanthracene were relatively more effected by the matrix effect, prob-ably as a results of non specific matrix interferences caused byco-elution of organic and inorganic components which can affectthe ionization process of the analytes [6]. The repeatability ofLC/MS/MS measurement was tested by means of 5 replicate injec-

tions of the spiked sediment sample with concentrations of addedanalytes of about 12 �g/l (nitrogen containing compounds) andabout 2300 �g/l (PAHs). The relative standard deviation was lessthan 9% for all compounds, except for the two benzoacridines with

J. Brulik et al. / J. Chromatogr. A 1294 (2013) 33– 40 37

TCa

Fig. 2. Examples of fragmentation spectra generated by LC–APPI-MS/M

able 2alculated relative matrix effect coefficient using internal standard correlation (MIs)nd internal standard used for compound evaluation.

MIs Internal standard

4-OH-quinoline 0.85 D-Quinoline6-OH-quinoline 0.94 D-Quinoline5-OH-quinoline 0.98 D-Quinoline2-OH-quinoline 0.89 D-QuinolineQuinoline 1.02 D-Quinoline9(10H)acridon 1.04 D-Quinoline6(5H)phenanthridinon 0.99 D-QuinolineAcridine 0.97 D-Acridine5,6-Benzoquinoline 1.05 D-AcridinePhenanthridine 1.07 D-Acridine7,8-Benzoquinoline 1.04 D-AcridineNaphthalene 1.14 D-AcridineBenzo[a]acridine 1.08 D-AcridineAnthracene 1.15 D-AcridinePhenanthrene 1.05 D-AcridineBenzo[c]acridine 1.04 D-AcridineBenzo[a]anthracene 1.03 D-Acridine

S of Quinoline (1), 5-OH Quinoline (2), and 9(10H)acridone (3).

a relative standard deviation of 16%. Repeatability of the analysesincluding both extraction and chromatographic step was shown byparallel (n = 2) extraction and measurement of 9 different sedimentsamples. The relative standard deviation of those parallel measure-ments was less than 16% for azaarones, less than 21% for azaarenesand less than 40% (23% if naphthalene was excluded) for PAHs.

The average (n = 2) was used for construction of the calibra-tion line. The accuracy of calibration line points was automaticallycalculated by Analyst software and was between 90 and 110%for all points above LOQ. The linearity of the MS/MS detectorresponses was tested in the range 0.2–60 �g/l for both azaarenesand azaarones and in the range 2–5000 �g/l for PAHs. This corre-sponds to ranges of levels in real samples. Regression coefficientswere invariably equal to or higher than 0.998, the detector hav-ing a linear response for all measured values above the limitof quantification, except for Benzo[a]acridine, where a nonlinearresponse was observed at a concentration above 14 �g/l, and forBenzo[c]acridine, where an overall regression coefficient of 0.985

was obtained.

The limit of quantification (LOQ) was defined as the lowestcalibration level that had a signal to noise ratio (S/N) above 10calculated automatically using Analyst software. The LOQ for the

38 J. Brulik et al. / J. Chromatogr. A 1294 (2013) 33– 40

Fig. 3. MRM of chromatograms of calibration solution containing 60 �g/L of PANHs and azaarones and 600 �g/L of PAHs (Top) and chromatogram of a real non spikeds . 6-OH9 inolin1

rftfcoprf

t[sdr

w

ediment sample from Rotterdam-Botlek sampling location 6. 1. 4-OH-Quinoline, 2(10H)acridon, 8. 6(5H)phenanthridone, 9. D-Acridine, 10. Acridine, 11. 5,6-Benzoqu6. Anthracene, 17. Phenanthrene, 18. Benzo[c]acridine, 19. Benzo[a]anthracene.

eal matrix samples was estimated on the basis of the S/N × 10or each compound and each run and ranged approx. from 0.15o 2.53 �g/l for azaarenes and azaarones and from 22 to 1400 �g/lor PAH. The LOQ of 0.15 �g/l of 9(10H)acridon corresponds in thisase with a sample concentration of 0.05 ng/g (dw); that of 22 �g/lf benzo[a]anthracene corresponds with 6.8 ng/g (dw). Only theeaks with a signal to noise ratio ≥10 were reported as positiveesults. An overview of the LOQs in pure solvent and retention timesor all compounds is shown in Table 3.

