Validating Analytical Protocols to Determine Selected ...downloads.hindawi.com/journals/jamc/2017/9796457.pdf · 4 JournalofAnalyticalMethodsinChemistry Concentrate under nitrogen

Research ArticleValidating Analytical Protocols to Determine Selected Pesticidesand PCBs Using Routine Samples

Oscar Pindado Jimeacutenez Susana Garciacutea Alonso and Rosa Mariacutea Peacuterez Pastor

Chemistry Division CIEMAT Av Complutense 40 28040 Madrid Spain

Correspondence should be addressed to Oscar Pindado Jimenez oscarpindadociemates

Received 4 July 2017 Accepted 18 September 2017 Published 29 October 2017

Academic Editor Paolo Montuori

Copyright copy 2017 Oscar Pindado Jimenez et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

This study aims at providing recommendations concerning the validation of analytical protocols by using routine samples It isintended to provide a case-study on how to validate the analytical methods in different environmental matrices In order to analyzethe selected compounds (pesticides and polychlorinated biphenyls) in two different environmental matrices the current work hasperformed and validated two analytical procedures by GC-MS A description is given of the validation of the two protocols by theanalysis of more than 30 samples of water and sediments collected along ninemonthsThe present work also scopes the uncertaintyassociated with both analytical protocols In detail uncertainty of water sample was performed through a conventional approachHowever for the sedimentsmatrices the estimation of proportionalconstant bias is also included due to its inhomogeneity Resultsfor the sediment matrix are reliable showing a range 25ndash35 of analytical variability associated with intermediate conditions Theanalytical methodology for the water matrix determines the selected compounds with acceptable recoveries and the combineduncertainty ranges between 20 and 30 Analyzing routine samples is rarely applied to assess trueness of novel analytical methodsand up to now this methodology was not focused on organochlorine compounds in environmental matrices

1 Introduction

Organochlorine pesticides (OC) and polychlorinated bi-phenyls (PCBs) are of special interest because of their persist-ence and bioaccumulation in the environment [1ndash4] Thisresults in having been categorized as persistent organic pollu-tant compounds since 1995 by United Nations EnvironmentProgramme Concern over pesticides and PCBs in environ-ment has increased during last decades Consequently sev-eral studies have been conducted to increase the knowledgeof the presence and effects related to these compounds [5ndash8] At the beginning many of the published works had theaim of quantifying as many pollutants as possible Nowadaysstudies are more concise for example about determinationof pesticides residues from water [9] the occurrence of PCBsin sediments [10] transport of pesticides along a river [11 12]or its presence in coastal areas [13]

There is an increasing demand for new analyticalmethodsfor monitoring and quantifying pesticides and PCBs at tracelevels Due to the complexity of environmental matrices and

the low concentrations of these compounds determination ofpesticides and PCBs usually carried out many steps such asconcentration separation or extraction processes followedby gas chromatography-mass spectrometry detection [14ndash16]Therefore possibility of losses or specific contamination oftenoccurs Conventional methods such as Soxhlet extraction orliquid-liquid extraction can be improved to adjust to matri-ces characteristics It is important that the developed meth-ods should be simple fast efficient and reliable The newchemical analysismust focus on reducing sample size solventvolumes and cost of analysis It is necessary to highlight thatdeveloped methods must be always validated and their relia-bilitymust be also proved It is well known that determinationof organic compounds into interest environmental matricesis complex and shows a high variability Some studies haveestablished variability around 30ndash50 when these com-pounds are analyzed at trace level [17ndash19] Nevertheless thenumbers of studies that have evaluated the analytical uncer-tainty are still less than desired [4 20ndash23]

HindawiJournal of Analytical Methods in ChemistryVolume 2017 Article ID 9796457 13 pageshttpsdoiorg10115520179796457

2 Journal of Analytical Methods in Chemistry

Within the framework of CIEMAT research an analyticalmethodology might be performed in order to quantify pesti-cides and PCBs in environmental samples including a widerange of concentrations The aims of this study were on theone hand developing two analytical methods to determinethe selected pesticides and PCBs as fast robust and simpleas possible in waters and sediments and on the other handdemonstrate the reliability of those methods by means of theanalysis of routine samples It is necessary to highlight thatthe present work is also intended to provide an easy to followcase-study on how to validate analytical protocols in differentenvironmental matrices using routine samples

Development and application of analytical methods fordetermination of organochlorine compounds in severalmatrices have already been published [9 17 24 25] Takinginto account the inherent characteristics of each matrix theestimation of possible constantproportional bias should bealso carried out to check the quality of analytical resultsTherestudies aim to formulate mathematical algorithms to arrive atfull validation but sometimes they are difficult to apply [26]

There is still a lack of information on the variabilityassociated with concentration level matrix of samples oranalysis dates To our knowledge this approach is rarelyapplied for environmental studies and has not yet beendone to the determination of organochlorine compounds insediments and watersThe present work can be considered tobe of the highest relevance in order to evaluate quality of anyanalytical method

2 Materials and Methods

21 Reagents andMaterials Pesticides standardweremixtureof 8 pesticides (pesticide mix-5 of 10 ng 120583Lminus1 in cyclohexane)mixture of 4 pesticides (pesticideMix 7 of 10 ng 120583Lminus1 in cyclo-hexane) mixture of 6 pesticides (pesticide Mix 31 of10 ng 120583Lminus1 in acetone) mixture of 26 pesticides and PCBs(pesticide Mix 33 of 10 ng 120583Lminus1 in isooctane) and penta-chlorobenzene and pentachlorophenol (10 ng 120583Lminus1 in cyclo-hexane) fromDr Ehrenstorfer GmbH (Augsburg Germany)Anthracene deuterated (10 g 120583Lminus1 in cyclohexane) from DrEhrenstorfer GmbH (Augsburg Germany) and o-terphenylfrom Supelco (Bellefonte USA) were used as surrogateand internal standard respectively Two certified referencematerials were used CNS391 (CRM) of pesticides andPCBs in sediment of natural water and WatRTM pollutionorganochlorine pesticides 709 (ERA) Acetonitrile dichloro-methane ethyl acetate and hexane of residue analysis gradewere supplied by J T Baker (Deventer Holland) Sodiumsulfate anhydrous was obtained from Merck (DarmstadtGermany)

The SPE procedure was performed using Oasis HLBExtraction Cartridge (6 cc and 500mg) acquired to Waters(Massachusetts USA) in a vacuum manifold Agilent (SantaClara USA)

22 Instruments AnAgilent 6890 chromatograph with auto-matic sampler and a splitsplitless programmed temperatureinjector coupled to an Agilent 5975 mass spectrometer was

employed Operating conditions were carrier gas heliumcapillary column HP-5MS (30m times 025mm times 025 120583m) withflow rate of 1mLminminus1 PTV injector with initial temperature110∘C held for 05min and then increased at a rate of500∘Cminminus1 to 200∘C operation mode splitless with 3 rowinjections of 1 120583L initial oven temperature 100∘C held for4min increased at a rate of 8∘Cminminus1 to 250∘C held for1min and finally increased at a rate of 8∘Cminminus1 to 290∘Cand detector operated in electronic impact mode (70 eV) inSIM mode

23 Sampling To evaluate the level of selected pesticides andPCBs during the time period among March 2015 to October2015 weekly samples of sediment and water were collectedfrom a confined area of continental waters in which therewere industrial discharges

Water samples were collected in 1-liter glass bottles andstored at 4∘C until analysis 100 g of sediment samples wascollected in amber glass bottles and stored at 4∘C untilanalysis without any additional treatment

