Op

EDD

a

ARRAA

KLMMPSU

1

fteihc[wAadd

otah

h1

Journal of Chromatography B, 958 (2014) 48–54

Contents lists available at ScienceDirect

Journal of Chromatography B

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

ptimization of diclofenac quantification from wastewater treatmentlant sludge by ultrasonication assisted extraction

mel Topuz, Sevgi Sari, Gamze Ozdemir, Egemen Aydin, Elif Pehlivanoglu-Mantas,idem Okutman Tas ∗

epartment of Environmental Engineering, Istanbul Technical University, Maslak, Istanbul 34469, Turkey

r t i c l e i n f o

rticle history:eceived 31 October 2013eceived in revised form 26 February 2014ccepted 28 February 2014vailable online 1 April 2014

eywords:

a b s t r a c t

A rapid quantification method of diclofenac from sludge samples through ultrasonication assisted extrac-tion and solid phase extraction (SPE) was developed and used for the quantification of diclofenacconcentrations in sludge samples with liquid chromatography/tandem mass spectrometry (LC–MS/MS).Although the concentration of diclofenac in sludge samples taken from different units of wastewatertreatment plants in Istanbul was below the limit of quantification (LOQ; 5 ng/g), an optimized methodfor sludge samples along with the total mass balances in a wastewater treatment plant can be used to

C–MS/MSethodicropollutants

harmaceuticalolid phase extractionltrasonication

determine the phase with which diclofenac is mostly associated. Hence, the results will provide infor-mation on fate and transport of diclofenac, as well as on the necessity of alternative removal processes.In addition, since the optimization procedure is provided in detail, it is possible for other researchers touse this procedure as a starting point for the determination of other emerging pollutants in wastewatersludge samples.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

The presence of pharmaceuticals and their metabolites in sur-ace waters, ground water, soil, or sediment is mainly due tohe discharge of municipal wastewater treatment plant (WWTP)ffluents. The occurrence and detected concentration of emerg-ng contaminants in surface waters and ground water samplesas prompted public interest regarding potential adverse ecologi-al effects and potential contamination of drinking water sources1,2]. Non-steroidal anti-inflammatory drugs are one of the mostidely detected pharmaceutical groups in the environment [3–6].mong them, diclofenac is a commonly used over-the-counternti-inflammatory drug worldwide [7]. Even though only 15% oficlofenac administered is excreted as the unchanged form [8],iclofenac is frequently detected in wastewaters.

Studies point out that certain and suspected effects of diclofenacn different organisms occur even at low concentrations [9–14] andhat it has the highest acute toxicity on aquatic organisms among

nti-inflammatory drugs [15]. Moreover, diclofenac is listed in theigh priority category by Global Water Research Coalition [16].

∗ Corresponding author. Tel.: +90 212 285 37 87; fax: +90 212 285 65 45.E-mail address: okutmand@itu.edu.tr (D. Okutman Tas).

ttp://dx.doi.org/10.1016/j.jchromb.2014.02.047570-0232/© 2014 Elsevier B.V. All rights reserved.

Diclofenac is not readily biodegradable [17] and hence highconcentrations of diclofenac are discharged from WWTPs to theenvironment [18]. Diclofenac have been reported to be present inground water and surface waters at significant levels [19–24], eventhough it is susceptible to photolysis in surface waters [25]. Theresults of studies showed that diclofenac concentration in mostof the investigated WWTPs effluents exceeded the proposed Envi-ronmental Quality Standards (0.1 �g/L) [26] thus indicating thatthe fate of diclofenac in municipal WWTPs needs to be furtherevaluated in order to improve its removal efficiency in existingWWTPs. Moreover, rapid measurement of diclofenac in WWTPscan also assist in providing more effective conceptual treatmentalternatives and in the determination of environmental compart-ment(s) where diclofenac can pose a risk to the ecosystem.

The fate of a pollutant in a WWTP cannot be determined byits quantification in the dissolved phase alone. Thus, the inves-tigation of the fate of pollutants in WWTPs requires knowledgeon the presence and speciation of these compounds in both liquidand solid phases to identify whether the removal from the liquidphase is due to biological processes or sorption to the sludge phase[27]. Mass balances based on only the dissolved phase may lead to

significant errors if sorption to the sludge plays a significant roleduring removal of pollutants. In addition, the possibility of agricul-tural application of WWTP sludge requires careful assessment ofmicropollutants in the sludge phase [28–30].

matog

amodtc

p(tiseatartlssfis

2

2

mg1nGt(swww

2

lt9ssU

ftdTtt

2

SMs

E. Topuz et al. / J. Chro

In the literature, there is a lack of knowledge on the occurrencend fate of pharmaceuticals in solid matrixes (e.g. sludge, sedi-ents) [31–35] with most of the studies focusing on quantification

f diclofenac in the liquid phase. The reason for the lack of studiesealing with both liquid and solid phases [18,36] can be attributedo current limitations of analytical measurement techniques foromplex matrices [37,38].