The LOQs obtained for azaarenes were at the same level or lowerhan the values previously published for APPI and ESI ionization6,8]. Quantification limits for PAHs were higher than values pre-ented in several papers; however, they were sufficient for the

etermination of PAHs in collected samples [19,29]. Only limitedelevant data for azaarones were found in the literature [12].

Dry weight of the sediments was determined by simple dryinghen 3 times approx. 5 g of wet sediment was spread out in a thin

-Quinoline, 3. 5-OH-Quinoline, 4. 2-OH-Quinoline, 5. D-Quinoline, 6. Quinoline, 7.e, 12. Phenathridine, 13. 7,8-Benzoquinoline, 14. Naphthalene, 15. Benzo[a]acridine,

layer onto a ceramic plate and kept in laboratory conditions (20 ◦C)for 3 days until no further weight loss was recorded. The standarddeviation of parallel measurements was always less than 3%.

3.3. Field samples analyses

The optimized method was applied to the 7 sediment samplesfrom the River Rhine collected from Rotterdam-Botlek Harbor, andto 3 sediment samples collected from the River Scheldt estuarydownstream of Antwerp Harbor (Western Scheldt). Each samplewas measured twice. These locations were selected because of theintensive passage of ships and because of the intensive chemicalindustry located upstream of the rivers. The concentrations of ana-

lytes in sediment and water samples were calculated using theinternal standards D-quinoline and D-acridine and external cal-ibration lines prepared in acetonitrile. The concentrations werecalculated using the special regression line on the base of the ratio

J. Brulik et al. / J. Chromatogr. A 1294 (2013) 33– 40 39

AHs (right panel) in the sediments of the Dutch rivers from location Rotterdam 4.

old

crmtwcouTd5abwrwctattttc

TLb

Fig. 4. Mean concentrations (n = 2) of azaarenes, azaarones (left panel) and P

f peak areas of standards and internal standard. This way calcu-ated concentrations were not corrected for obtained recoveries. Allata are presented in Electronic Supplementary Material.

The results obtained for sediment samples show typical con-entration ratios between PAHs and PANHs mentioned in previousesearch [8,30]. Extremely high values for azaarenes and PAHs wereeasured in sampling location Rotterdam 2 probably because of

he spot contamination at this sampling location and this sampleas excluded from final evaluation. In general, the sum of con-

entrations of all investigated azaarenes ranged from 3 to 10%f the sum of all investigated PAHs. Concentrations of individ-al azaarones in collected samples were significantly different.he reason would be a different grade of oxidation and degra-ation processes in the different matrices. 4-hydroxyquinoline,-hydroxyquinoline, 6-hydroxyquinoline, naththalene, anthracenend phenanthrene were not extracted or concentrations wereelow the quantification limit in some samples. 2-Hydoxyquinolineas the most abundant hydroxyquinoline and its concentration

anged from 1.5 to 22.3 ng/g (dw). All other hydroxyquinolinesere found in concentrations from 0.3 to 1.6 ng/g (dw). The con-

entration of 9(10H)acridone ranged from 2.0 to 8.3 ng/g (dw) andhat of 5(6H) phenanthridone ranged from 4.8 to 25.1 ng/g (dw). Inll cases, the total concentration of hydroxyquinolines was 0.3–5imes higher than the concentration of quinoline, the concentra-ion of acridone was 2–7 times lower than that of acridine, and

he concentration of phenanthridone was 4–8 times higher thanhat of phenanthridine. Azaarenes were present in all locations inoncentrations in range from 1.0 (5.6-benzoquinoline) to 46.4 (acri-

able 3imits of quantification (LOQ) and retention times of the tested analytes obtainedy LC–APPI-MS/MS method.