24 Description of Analytical Method

241 Water Samples Figure 1 shows a scheme of the pro-posed methodology to determine selected compounds in awater matrix In this method 1 L of continental water spikedwith 5 120583L of anthracene deuterated (10 ng120583Lminus1) as surrogateis employed to determine the selected organochlorine pesti-cides (pp1015840-DDD op1015840-DDD pp1015840-DDE op1015840-DDE pp1015840-DDTop1015840-DDT 120572-HCH 120573-HCH 120574-HCH atrazine simazinemetolachlor terbuthylazine aldrin dieldrin endrin isodrinand hexachlorobenzene) Firstly cartridges were conditionedwith 12mL of dichloromethane and 12mL of acetonitrileand later were rinsed with 12mL of water Secondly whencartridges are still wet water sample is passed through themat a flow rate of 5ndash10mLminminus1 Cartridges were washed with6mL of water and then dried for 20min under a stream ofnitrogen Later extracts were eluted with 25mL of acetoni-triledichloromethane (1 1) and 32mL of dichloromethaneThen extracts were concentrated with nitrogen to 3mLand 1 g of anhydrous sodium sulfate was added to take outhumidity Finally extracts were concentratedwith nitrogen to90 120583L of ethyl acetate and 10 120583L of internal standard solution(10 ng120583Lminus1 of o-terphenyl in hexane) was added and injectedinto the GC-MS

242 Sediment Samples Figure 2 shows a scheme of theproposed methodology to analyze sediments matrix Themethod was optimized for determining the selected organicpesticides and PCBs (pp1015840-DDD op1015840-DDD pp1015840-DDEop1015840-DDE pp1015840-DDT op1015840-DDT 120572-HCH 120573-HCH 120574-HCHhexachlorobenzene pentachlorobenzene PCB-28 PCB-52PCB-101 PCB-118 PCB-138 PCB-153 and PCB-180) Themethod begins as follows 05ndash2 g of sediment spiked with10 120583L of anthracene deuterated (10 ng 120583Lminus1) as surrogatewas weighed and 25mL of hexaneacetone mixture wassubsequently added The first extraction was performed byshaking the mixture for 1min in a Vortex and later goes

Journal of Analytical Methods in Chemistry 3

Dry cartridge20 minutes

Chromatographic analysisGCMS (SIM)

River water1 litre

Add standard surrogate5 uL an D10 (10 ngmL)

Add internal standard(10 ngmL)10 uL o-terphenyl

SPEOasis HLB

(6 cc 500 mg)

Active cartridge(1) 12 mL （2l2(2) 12 mL acetonitrile

Elute cartridge

(2) 32 mL （2l2(1) 25 mL acto（2l2

Condition cartridge12 mL water

Concentrate under nitrogento 3 mL

Concentrate under nitrogento 90 uL

Wash cartridge6 mL water

Load sample5ndash10 mLmin(5---15 in Hg)

Remove moisture1-2 g sodium sulfate anhydrous

Figure 1 Scheme of the proposed analytical method to quantify the selected organochlorine pesticides in water

toward an ultrasonic bath for 10 minutes The extracts werecentrifuged for 3 minutes and two phases were obtainedThe top phase contains organic phase and the bottom phase(aqueous phase and sediment) was extracted twice again withthe mixture hexaneacetone All organic phases were gettingtogether and the final extract was concentrated with nitrogento 1mL of hexane and 100 120583L of internal standard solution(10 ng 120583Lminus1 of o-terphenyl in hexane) was added and injectedinto the GC-MS

25 Preparation of Spiked Sediment Samples Two sedimentssamples containing a low chlorinated mass fraction werechosen to prepare spiked samplesThe spikingwas conducted

by adding a specific volume of chlorinated standard solutionsto the sediments In detail volumes of 50 and 200 120583L ofstandard solutions were added to 1 gram of sediments Spikedsubsamples were agitated for a short time using automaticshaker and left 1 h for solvent evaporation Trials of spiking oneach sample were conducted according to the flow diagrampresented in Figure 1S (in Supplementary Material availableonline at httpsdoiorg10115520179796457)

26 Mathematical Concepts Related to Constant and Propor-tional Bias According toMaroto et al [27ndash29] the analyticaltrueness can be assessed by recovery trials using routinesamples and analyses of different sample masses Taking into


Concentrate under nitrogento dryness


Organic phase (top)Reserve

Add standardsurrogate

10 uL an D10 (10 ngmL)

Sediment05ndash2 grams

Add solvent25 mL hexaneacetone

(1 1)

times2

Solid + aqueousphase (lower)

hexaneacetone (1 1)Add 1 mL

Dissolve1 mL hexane

STIRVortex + ultrasounds

(10 min)

Centrifugate

3 min2500 rpm


Figure 2 Scheme of the proposed analytical method to quantify the selected organochlorine pesticides and PCBs in sediments

account that the results of sediment analysis vary within awide range of mass fraction the proportional and constantbias must be addressed

261 Proportional Bias If the bias of analytical procedurevaries with concentration it can be assumed there is pro-portional bias It can be estimated performing recoverymeasurements with spiked samples The proportional bias iscalculated from the recovery by (1) Specifically the presentwork has performed two spiked levels (50 and 200 ng)

119877119894 = 119908found(119894)119908added(119894) (1)

where 119908found(119894) refers to the mass fraction calculated bysubtracting the mass fraction of selected organochlorinecompounds measured in the spiked subsamples from that in

unspiked subsample The added mass fraction is denoted as119908added(119894)A significance test is then used to determine whether the

mean overall recovery (119877) is significantly different from 10That is calculated as the average of119877119894 related to the two spikedlevels The test is also used to verify if the ratio between thevalue of 119877 and its uncertainty 119906(119877) is significant higher thanthe corresponding tabulated two-side value The test statistic(119905) was deduced according to

119905119877 = |119877 minus 1|119906 (119877) (2)

where 119906(119877) is calculated as a combined relative standarddeviation 119904119903(119894) at each mass fraction level (spiked samplesat low level and at high level) and considering also thecorresponding results from analyses of reference materialaccording to

119906119877mean = radic1199042119903 (ref mat) + 1199042119903 (LSpk 119860) + 1199042119903 (LSpk 119861) + 1199042119903 (HSpk 119860) + 1199042119903 (HSpk 119861) (3)


119904119903(119894) is calculated from the independentmeasurements corre-sponding to each subsample mass and divided by the squareroot of number of repetitions (119899 = 4 for reference materialand 119899 = 2 for each investigated mass fraction level of samples119860 and 119861)

On the basis of the study performed along this paper thetabulated student statistic at a probability 120572 = 005 and 11degrees of freedom is 119905tab = 220262 Constant Bias The constant bias of the analyticalprocedure can be evaluated by analyzing two differentmassesof samples and checking if both samples deliver the sameresult It is estimated with the Youden method [29 30]according to (4)

119909119894119887cte119886119894 = 119909119895119887cte119886119895 119887cte = (119886119894119909119895 minus 119886119895119909119894119886119894 minus 119886119895 )

(4)

where 119887cte denotes the constant bias (expressed in ng) 119886119894 and119886119895 are themean values of the twomasses of the subsamples forcomparison and 119909119894 and 119909119895 are the mean masses of quantifiedchlorinated compounds (ng)

To assess whether the constant bias is significant itsuncertainty must be obtained according to (5) More detailscan be seen in a previous work already published [27]

119906119903 (119887cte) = 1(119886119894 minus 119886119895)radic(119886119895119906 (119909119894)119909119894 )2 + (119886119894 119906 (119909119895)119909119895 )2 (5)

Similar to the above described the significant test must beused to evaluate the constant bias by applying

119905119894 = 119887cte119909119894119906119903 (119887cte) (6)