The most frequently used extraction methods for sludge sam-les are Microwave Extraction (MWE), Accelerated Solid ExtractionASE), Pressurized Liquid Extraction (PLE) and Ultrasonic Extrac-ion (USE) [39]. USE is the most practical technique because theres no need for special equipment, but can be both time andolvent consuming [38,40]. However, solvent consumption andxtraction duration can be decreased by optimizing factors thatffect extraction recovery (e.g. extraction time, extraction solventype, extraction solvent volume, number of extraction cycles, andmount of sludge sample). In the literature, some studies onlyeport the extraction procedure and do not present details ofhe optimization procedure [34–42], while others focus on only aimited number of factors affecting the extraction procedure (e.g.olvent type, sample amount) [40,43–45]. Therefore, the aim of thistudy was to establish an easy-to-apply and efficient USE methodor the extraction of diclofenac from WWTP sludge. The applicabil-ty of the optimized extraction method was assessed with WWTPludge samples taken from different units of the WWTPs.

. Materials and methods

.1. Reagents

All of the solvents used for USE, SPE and LC–MS/MS measure-ents were HPLC grade (Sigma–Aldrich, USA) except for LC–MS

rade water (Merck, Germany). Diclofenac sodium salt (98%; CAS5307-79-6) and isotope labeled diclofenac-d4 used as inter-al standards were purchased from Sigma–Aldrich (Steinheim,ermany) and C/D/N Isotopes (Canada), respectively. Stock solu-

ions of diclofenac and diclofenac-d4 were prepared in methanolMeOH) and working solutions were prepared daily by dilutingtock solutions in LC–MS grade water. All non-volumetric glass-are used for preparation of solutions and recovery experimentsas kept at 550 ◦C for 3 h after rinsing with MeOH and deionizedater (DI).

.2. Sample preparation

Sludge samples were collected from two municipal WWTPsocated in Istanbul, Turkey. All sludge samples, except the onesaken from drying and centrifugation units, were centrifuged at000 rpm for 10 min to concentrate the sludge phase and then allamples were freeze dried (ThermoSavant, Modulyo D, USA) andtored at −20 ◦C. Samples were weighed before the application ofSE.

Recovery studies were performed with the sludge sample takenrom the drying unit of the WWTP-II. 2 mL MeOH was addedo freeze dried sludge samples before spiking diclofenac andiclofenac-d4 to have a homogenous concentration distribution.hen samples were mixed for 5 min with vortex. USE was appliedo the samples which were kept overnight in a darkened fume hoodo evaporate MeOH.

.3. Ultrasonic Extraction (USE)

The performances of the ultrasonic homogenizator (Bandelin,onoplus HD 2200, Germany) and ultrasonic bath (Intersonik,IN12, Turkey) were tested for the extraction of diclofenac from

ludge samples. Experiments were performed in a basic ultrasonic

r. B 958 (2014) 48–54 49

bath with a 25–40 kHz ultrasonic frequency. The temperature dur-ing extraction was kept constant at room temperature. Since thereare numerous combinations to test the factors affecting extrac-tion efficiency, a systemic approach was used for optimization.A method was selected from literature [46] and several factors(e.g. optimum sludge amount, type and volume of extraction sol-vent, ultrasonication time, extraction cycle), were optimized ina stepwise manner. All sludge samples were ultrasonicated in50 mL Teflon centrifuge tubes (Nalgene, USA). At the end of thesolvent extraction, collected supernatants were evaporated to dry-ness at 45 ◦C by using a rotary evaporator (Heidolp, Laborota 4000,Germany). The test conditions for extraction are provided for eachoptimization step in related sections.

2.4. Solid phase extraction (SPE)

SPE was applied at the end of the USE process in order toclean-up and concentrate the samples. Several types of SPE car-tridges (Oasis HLB cartridge 6 cc/200 mg; Oasis MCX Cartridges3 cc/60 mg (Waters, Millford, MA, USA); Cleanert PEP-H car-tridge 6 cc/200 mg; Cleanert PAX cartridge 6 cc/200 mg; CleanertPCX 6 cc/200 mg (Bonna-Agela, Willmington, DE, USA)) weretested (data not shown). Oasis HLB cartridge (6 cc/200 mg) whichprovided the best recovery was selected for SPE analysis. An auto-matic SPE vacuum manifold (VacMaster, Biotage, NC, USA) wasused in order to have homogeneous extraction efficiency. MeOH,MeOH:acetonitrile (ACN), MeOH:acetone (Ace), methyl tert-butylether (MTBE) and ethyl acetate (EtOAc) were tested as eluent byusing DI spiked with diclofenac and diclofenac-d4. Glassware andPTFE lined materials were used in order to prevent adsorption ofdiclofenac.