Compound LOQ (�g/L) Retention time (min)

4-OH-quinoline 1.58 11.56-OH-quinoline 0.57 17.25-OH-quinoline 0.92 17.92-OH-quinoline 1.41 19.1Quinoline 1.12 28.29(10H)acridon 0.23 28.86(5H)phenanthridinon 0.46 41.6Acridine 0.48 47.65,6-Benzoquinoline 0.38 48.8Phenanthridine 0.41 49.37,8-Benzoquinoline 0.40 52.7Benzo[c]acridine 0.27 73.9Benzo[a]acridine 0.21 64.6Naphthalene 769 55.1Anthracene 76.8 69.3Phenanthrene 46.1 71.9Benzo[a]anthracene 31.9 82.8

Fig. 5. Sum of mean concentrations of azaarenes, azarones, and PAHs at differentlocations and periods of years.

dine) ng/g (dw). PAHs were present in range of concentrations from29.0 (anthracene) to 1052 (naphthalene) ng/g (dw) An example ofthe concentrations of all compounds from location Rotterdam 4, ispresented in Fig. 4.

The ratios between the concentrations of azaarenes andazaarones are comparable with the results published by de Voogtand Laane, who assessed the presence of azaarenes and azaaronesin the Dutch coastal zone by GC–MS [4], although only sedimentswith a fraction less than 63 �m were used in that work. The ratiobetween azaarenes and PAHs in sediments observed in the presentstudy is in agreement with results from Lintelman et al. [6] obtainedfor airborne particles.

The highest sum of average concentrations of PAHs and PANHswas in samples from Rotterdam Harbor collected in April (seeFig. 5), but the highest sum of concentrations of azaarones at thesame location was observed in November. This difference may becaused by the microbial and photo degradation of azaarenes, whichis more effective during the summer period than in the winter. Theresults obtained for real samples in the present study demonstratethe applicability of the LC–APPI-MS/MS method for the determina-tion of these analytes in environmental samples at environmentallyrelevant concentrations.

4. Conclusions

The LC–APPI-MS/MS method was developed, optimized andvalidated for the determination of selected azaarenes and their

degradation products (azaarones) in a single run. Additionally,selected PAHs can also be measured simultaneously by this method.The method is highly sensitive and selective for azaarenes andazaarones, in particular. Soxhlet extraction and SPE extraction were

4 atogr

urfiRiet

A

(

pa

A

f2

R

[[

[

[

[[

[

[[

[[[

[[[

[

[[

0 J. Brulik et al. / J. Chrom

sed for sample preparation. The reliability of the method for envi-onmental monitoring and routine analyses was demonstrated oneld sediment samples from the Dutch Rhine and Western Scheldtivers. This method provides a new, efficient and easy tool for mon-

toring the content of azaarenes and azaarones in the environment,specially for the investigation of potential impacts associated withheir presence in the aquatic environment.

cknowledgements

The research was supported by the Czech Ministry of EducationMSM0021622412) and by the project CETOCOEN (ED 0001/01/01).

Jan Brulík gratefully acknowledges the ERASMUS mobility forlacement grant that provided the support for a visiting scholarshipt IBED-University of Amsterdam.

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.chroma.013.03.079.

eferences

[1] J.P. Kaiser, Y.C. Feng, J.M. Bollag, Microbiol. Rev. 60 (1996) 483.[2] J.M. Schmitter, H. Colin, J.L. Excoffier, P. Arpino, G. Guiochon, Anal. Chem. 54

(1982) 769.

[3] M. Blumer, T. Dorsey, J. Sass, Science 195 (1977) 283.[4] P. de Voogt, R.W.P.M. Laane, Chemosphere 76 (2009) 1067.[5] B. Janoszka, Food Chem. 126 (2010) 1344.[6] J. Lintelmann, M.H. Franca, E. Hubner, G. Matuschek, J. Chromatogr. A 1217

(2010) 1636.