The present work has evaluated three different masses 0616 and 26 grams of sediments comparing them in pairsThetabulated student statistic at a probability 120572 = 005 and 4degrees of freedom is 119905tab = 2783 Results and Discussions

31 Validation of Analytical Procedures During a time periodof nine months the linearity of the GC-MS device wasexamined over the range 50ndash1000 ng Lminus1 which is in linewith the environmental levels for pesticides currently foundin the literature The calculation of a four-point linear plotestablishes several linear regression and squared correlationcoefficient The GC-MS method showed good linearity withregression correlation coefficients 1199032 gt 0990 for all com-pounds along the studied period In case of op1015840-DDT pp1015840-DDT and several PCBs linearity was slightly lower Theauthors have associated it with generation of active pointsinto the liner through repeated injections These compounds

would be degraded so that linearity can be affected Moredetails can see in the supplementary material (Table S3)

Detection limits for each compound were calculated asthree time the standard deviation of four spiked samplesat a final concentration of 10 ng Lminus1 The LOD and LOQare summarized in Table S3 of supplementary material Thelimits were calculated according to IUPAC recommendationand the obtained values were below the first calibration level

Accuracy was evaluated by analyzing two certified ref-erence materials ERA 709 of several pesticides in waterand CNS391 of several pesticides and PCBs in sedimentsAnalyzed compounds showed values in the performanceacceptance limits at 95 confidence interval so accuracy isdemonstrated Results are summarized in the supplementarymaterial (Table S4)32 Assessing Trueness Study of Recovery and Bias

321WaterMatrices In order to perform the recovery studysix uncontaminatedwater samples (500mL)were spikedwiththe pesticide standards at three fortification levels (250 500and 1000 ng Lminus1) The lower and higher levels were spikedwith standards Mix 5 Mix 7 andMix 3 and the intermediatelevel was spiked with the standard Mix 33 Each fortificationlevel was analyzed along three different weeksThe aimwas tohighlight how to vary the analytical response along the timeMean recovery data obtained are summarized in Table 1

Along those three weeks the GCMS device was inideal conditions at the beginning and proceeds to nonidealconditions with time It is noticeable that recoveries obtainedin that period ranged from40 to 120with a relative standarddeviation below 20 for the majority of compounds On thebasis of those results the variation of analytical responsewith time affects mainly isomers of DDT This fact can beexplained thought generation of active points into injectiondevice This compound could be degraded into injectionsystem of the gas chromatograph In the case of recoveriesbelow 40 determination of these compounds must bemodified

322 Sediment Matrices The sediment samples are morecomplex and less homogeneous than water samples so therecovery study must be performed in more detail Hence theproportional and constant biases are evaluated by employingroutine samples spiked samples and also referencematerials

Two samples with low mass fraction are selected in orderto evaluate the proportional bias [4 27ndash30] The routinesamples are divided into three subsamples One subsampleis not spiked while the other two are spiked at two differentlevels as was described in the preparation of spiked sedimentsamples section (Figure S1) Also the bias study is carried outon two different days

Table 2 shows the mean of recoveries their relative stan-dard deviation and the statistic ldquo119905rdquo parameter to determinethe significance test These parameters are obtained fromthe three measurement results at each mass fraction levelWhen the value of 119905exp is higher than the tabulated two-sidevalue (119905tab = 220 at a probability of 120572 = 005 119899 = 11)


Table 1 Recovery study for the analysis of six spiked water samples at three levels of concentration in different weeks

Pesticide Recovery ()Week 1 (500 ng Lminus1) Week 2 (1000 ng Lminus1) Week 3 (250 ng Lminus1)

Aldrin 59 plusmn 14 mdash mdashAtrazine mdash 117 plusmn 14 110 plusmn 16op1015840-DDD mdash 60 plusmn 8 51 plusmn 8pp1015840-DDD 89 plusmn 23 70 plusmn 10 61 plusmn 6op1015840-DDE mdash 41 plusmn 4 33 plusmn 1pp1015840-DDE 73 plusmn 21 45 plusmn 6 34 plusmn 1op1015840-DDT mdash 43 plusmn 4 36 plusmn 4pp1015840-DDT 75 plusmn 30 68 plusmn 8 50 plusmn 9Dieldrin 101 plusmn 24 88 plusmn 21 75 plusmn 10Endrin 123 plusmn 21 73 plusmn 9 89 plusmn 20Hexachlorobenzene 78 plusmn 20 mdash mdash120572-HCH 111 plusmn 11 114 plusmn 28 117 plusmn 17120573-HCH 117 plusmn 12 106 plusmn 12 116 plusmn 28120574-HCH 113 plusmn 14 112 plusmn 27 116 plusmn 23Isodrin 64 plusmn 16 mdash mdashMetolachlor mdash 91 plusmn 9 109 plusmn 11Simazine mdash 113 plusmn 12 120 plusmn 19Terbuthylazine mdash 76 plusmn 9 109 plusmn 14

the proportional bias should be considered significant soa correction factor (1119877119894) might be applied to correct theanalytical result

According to the summarized values in Table 2 therecovery from pp1015840-DDT 120572-HCH 120573-HCH and 120574-HCH isstatistically different of 100 In the case of pp1015840-DDT anobvious matrix effect is observed in the recovery of spikedmass fraction While recovery values in the first sample trialswere acceptable (sim90) recovery results were reduced toa poor 20 in the second sediment sample trials Theseobservations can be related to the high susceptibility of thecompound to have degradation losses into injection systemof the chromatograph On the other hand differences for 120572-HCH 120573-HCH and 120574-HCH can be attributed to their highervolatility compared to other analyzed compounds

The constant bias study is performed by testing a rou-tine sample heterogeneous with a high mass fraction oforganochlorine compoundsThis sample is divided into threesubsamples of 06 g 16 g and 26 gThe analyses are done onalternate days by triplicate More details can be seen in thesupplementary material (Tables S1-S2 and Figure S2)

Table 3 compiles the values of the constant bias (119887cte) itsrelative standard uncertainty 119906119903(119887cte) and the correspondingvalues of 119905119894 When 119905119894 was below the tabulated student statistic(119905tab = 278 at a probability 120572 = 005 and 4 degrees of free-dom) no significant constant bias is found However if thevalue of 119905119894 is above that tabulated value there is a significantconstant bias

As can be seen in Table 3 significant constant bias isobtained for pp1015840-DDD pp1015840-DDE 120572-HCH 120574-HCH PCB-138 PCB-153 and PCB-180 when the analyzed subsampleis lower than 1 gram However if the mass of the analyzed

subsample is greater than 1 gram the constant bias is insignifi-cant with the exception of120572-HCHdue to his higher volatilityTherefore it can be concluded that analyzing subsampleslower than 1 gram implies significant constant bias in resultsSo it is recommended to analyze subsamples higher than2 gramsThismass is the best choice when this heterogeneoussediment is analyzed

33 Estimation of Uncertainty There are many sourcesof uncertainty associated with an analytical measurementHowever not all the components have a significant contri-bution to it The cause and effect diagram facilitates identi-fication of each contribution The electronic supplementarymaterial (Figures S3-S4) shows it and each mathematicalequation In this study we have identified three maincontributions to uncertainty the concentration measuredthrough a GCMS calibration the recovery study and theintermediate precision The remainder contributions havebeen already estimated elsewhere [20 31 32] and showed lowvalues so those were omitted in this study

331 ConcentrationMeasured through Calibration This con-tribution (119906119909) is affected by the preparation of each standard(119906119862std) and also by the chromatographic calibrates (119906119862119909) fromlinear least squares