2.5. Liquid chromatography/tandem mass spectrometry(LC–MS/MS)

Diclofenac was measured by using Thermo Electron Cooper-ation TSQ Quantum Access triple quadruple mass spectrometercoupled with Accela Ultra Performance Liquid Chromatograph(UPLC) with a Thermo Hypersil Gold column (1.2 mm × 100 mm;1.9 �m; Thermo Scientific, Germany) and an electrospray ioniza-tion interface. LC–MS/MS method used for diclofenac measurementwas modified from the study of Aydinli and Talinli [20]. Chro-matographic separation (Table S1) and MS parameters (Table S2)were optimized for diclofenac measurement. While quantificationwas made by using transitions of 294–214 m/z for diclofenac and298–253.9 m/z for diclofenac-d4, confirmations of these ions wereprovided with transitions of 294–249.7 m/z and 298–281.6 m/z fordiclofenac and diclofenac-d4, respectively. A chromatogram exam-ple for the optimized conditions can be found in Fig. S1.

2.6. Quality assurance and control

All samples were injected three times and relative standarddeviations (RSD) were below 20%, which indicates a high precisionfor this method. Diclofenac-d4 was used as an internal/surrogatestandard for quantification of diclofenac. Correlation coefficient(R2) of calibration curves used for the quantification were alwaysgreater than 0.995. Limit of Detection (LOD) and Limit of Quantifi-cation (LOQ) were determined using signal to noise ratios of 3 and10, respectively.

The accuracy of this method was tested by performing recov-ery experiments. Sludge samples were spiked with 100 or 400 ng/g

diclofenac (corresponding to low and high concentration levels,respectively) and 50 ng/g diclofenac-d4. The extraction methodwas optimized by determining the conditions that had the highestrecovery (≥80%). For method validation, diclofenac was quantified

50 E. Topuz et al. / J. Chromatogr. B 958 (2014) 48–54

Table 1Extraction recoveries for ultrasonic homogenizator and ultrasonic bath (mean valueof the three replicates).

Equipment Diclofenacconcentration (ng/gdry matter)

Recovery(%)

Std dev(%)

Ultrasonichomogenizator

100 85 10400 77 10

Ultrasonic bath 100 94 2400 84 7

E(c

itsr

3

wuedmsI

3

3

[ctfwAttvl

cmspcvhbeifd

3

lmsao

0

10

20

30

40

50

60

70

80

90

100

MeOH AC N Ac e MeOH :AC N MeOH :Ac e

Recovery

(%

)

100 ng /g 400 ng /g

Fig. 1. Extraction recoveries for different type of solvents tested (mean value ofthe three replicates). (Extraction conditions: sample amount, 0.5 g; 10 mL of solvent

xtraction conditions: sample amount, 0.5 g; extraction solvent, 10 mL of MeOH:Ace1:1); centrifugation at 9000 rpm for 15 min followed by 15 min ultrasonication;ycle number, 3; reconstitution, 18 mL DI + 2 mL MeOH:Ace (1:1).

n sludge samples acquired via different WWTP units by usinghe sludge extraction method optimized in this study. Extractedamples were measured with LC–MS/MS in the same way as theecovery experiments (Tables S1 and S2).

. Results and discussion

The recovery experiments were performed using municipalastewater treatment plant sludge taken from the sludge dryingnit. Comparisons of the results were based on extraction recov-ries obtained through LC–MS/MS measurements after applyingifferent USE and SPE procedures (Section 3.1). The optimizedethod is summarized in Section 3.2 which was applied to sludge

amples taken from different units of two municipal WWTPs instanbul, Turkey (Section 3.3).

.1. Optimization of the extraction procedure

.1.1. Effect of ultrasonication equipmentThere are various studies using ultrasonic homogenizators

43,45,47] or ultrasonic baths [39,42] for the extraction of pharma-euticals from different matrices. However, to our knowledgehe performances were not compared in these studies. There-ore, in this study, performances of two ultrasonication equipmentere tested for the extraction of diclofenac from sludge samples.lthough the PLE method was reported to be more efficient than

he USE method applied with ultrasonic bath in the literature [39],he ultrasonic bath provided the highest extraction recoveries withery low standard deviations for both low and high concentrationevels of diclofenac (Table 1).