[[

[

. A 1294 (2013) 33– 40

[7] K. Minomo, N. Ohtsuka, K. Nojiri, Y. Kurata, M. Karaushi, Y. Isobe, J. Mater. CyclesWaste Manage. 11 (2009) 73.

[8] R. Svabensky, M. Oravec, Z. Simek, Int. J. Environ. Anal. Chem. 89 (2009)167.

[9] E.A.J. Bleeker, S. Wiegman, P. de Voogt, M. Kraak, H.A. Leslie, E. de Haas, W.Admiraal, Reviews of Environmental Contamination and Toxicology, vol. 173,Springer-Verlag, New York, 200239.

10] I. Negron-Encarnacion, R. Arce, Atmos. Environ. 41 (2007) 6771.11] S. Wiegman, P.L.A. van Vlaardingen, E.A.J. Bleeker, P. de Voogt, M.H.S. Kraak,

Environ. Toxicol. Chem. 20 (2001) 1544.12] A.K. Reineke, T. Goen, A. Preiss, J. Hollender, Environ. Sci. Technol. 41 (2007)

5314.13] M.L. Paumen, P. de Voogt, C.A.M. van Gestel, M.H.S. Kraak, Sci. Total Environ.

407 (2009) 4605.14] M. Mundt, J. Hollender, J. Chromatogr. A 1065 (2005) 211.15] R.E. Cochran, N. Dongari, H. Jeong, J. Beranek, S. Haddadi, J. Shipp, A. Kubatova,

Anal. Chim. Acta 740 (2012) 93.16] N. Barco-Bonilla, R. Romero-Gonzalez, P. Plaza-Bolanos, J.L. Fernandez-Moreno,

A.G. Frenich, J.L.M. Vidal, Anal. Chim. Acta 693 (2011) 62.17] B.M. Showalter, J.R. Locker, Abstr. Pap. Am. Chem. Soc. 207 (1994) 173.18] R. Svabensky, K. Koci, Z. Simek, Int. J. Environ. Anal. Chem. 87 (2007)

337.19] M. Smoker, K. Tran, R.E. Smith, J. Agric. Food Chem. 58 (2010) 12101.20] O. Delhomme, M. Millet, Polycyclic Aromat. Compd. 28 (2008) 518.21] M. Amati, S. Belviso, P.L. Cristinziano, C. Minichino, F. Lelj, I. Aiello, M. La Deda,

M. Ghedini, J. Phys. Chem. A 111 (2007) 13403.22] J. Kochany, R.J. Maguire, Chemosphere 28 (1994) 1097.23] J. Blotevogel, A.K. Reineke, J. Hollender, T. Held, Grundwasser 11 (2006) 295.24] J. Neuwoehner, A.K. Reineke, J. Hollender, A. Eisentraeger, Ecotoxicol. Environ.

Saf. 72 (2009) 819.25] L.G. Song, D.S. Cho, D. Bhandari, S.C. Gibson, M.E. McNally, R.M. Hoffman, K.D.

Cook, Int. J. Mass Spectrom. 303 (2011) 173.26] D.B. Robb, T.R. Covey, A.P. Bruins, Anal. Chem. 72 (2000) 3653.27] Y.S. Ding, D.L. Ashley, C.H. Watson, J. Agric. Food Chem. 55 (2007) 5966.

28] J.A. Syage, J. Am. Soc. Mass Spectrom. 15 (2004) 1521.29] H. Moriwaki, M. Ishitake, S. Yoshikawa, H. Miyakoda, J.F. Alary, Anal. Sci. 20

(2004) 375.30] K. Kobeticova, J. Buchlebova, J. Lana, I. Sochova, J. Hofman, Ecotoxicol. Environ.

Saf. 71 (2008) 650.