Uncertainty associated with preparation of each calibra-tion solution includes the uncertainty of the stock solutionaccording to certificates and the uncertainty of dilutionchain The uncertainty associated with stock solution is sup-plied by the manufacturer as plusmn1 Contributions of dilutionchain include repeatability temperature and specification


Table 2 Values of recovery and relative standard deviations (expressed in ) obtained from the analysis of two sediments spiked at two levelsof concentration

Compound Spiked 50 ng Spiked 200 ng 119877mean () 119906(119877mean) () 119905Sediment 1 () Sediment 2 () Sediment 1 () Sediment 2 ()

pp1015840-DDD 129 plusmn 57 104 plusmn 58 124 plusmn 29 82 plusmn 10 109 97 071pp1015840-DDE 103 plusmn 16 75 plusmn 12 92 plusmn 39 82 plusmn 57 88 73 162pp1015840-DDT 96 plusmn 13 26 plusmn 35 88 plusmn 16 15 plusmn 15 56 36 12HCB 60 plusmn 18 122 plusmn 200 59 plusmn 70 100 plusmn 61 85 210 074120572-HCH 92 plusmn 88 85 plusmn 17 56 plusmn 52 87 plusmn 25 72 90 306120573-HCH 134 plusmn 63 126 plusmn 52 123 plusmn 60 126 plusmn 47 127 81 337120574-HCH 71 plusmn 10 89 plusmn 53 67 plusmn 43 93 plusmn 10 80 65 310PCB-28 108 plusmn 26 137 plusmn 60 99 plusmn 58 89 plusmn 160 108 49 164PCB-52 112 plusmn 11 109 plusmn 29 101 plusmn 27 111 plusmn 75 108 150 054PCB-101 120 plusmn 10 101 plusmn 33 112 plusmn 35 104 plusmn 43 109 120 077PCB-118 113 plusmn 40 103 plusmn 36 118 plusmn 24 104 plusmn 16 110 102 093PCB-138 120 plusmn 15 90 plusmn 46 114 plusmn 11 92 plusmn 11 104 119 035PCB-153 120 plusmn 10 94 plusmn 37 113 plusmn 19 97 plusmn 15 106 84 071PCB-180 129 plusmn 12 82 plusmn 83 123 plusmn 10 79 plusmn 11 103 71 048

Table 3 Study of constant bias by means the analysis of three sediment subsamples of 06 16 and 26 grams respectively

(ng) Subsample 1 Subsample 2 Subsample 3 119887cte 119906(119887cte) 119905(119887cte)1199091=06 1199092=16 1199093=26 1199091-1199092 1199092-1199093 1199091-1199092 1199092-1199093 1199091-1199092 1199092-1199093op1015840-DDD 44 plusmn 52 102 plusmn 52 160 plusmn 28 10 minus75 0109 0209 22 04pp1015840-DDD 45 plusmn 34 99 plusmn 12 143 plusmn 83 14 minus29 0080 0195 37 15op1015840-DDE 43 plusmn 026 11 plusmn 041 17 plusmn 34 047 minus058 0059 0231 19 02pp1015840-DDE 23 plusmn 21 52 plusmn 16 78 plusmn 11 65 minus89 0085 0165 33 10HCB 3524 plusmn 209 8451 plusmn 198 13832 plusmn 421 594 154 0055 0049 31 04120572-HCH 29 plusmn 006 15 plusmn 26 41 plusmn 60 minus41 27 0066 0319 22 59120573-HCH 61 plusmn 46 145 plusmn 29 261 plusmn 16 11 40 0098 0306 18 09120574-HCH 82 plusmn 17 18 plusmn 19 40 plusmn 12 25 18 0190 0373 58 27PCB-28 442 plusmn 49 997 plusmn 77 1607 plusmn 138 112 minus23 0105 0152 24 01PCB-52 143 plusmn 18 352 plusmn 25 576 plusmn 59 19 71 0119 0158 16 01PCB-101 68 plusmn 58 178 plusmn 16 252 plusmn 28 28 minus60 0085 0135 05 25PCB-118 30 plusmn 50 63 plusmn 44 104 plusmn 55 11 15 0154 0120 23 02PCB-138 33 plusmn 015 77 plusmn 059 13 plusmn 11 065 14 0050 0149 39 12PCB-153 10 plusmn 063 22 plusmn 21 40 plusmn 30 30 52 0065 0162 45 14PCB-180 14 plusmn 21 25 plusmn 26 38 plusmn 34 72 minus38 0144 0186 36 08

limits of syringes According to previous studies contribu-tions of dilution chain were not significant As a result therelative uncertainty associated with each concentration ofcalibration standard (119906119862std119862std) was estimated as 002 (1supplied by Dr Ehrenstorfer 50120583gsdotmLminus1) and was used forall standards in the studied concentration range

The uncertainty from linear least squares calibration ismainly due to variability in the responses of the GCMSThe uncertainty is evaluated from the standard deviations ofslope and intercept in calibration line For this a set of fourconcentration levels is independently prepared and analyzedfor a period of 4months In definitive the relative uncertaintyof the predicted analyte concentration (119906119862119909119862119909) from linearleast squares calibration is calculated applying

11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)

where ldquo119910rdquo is the mean value of the analytical response offour standard solutions (50 ngmLminus1 of selected organochlori-nated) analyzed for a period of 4 months ldquo119906119910rdquo is uncertaintyapplying the expression for the lineal regression of leastsquares ldquo119887rdquo is the 119910-intercept of the calibration graph ldquo119906119887rdquois the uncertainty deduced from standard deviation of 119910-intercept ldquo119898rdquo is the slope of calibration graph and ldquo119906119898rdquo is theuncertainty of slope


In the electronic supplementary material each detailrelated to uncertainty associated with concentration mea-sured through calibration by GCMS is thoroughly described(Tables S5ndashS7) Table 4 summarizes the uncertainty associ-ated with concentration measured through calibration

The uncertainty associated with the concentration mea-sured through calibration (119906119909119909) is about 15ndash20 for mostof the selected organochlorinated compounds Uncertaintiesfrom the calibration graphs are clearly higher than thoseof standards (Table 4) It is noticeable that measurementsof pentachlorobenzene isodrin dieldrin endrin and PCBsnumbers 101 118 and 138 rise to 30ndash50 for the whole periodThese results are consequence of the occurrence with time ofactive points into liner which degrades several compoundsand so a low analytical sensitivity for those It is important topoint out that this uncertainty study was performed on inter-mediate terms So those values are a clear proof of the require-ment to analyze standards and sample extracts on the sameday or days very close trying to keep the same chromato-graphic conditions as liner columns state and others

332 Recovery Study In environmental studies there arerecovery factors different from 100 So recovery is a signifi-cant uncertainty source In order to estimate this contribution(119906rec) the recovery study described above was employedFor water samples the standard relative deviation from therecovery study of uncontaminated water samples spiked withpesticides at three fortification levels was employed Mean-while for sediment samples the standard relative deviationfrom reference material and spiking assays were used Moredetails can be seen in the supplementary electronic material(Tables S8 and S9) Table 5 summarizes the final results

As can be seen the influence of the matrix type in thedispersion of analytical results is more significant in the caseof sediment samples with values around 10 while those forwater samples are mainly around 5

333 Intermediate Precision In water samples the interme-diate precision is assessed by analyzing by duplicating a setof water samples spiked at three concentration levels for aperiod of three weeksThe supplementary electronicmaterial(Table S10) shows more details The intermediate precisionexpressed as pooled relative standard deviation was deducedas follows

RSDpooled = radicsumRSD2119894 (119899119894 minus 1)sum (119899119894 minus 1) (8)

Being grouped the relative standard deviations (RSD) of theanalytical results by sample are replicated The uncertaintyassociated with the intermediate precision for water sampleswas calculated as follows

119906 (RSDpooled) = rsdpooled (9)

Table 6 summarizes the results According to that values arelower than 20

In the case of sediment samples the intermediate preci-sion is also calculated from the analyses of a set of 33 samples

collected in the area of study which are analyzed by duplicatefor a period of 9 months More details can be seen in theelectronic supplementary material (Table S11) Results ofuncertainty associated with intermediate precision are sum-marized in Table 7

According to values summarized in Table 7 the interme-diate precision ranges between 25 and 40 for most of theanalyzed compounds These values are in line with expectedlevels for the analysis of organic samples in complex andheterogeneous samples It is noteworthy that precision is eval-uated with samples collected in the study area along 9monthsand are analyzed by duplicate And thus this study is a clearand detailed item to demonstrate variability and difficultyassociated with these analyses

Particularly the op1015840-DDT and pp1015840-DDT show the high-est values of RSD It is well known that these compounds canbe degraded into injection system causing analytical prob-lems Also the determination of both compounds in theanalyzedmatrices is at trace level close to detection limit andin a reduced number of samples The authors have identifiedthese issues as responsible for those levels In the case ofpentachlorobenzene the value measured is due to volatilityof this compound which can generate losses along theextraction process

334 Estimation of Total Uncertainty Once the three mainuncertainty contributions are calculated the total uncertaintycan be achieved from different waysThe authors have clearlyreflected the different approach to be adopted to analyze theselected compounds in differentmatricesThemain problemsof sediments matrices are inhomogeneity and in case of thewater low concentrations

The total uncertainty associated with the determinationof selected organochloride compounds from water samplesis estimated as a combined uncertainty The relationshipbetween the combined standard uncertainty and the uncer-tainty of the independent contributions on which it dependsis 119906119888 (Pesticideswater)

= radic119906119903 (cal)2 + 119906119903 (rec)2 + 119906119903 (rsdpooled)2 (10)

Figure 3 displays the combined uncertainty for the determi-nation of each organochlorine compound in water As can beseen the combined uncertainty shows values lower than 30for many compounds The uncertainty due to recovery hasgot the least influence in the final results However thisuncertainty could be reduced somewhat This contributioncan be associated with losses of more volatile compoundsduring concentration under a gentle stream of nitrogen andalso with incomplete retention by the SPE cartridges If bothissues are improved the uncertainty of recovery will beslightly decreased Secondly the contribution of intermediateprecision shows values ranged between 10 and 20This con-tribution involves variations along the time for the analysis ofthe spiked water samples at different levels so robustness ofmethodology is clearly verified And finally the contributiondue to the analysis shows the highest values It would be


Table 4 Uncertainty associated with concentration measured through calibration

(119906119862std)21198622std (119906119862119909)21198622119909 11990621199091199092 119906119909119909Aldrin 00004 0032 0032 018Atrazine 00004 0014 0014 012Dieldrin 00004 0212 0212 046op1015840-DDE 00004 0007 0007 008pp1015840-DDE 00004 0011 0012 011op1015840-DDD 00004 0019 0019 014pp1015840-DDD 00004 0026 0026 016op1015840-DDT 00004 0025 0026 016pp1015840-DDT 00004 0021 0021 014Endrin 00004 0251 0252 050Hexachlorobenzene 00004 0027 0027 016Isodrin 00004 0061 0061 025120572-HCH 00004 0018 0018 013120573-HCH 00004 0015 0016 013120574-HCH 00004 0022 0023 015Metolachlor 00004 0016 0016 013Pentachlorobenzene 00004 0120 0120 035PCB-28 00004 0027 0027 016PCB-52 00004 0050 0050 022PCB-101 00004 0093 0093 030PCB-118 00004 0090 0091 030PCB-138 00004 0158 0159 040PCB-153 00004 0053 0053 023PCB-180 00004 0033 0034 018Simazine 00004 0040 0041 020Terbuthylazine 00004 0017 0018 013

AldrinAtrazineDieldrin

op-DDDpp-DDDop-DDEpp-DDEop-DDTpp-DDT

EndrinHexachlorobenzene

Isodrin

MetolachlorSimazine

Terbuthylazine

Combined uncertainty

10 20 30 40 50 60 700()

-HCH-HCH-HCH

Figure 3 Combined uncertainty for the analysis of the selected pesticides in water samples


Table 5 Results of uncertainty associated with recovery of selectedpesticides in water and sediment samples (na nonanalyzed)

Water Sediment119906rec 119906(119877mean)Aldrin 0019 naAtrazine 0037 naDieldrin 0058 naop-DDE 0062 napp-DDE 0086 0089op-DDD 0070 napp-DDD 0072 0103op-DDT 0073 napp-DDT 0114 0047Endrin 0130 naHexachlorobenzene 0066 0205Isodrin 0033 na120572-HCH 0062 0090120573-HCH 0051 0081120574-HCH 0054 0065Metolachlor 0035 naPCB-28 na 0137PCB-52 na 0100PCB-101 na 0113PCB-118 na 0102PCB-138 na 0119PCB-153 na 0083PCB-180 na 0087Simazine 0044 naTerbuthylazine 0060 na

therefore very difficult to reduce the associated uncertaintyOnly the use of new chromatographic columns inlet systemsor more sensitive detectors will enable us to make practicalimprovements In the case of the value of endrin uncertaintyis associated with well-known degradation of this compoundto endrin ketone and endrin aldehyde into the inlet systemsNevertheless it is highlighted that the uncertainty measuredinvolves each calibration performed along nine months Andthus nonoptimal conditions are also considered in order topoint the need of periodic evaluation of this degradationThe authors have finally concluded that inlet system must besubstituted periodically and thus that issue does not affect thefinal results

The total uncertainty associated with the determinationof selected organochloride compounds from sediment sam-ples is evaluated differently In this case the three maincontributions cannot be combined because the three contri-butions are not independent

Uncertainty contributions estimated in the present workfromproportionalconstant bias of sediment subsamplemea-surements are deduced from the values above included inTables 2 and 3 Hence component of proportional bias isdeduced as a combined value of the variances from respectiveassays as it was previously indicated Taking into account

Table 6 Intermediate precision calculated for determination ofselected pesticides in water

Compound Average 119899 rsdpooled 119906(rsd119901) 119906119903(rsd119901) ()Aldrin 049 3 0082 0082 17Atrazine 111 6 0104 0104 9op1015840-DDD 066 9 0172 0172 15pp1015840-DDD 063 6 0150 0150 26op1015840-DDE 037 6 0085 0085 24pp1015840-DDE 050 9 0096 0096 23op1015840-DDT 060 6 0063 0063 19pp1015840-DDT 076 9 0160 0160 10Dieldrin 081 9 0217 0217 21Endrin 082 9 0120 0120 27HCB 078 3 0090 0090 12120572-HCH 128 9 0199 0199 16120573-HCH 117 9 0182 0182 16120574-HCH 120 9 0176 0176 15Isodrin 054 3 0121 0121 22Metolachlor 098 6 0076 0076 8Simazine 117 6 0121 0121 10Terbuthylazine 091 6 0095 0095 11

Table 7 Intermediate precision assessed by the analysis of differentsediments collected in the area of study The number of sedimentsamples corresponds to those of samples with mass fraction abovedetection limit

Compound Number of sediment samples RSDpooled ()op1015840-DDD 32 23pp1015840-DDD 32 36op1015840-DDE 32 23pp1015840-DDE 31 24op1015840-DDT 14 54pp1015840-DDT 10 63Hexachlorobenzene 31 39120572-HCH 32 31120573-HCH 29 36120574-HCH 27 25Pentachlorobenzene 27 43PCB-28 31 27PCB-52 32 24PCB-101 31 30PCB-118 18 24PCB-138 11 37PCB-153 29 32PCB-180 9 33

that uncertainty components associated with constant biasare estimated from assays carried out with different samplesthe corresponding uncertainty contribution is estimated asan intermediate precision by means of a pooled relativestandard fromdeduced uncertainty contributions of constantbias (119906(119887cte))


HCB

PCB-28PCB-52

PCB-101PCB-118PCB-138PCB-153PCB-180

Intermediate precisionCalibration

10 20 30 40 50 60 700Standard uncertainty ()

-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD

Figure 4 Standard uncertainty for intermediate precision and calibration contributions for the analysis of selected pesticides and PCBs insediments

HCB

PCB-28PCB-52


Constant biasProportional bias

-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD

5 10 15 20 25 30 35 400Standard uncertainty ()

Figure 5 Standard uncertainty for constant and proportional bias contributions for the analysis of selected pesticides and PCBs in sediments

Figures 4 and 5 are the representation for comparisonamong the uncertainties contributions deduced in this workIn general the analytical variability from intermediate pre-cision is the highest contribution to uncertainty The valuesof intermediate precision correspond to the set of analyticalmeasurements made during 9 months with different com-plexity of different sediment samples It can be consideredthat those values are indicative of the total uncertaintyIn general the variability ranged between 25 and 35 formost of the studied compounds except for pp1015840-DDT andhexachlorobenzene due to their degradation and volatilityrespectively

From the results the variability data associated with thepresence of bias indicates that the values of constant bias arehigher than those of proportional bias This effect is mainlyobserved for DDT HCB and isomers of HCH type

4 Conclusions

Thiswork describes two analytical procedures for quantifyingselected organochlorine compounds in water at concentra-tions of ng Lminus1 and in sediments at mass fraction of mgKgminus1The analytical protocols use a minimal handling and reagentsfor analyzing the goal compounds in sediments and watersBoth analytical methods show adequate validation param-eters for quantifying those compounds at trace levels Themethod for analyzing water samples involves a solid phaseextraction and a GCMS analysis meanwhile the methodfor analyzing sediments comprises a sequentialmechanic andultrasonic extractions followed by GCMS analysis

The novelty of this work focuses on themethod validationprocedure using routine samples That assessment is rarelyapplied Hence the present work is of highest significance in


order to assure quality results for monitoring organochlorinecompounds in environmental matrices

The proposed analytical method for sediments providesan important reduction of costs associated with analysesminimizing extraction time volume of solvents and sampleweight These aspects make an easier cheaper and fastermethod than classical extraction procedures

In order to evaluate the performance of the proposedmethod the bias and precision have been evaluated by theanalysis of routine samples One of the main objectives wasto optimize the sample size while maintaining the absence ofa significant bias The proportional bias from recovery testsis only significant in the case of pp1015840-DDT and HCH isomersSimilarly the constant bias is not significant when aliquots of2 g were chosen Lower weights lead to significant constantbias which could imply revision or correction of the finalresults Moreover the evaluation of uncertainty is carried outby grouping the intermediate precision and the bias studyIn general the results are very consistent showing a rangebetween 25 and 35

The analytical methodology to determine the selectedcompounds in water allows quantifying them at trace levelswith acceptable recoveries Due to homogeneity of watermatrices the assessing of trueness is performed by a recoverystudyMoreover threemain sources of uncertainty have beenidentified recovery calibration of GCMS device and inter-mediate precision Finally the combined uncertainty rangesbetween 20 and 30 for all compounds with the exception ofendrin That value is due to well-known degradation of thiscompound into the inlet systems of a gas chromatographyWe have concluded that inlet system must be substitutedperiodically in order to avoid errors

Finally the authors give case-study on how to validate andestimate uncertainty for the determination of selected com-pounds in function of type of matrix This work also aimedat highlighting that validation and the evaluation of theuncertainty associated must be adapted to type of matricesbeing an essential part of an environmental analysis

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

References

[1] Y-F Guan J-Z Wang H-G Ni and E Y Zeng ldquoOrganochlo-rine pesticides and polychlorinated biphenyls in riverine runoffof the Pearl River Delta China Assessment of mass loadinginput source and environmental faterdquo Environmental Pollutionvol 157 no 2 pp 618ndash624 2009

[2] S H Hong U H Yim W J Shim J R Oh P H Viet andP S Park ldquoPersistent organochlorine residues in estuarine andmarine sediments fromHaLongBayHai PhongBay andBa LatEstuary Vietnamrdquo Chemosphere vol 72 no 8 pp 1193ndash12022008

[3] N Ratola L Santos P Herbert and A Alves ldquoUncertaintyassociated to the analysis of organochlorine pesticides in waterby solid-phase microextractiongas chromatography-electron

capture detection-Evaluation using two different approachesrdquoAnalytica Chimica Acta vol 573-574 pp 202ndash208 2006

[4] S Yenisoy-Karakas ldquoValidation and uncertainty assessment ofrapid extraction and clean-up methods for the determinationof 16 organochlorine pesticide residues in vegetablesrdquoAnalyticaChimica Acta vol 571 no 2 pp 298ndash307 2006

[5] S Lacorte J Quintana R Tauler F Ventura A Tovar-Sanchezand C M Duarte ldquoUltra-trace determination of PersistentOrganic Pollutants in Arctic ice using stir bar sorptive extrac-tion and gas chromatography coupled to mass spectrometryrdquoJournal of Chromatography A vol 1216 no 49 pp 8581ndash85892009

[6] P Castells J Parera F J Santos and M T Galceran ldquoOccur-rence of polychlorinated naphthalenes polychlorinated bi-phenyls and short-chain chlorinated paraffins in marine sedi-ments from Barcelona (Spain)rdquo Chemosphere vol 70 no 9 pp1552ndash1562 2008

[7] J O Bustnes Oslash Miland M Fjeld K E Erikstad and J USkaare ldquoRelationships between ecological variables and fourorganochlorine pollutants in an artic glaucous gull (Larushyperboreus) populationrdquoEnvironmental Pollution vol 136 no1 pp 175ndash185 2005

[8] H W Vallack D J Bakker I Brandt et al ldquoControlling persis-tent organic pollutants-what nextrdquo Environmental Toxicologyand Pharmacology vol 6 no 3 pp 143ndash175 1998

[9] M Shamsipur N Yazdanfar and M Ghambarian ldquoCombi-nation of solid-phase extraction with dispersive liquid-liquidmicroextraction followed by GC-MS for determination ofpesticide residues fromwater milk honey and fruit juicerdquo FoodChemistry vol 204 pp 289ndash297 2016

[10] Z Yang Z Shen F Gao Z Tang and J Niu ldquoOccurrenceand possible sources of polychlorinated biphenyls in surfacesediments from the Wuhan reach of the Yangtze River ChinardquoChemosphere vol 74 no 11 pp 1522ndash1530 2009

[11] A Navarro R Tauler S Lacorte and D Barcelo ldquoOccurrenceand transport of pesticides and alkylphenols in water samplesalong the Ebro River Basinrdquo Journal of Hydrology vol 383 no1-2 pp 18ndash29 2010

[12] I K Konstantinou D G Hela and T A Albanis ldquoThe statusof pesticide pollution in surface waters (rivers and lakes) ofGreece Part I Review on occurrence and levelsrdquo EnvironmentalPollution vol 141 no 3 pp 555ndash570 2006

[13] P N Carvalho P N R RodriguesM C P Basto andM T S DVasconcelos ldquoOrganochlorine pesticides levels in Portuguesecoastal areasrdquo Chemosphere vol 75 no 5 pp 595ndash600 2009

[14] N Yazdanfar Y Yamini and M Ghambarian ldquoHomogeneousliquid-liquidmicroextraction for determination of organochlo-rine pesticides in water and fruit samplesrdquo Chromatographiavol 77 no 3-4 pp 329ndash336 2014

[15] B Cavaliere M Monteleone A Naccarato G Sindona and ATagarelli ldquoA solid-phase microextraction-gas chromatographicapproach combined with triple quadrupole mass spectrometryfor the assay of carbamate pesticides in water samplesrdquo Journalof Chromatography A vol 1257 pp 149ndash157 2012

[16] C Jia X Zhu J Wang et al ldquoExtraction of pesticides in watersamples using vortex-assisted liquid-liquid microextractionrdquoJournal of Chromatography A vol 1217 no 37 pp 5868ndash58712010

[17] F J Camino-Sanchez A Zafra-Gomez S Cantarero-Malagonand J L Vılchez ldquoValidation of a method for the analysis of77 priority persistent organic pollutants in river water by stir


bar sorptive extraction in compliance with the EuropeanWaterFramework Directiverdquo Talanta vol 89 pp 322ndash334 2012

[18] P Lepom B Brown G Hanke R Loos P Quevauviller and JWollgast ldquoNeeds for reliable analytical methods for monitoringchemical pollutants in surface water under the EuropeanWaterFramework Directiverdquo Journal of Chromatography A vol 1216no 3 pp 302ndash315 2009

[19] A Gustavo Gonzalez and M Angeles Herrador ldquoA practicalguide to analytical method validation including measurementuncertainty and accuracy profilesrdquo TrAC - Trends in AnalyticalChemistry vol 26 no 3 pp 227ndash238 2007

[20] O Pindado Jimenez and R M Perez Pastor ldquoEstimationof measurement uncertainty of pesticides polychlorinatedbiphenyls and polyaromatic hydrocarbons in sediments byusing gas chromatography-mass spectrometryrdquo AnalyticaChimica Acta vol 724 pp 20ndash29 2012

[21] A Drolc and A Pintar ldquoMeasurement uncertainty evaluationand in-house method validation of the herbicide iodosulfuron-methyl-sodium in water samples by using HPLC analysisrdquoAccreditation and Quality Assurance vol 16 no 1 pp 21ndash292011

[22] R Stepan J Hajslova V Kocourek and J Ticha ldquoUncertaintiesof gas chromatographic measurement of troublesome pesticideresidues in apples employing conventional and mass spectro-metric detectorsrdquo Analytica Chimica Acta vol 520 no 1-2 pp245ndash255 2004

[23] L Cuadros-Rodrıguez M E Hernandez Torres E AlmansaLopez F J Egea Gonzalez F J Arrebola Liebanas and JL Martınez Vidal ldquoAssessment of uncertainty in pesticidemultiresidue analytical methodsMain sources and estimationrdquoAnalytica Chimica Acta vol 454 no 2 pp 297ndash314 2002

[24] A S D Souza J P M Torres R O Meire R C Neves MS Couri and C S Serejo ldquoOrganochlorine pesticides (OCs)and polychlorinated biphenyls (PCBs) in sediments and crabs(Chasmagnathus granulata Dana 1851) from mangroves ofGuanabara Bay Rio de Janeiro State Brazilrdquo Chemosphere vol73 no 1 pp S186ndashS192 2008

[25] J Klanova N Matykiewiczova Z Macka P Prosek K Laskaand P Klan ldquoPersistent organic pollutants in soils and sed-iments from James Ross Island Antarcticardquo EnvironmentalPollution vol 152 no 2 pp 416ndash423 2008

[26] EURACHEM CITAC Guide CG 4 Quantifying Uncertainty inAnalytical Measurement 3rd edition 2012

[27] S Garcıa-Alonso R M Perez-Pastor D Sanz-Rivera E Rojas-Garcıa and J Rodrıguez-Maroto ldquoPAH analysis in biomasscombustion wastes an approach to evaluate bias and precisionof analytical results using routine samplesrdquo Accreditation andQuality Assurance vol 22 no 3 pp 153ndash159 2017

[28] A Maroto R Boque J Riu and F Xavier Rius ldquoEstimation ofmeasurement uncertainty by using regression techniques andspiked samplesrdquo Analytica Chimica Acta vol 1-2 pp 131ndash1432001

[29] A Maroto R Boque J Riu and F X Rius ldquoMeasurementuncertainty in analytical methods in which trueness is assessedfrom recovery assaysrdquo Analytica Chimica Acta vol 440 no 2pp 171ndash184 2001

[30] W R Kelly and J D Fasset ldquoDetermination of picogramquanti-ties of uranium in biological tissues by isotope dilution thermalionization mass spectrometry with ion counting detectionrdquo inAnalytical Chemistry pp 1040ndash1044 1983

[31] S Garcıa-Alonso R M Perez-Pastor and F J Garcıa-FrutosldquoAn evaluation of analytical quality for selected PAH measure-ments in a fuel-contaminated soilrdquo Accreditation and QualityAssurance vol 16 no 7 pp 369ndash377 2011

[32] O P Jimenez R M P Pastor and S G Alonso ldquoAssessmentuncertainty associated to the analysis of organic compoundslinked to particulate matter of atmospheric aerosolsrdquo Talantavol 80 no 3 pp 1121ndash1128 2010

Submit your manuscripts athttpswwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Inorganic ChemistryInternational Journal of

Hindawi Publishing Corporation httpwwwhindawicom Volume 201

International Journal ofInternational Journal ofPhotoenergy


Carbohydrate Chemistry

International Journal ofInternational Journal of


Journal of

Chemistry


Advances in

Physical Chemistry

Hindawi Publishing Corporationhttpwwwhindawicom

Analytical Methods in Chemistry

Journal of

Volume 2014

Bioinorganic Chemistry and ApplicationsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

SpectroscopyInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Medicinal ChemistryInternational Journal of


Chromatography Research International


Applied ChemistryJournal of



Theoretical ChemistryJournal of


Journal of

Spectroscopy

Analytical ChemistryInternational Journal of


Journal of


Quantum Chemistry


Organic Chemistry International

ElectrochemistryInternational Journal of



CatalystsJournal of


Within the framework of CIEMAT research an analyticalmethodology might be performed in order to quantify pesti-cides and PCBs in environmental samples including a widerange of concentrations The aims of this study were on theone hand developing two analytical methods to determinethe selected pesticides and PCBs as fast robust and simpleas possible in waters and sediments and on the other handdemonstrate the reliability of those methods by means of theanalysis of routine samples It is necessary to highlight thatthe present work is also intended to provide an easy to followcase-study on how to validate analytical protocols in differentenvironmental matrices using routine samples

Development and application of analytical methods fordetermination of organochlorine compounds in severalmatrices have already been published [9 17 24 25] Takinginto account the inherent characteristics of each matrix theestimation of possible constantproportional bias should bealso carried out to check the quality of analytical resultsTherestudies aim to formulate mathematical algorithms to arrive atfull validation but sometimes they are difficult to apply [26]

There is still a lack of information on the variabilityassociated with concentration level matrix of samples oranalysis dates To our knowledge this approach is rarelyapplied for environmental studies and has not yet beendone to the determination of organochlorine compounds insediments and watersThe present work can be considered tobe of the highest relevance in order to evaluate quality of anyanalytical method

2 Materials and Methods

21 Reagents andMaterials Pesticides standardweremixtureof 8 pesticides (pesticide mix-5 of 10 ng 120583Lminus1 in cyclohexane)mixture of 4 pesticides (pesticideMix 7 of 10 ng 120583Lminus1 in cyclo-hexane) mixture of 6 pesticides (pesticide Mix 31 of10 ng 120583Lminus1 in acetone) mixture of 26 pesticides and PCBs(pesticide Mix 33 of 10 ng 120583Lminus1 in isooctane) and penta-chlorobenzene and pentachlorophenol (10 ng 120583Lminus1 in cyclo-hexane) fromDr Ehrenstorfer GmbH (Augsburg Germany)Anthracene deuterated (10 g 120583Lminus1 in cyclohexane) from DrEhrenstorfer GmbH (Augsburg Germany) and o-terphenylfrom Supelco (Bellefonte USA) were used as surrogateand internal standard respectively Two certified referencematerials were used CNS391 (CRM) of pesticides andPCBs in sediment of natural water and WatRTM pollutionorganochlorine pesticides 709 (ERA) Acetonitrile dichloro-methane ethyl acetate and hexane of residue analysis gradewere supplied by J T Baker (Deventer Holland) Sodiumsulfate anhydrous was obtained from Merck (DarmstadtGermany)

The SPE procedure was performed using Oasis HLBExtraction Cartridge (6 cc and 500mg) acquired to Waters(Massachusetts USA) in a vacuum manifold Agilent (SantaClara USA)

22 Instruments AnAgilent 6890 chromatograph with auto-matic sampler and a splitsplitless programmed temperatureinjector coupled to an Agilent 5975 mass spectrometer was

employed Operating conditions were carrier gas heliumcapillary column HP-5MS (30m times 025mm times 025 120583m) withflow rate of 1mLminminus1 PTV injector with initial temperature110∘C held for 05min and then increased at a rate of500∘Cminminus1 to 200∘C operation mode splitless with 3 rowinjections of 1 120583L initial oven temperature 100∘C held for4min increased at a rate of 8∘Cminminus1 to 250∘C held for1min and finally increased at a rate of 8∘Cminminus1 to 290∘Cand detector operated in electronic impact mode (70 eV) inSIM mode

23 Sampling To evaluate the level of selected pesticides andPCBs during the time period among March 2015 to October2015 weekly samples of sediment and water were collectedfrom a confined area of continental waters in which therewere industrial discharges

Water samples were collected in 1-liter glass bottles andstored at 4∘C until analysis 100 g of sediment samples wascollected in amber glass bottles and stored at 4∘C untilanalysis without any additional treatment

24 Description of Analytical Method

241 Water Samples Figure 1 shows a scheme of the pro-posed methodology to determine selected compounds in awater matrix In this method 1 L of continental water spikedwith 5 120583L of anthracene deuterated (10 ng120583Lminus1) as surrogateis employed to determine the selected organochlorine pesti-cides (pp1015840-DDD op1015840-DDD pp1015840-DDE op1015840-DDE pp1015840-DDTop1015840-DDT 120572-HCH 120573-HCH 120574-HCH atrazine simazinemetolachlor terbuthylazine aldrin dieldrin endrin isodrinand hexachlorobenzene) Firstly cartridges were conditionedwith 12mL of dichloromethane and 12mL of acetonitrileand later were rinsed with 12mL of water Secondly whencartridges are still wet water sample is passed through themat a flow rate of 5ndash10mLminminus1 Cartridges were washed with6mL of water and then dried for 20min under a stream ofnitrogen Later extracts were eluted with 25mL of acetoni-triledichloromethane (1 1) and 32mL of dichloromethaneThen extracts were concentrated with nitrogen to 3mLand 1 g of anhydrous sodium sulfate was added to take outhumidity Finally extracts were concentratedwith nitrogen to90 120583L of ethyl acetate and 10 120583L of internal standard solution(10 ng120583Lminus1 of o-terphenyl in hexane) was added and injectedinto the GC-MS

242 Sediment Samples Figure 2 shows a scheme of theproposed methodology to analyze sediments matrix Themethod was optimized for determining the selected organicpesticides and PCBs (pp1015840-DDD op1015840-DDD pp1015840-DDEop1015840-DDE pp1015840-DDT op1015840-DDT 120572-HCH 120573-HCH 120574-HCHhexachlorobenzene pentachlorobenzene PCB-28 PCB-52PCB-101 PCB-118 PCB-138 PCB-153 and PCB-180) Themethod begins as follows 05ndash2 g of sediment spiked with10 120583L of anthracene deuterated (10 ng 120583Lminus1) as surrogatewas weighed and 25mL of hexaneacetone mixture wassubsequently added The first extraction was performed byshaking the mixture for 1min in a Vortex and later goes




River water1 litre



SPEOasis HLB

(6 cc 500 mg)


Elute cartridge

(2) 32 mL （2l2(1) 25 mL acto（2l2




















(1 1)

times2



Dissolve1 mL hexane


(10 min)

Centrifugate

3 min2500 rpm





119877119894 = 119908found(119894)119908added(119894) (1)




119905119877 = |119877 minus 1|119906 (119877) (2)







(4)



119906119903 (119887cte) = 1(119886119894 minus 119886119895)radic(119886119895119906 (119909119894)119909119894 )2 + (119886119894 119906 (119909119895)119909119895 )2 (5)


119905119894 = 119887cte119909119894119906119903 (119887cte) (6)
































11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of




River water1 litre



SPEOasis HLB

(6 cc 500 mg)


Elute cartridge

(2) 32 mL （2l2(1) 25 mL acto（2l2




















(1 1)

times2



Dissolve1 mL hexane


(10 min)

Centrifugate

3 min2500 rpm





119877119894 = 119908found(119894)119908added(119894) (1)




119905119877 = |119877 minus 1|119906 (119877) (2)







(4)



119906119903 (119887cte) = 1(119886119894 minus 119886119895)radic(119886119895119906 (119909119894)119909119894 )2 + (119886119894 119906 (119909119895)119909119895 )2 (5)


119905119894 = 119887cte119909119894119906119903 (119887cte) (6)
































11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of









(1 1)

times2



Dissolve1 mL hexane


(10 min)

Centrifugate

3 min2500 rpm





119877119894 = 119908found(119894)119908added(119894) (1)




119905119877 = |119877 minus 1|119906 (119877) (2)







(4)



119906119903 (119887cte) = 1(119886119894 minus 119886119895)radic(119886119895119906 (119909119894)119909119894 )2 + (119886119894 119906 (119909119895)119909119895 )2 (5)


119905119894 = 119887cte119909119894119906119903 (119887cte) (6)
































11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of





(4)



119906119903 (119887cte) = 1(119886119894 minus 119886119895)radic(119886119895119906 (119909119894)119909119894 )2 + (119886119894 119906 (119909119895)119909119895 )2 (5)


119905119894 = 119887cte119909119894119906119903 (119887cte) (6)
































11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of






















11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of









11990621198621199091198622119909 = 1199062119910(119910 minus 119887)2 + 1199062119887(119910 minus 119887)2 + 11990621198981198982 (7)


























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of

























Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of







Isodrin

MetolachlorSimazine

Terbuthylazine


10 20 30 40 50 60 700()

-HCH-HCH-HCH














HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of













HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of


HCB

PCB-28PCB-52




-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD


HCB

PCB-28PCB-52



-HCH-HCH-HCH

pp-DDTpp-DDEpp-DDD





4 Conclusions











References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of









References













































Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of



























Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of










Journal of

Chemistry


Advances in

Physical Chemistry



Journal of

Volume 2014














Journal of

Spectroscopy



Journal of


Quantum Chemistry






CatalystsJournal of

Documents

Validating Analytical Protocols to Determine Selected ...downloads.hindawi.com/journals/jamc/2017/9796457.pdf · 4 JournalofAnalyticalMethodsinChemistry Concentrate under nitrogen