In addition, ultrasonic bath application resulted in less crossontamination because direct contact of the sample with equip-ent was eliminated. Whereas ultrasonic homogenizators have a

pecial probe merged directly into the sample that may result inossible contamination. In this study the RSD of the diclofenac con-entrations were at most 8%, whereas Gobel et al. reported RSDalues between 8 and 20% for samples extracted with an ultrasonicomogenizator [41]. In addition to its accessibility, the ultrasonicath is a more practical method, since there is no need to wash thequipment after each use, no extra solvent consumption for wash-ng and no frequent equipment maintenance which is inevitableor ultrasonic homogenizators due to frequent clogging of probesue to the high suspended solids content of sludge.

.1.2. Effect of solvent typeMost frequently used solvents for the extraction of organic pol-

utants are methanol, acetonitrile, acetone, dichloromethane or

ixtures of these solvents in different ratios [40,41,43,45,46]. The

olvent mixture should be selected in a way to increase the relativeffinity of diclofenac while decreasing the dissolution of non-targetrganic compounds originating from the sludge in solvent media.

tested; centrifugation at 9000 rpm for 15 min followed by 15 min ultrasonication;cycle number, 3; reconstitution, 18 mL DI + 2 mL MeOH:Ace (1:1).)

In this study, methanol (MeOH), acetonitrile (ACN), aceton (Ace),MeOH:ACN (1:1) and MeOH:Ace (1:1) were evaluated as extractionsolvents. Although all solvents tested provided sufficient extrac-tion recovery (approximately 80%) for high concentration levelsof diclofenac (400 ng/g), at low concentration levels (100 ng/g)diclofenac was not as efficiently extracted except when MeOH:Ace(1:1) mixture was used (Fig. 1). Moreover, the performances ofMeOH, ACN and Ace were individually below the accepted recoverylevel (70%) for diclofenac extraction from sludge at low concentra-tion levels. In a study conducted for the extraction of diclofenacfrom soil samples, the recoveries obtained with MeOH were similarto the results of this study, but Ace provided higher recoveries forsoil as compared to the results for sludge samples [43]. These resultssuggest that the application of a method developed for soil samplescould be irrelevant when establishing an extraction procedure forsludge samples. When MeOH and Ace (individual extraction effi-ciencies of 50% and 65%, respectively) were used as a mixture, theextraction recovery was high (>90%). ACN was the least efficientsolvent for the extraction of low concentration levels of diclofenac(30%) and did not enhance the extraction efficiency when mixedwith MeOH, which could be due to the low selectivity of ACN fordiclofenac.

3.1.3. Effect of cycle numberThe requirement for more than one extraction cycle is the main

drawback of the USE technique and hence the cycle number isthe critical factor in the application of USE as a practical method.In most studies, USE is applied in three cycles [39,43,44]; how-ever, there are few studies reporting the efficiencies of these cyclesindividually [45]. In this study, the supernatant of each cycle wasmeasured separately in order to determine the contribution of eachcycle to the overall extraction recovery. The results showed that thethird cycle was redundant in terms of extraction efficiency (Fig. 2).Although most of the diclofenac was extracted in the first cycle,the second cycle contributed greatly to the recovery, specifically inexperiments conducted with low diclofenac concentrations. Dur-ing the extraction of high concentration levels of diclofenac, therecovery was below the minimum acceptable recovery level in thefirst cycle, while the second cycle contributed significantly to over-

all extraction recovery. Thus, the use of only two cycles during USEis sufficient for an efficient extraction of diclofenac from sludgesamples, which improves the practicability of the method.

E. Topuz et al. / J. Chromatogr. B 958 (2014) 48–54 51

0

10

20

30

40

50

60

70

80

Cycle I Cycle II Cycle III

Recovery

(%

)

100 ng /g

400 ng /g

Fig. 2. Extraction recoveries for each extraction cycle applied in the experiment(1u

3

att(l5aelts

3

nocp8

Ftas

0

10

20

30

40

50

60

70

80

90

100

5 min 10 mi n 15 mi n 20 mi n

Recovery

(%

)

100 ng /g

400 ng /g

Fig. 4. Extraction recoveries for different cycle times tested (mean value of the three

mean value of the three replicates). (Extraction conditions: sample amount, 0.5 g;0 mL of MeOH:Ace (1:1); centrifugation at 9000 rpm for 15 min followed by 15 minltrasonication; reconstitution, 18 mL DI + 2 mL MeOH:Ace (1:1).)

.1.4. Effect of solvent volumeSolvent volumes used during USE should be optimized to extract

ll of the target compound with sufficiently low solvent volumeo prevent interferences and excess solvent consumption. Extrac-ion of diclofenac from sludge samples with 10 mL of MeOH:Ace1:1) mixture was successful for both high and low concentrationevels of diclofenac (90% extraction recoveries) (Fig. 3). However,

mL of MeOH:Ace (1:1) mixture could not extract diclofenac at minimum acceptable level (<70%). Similarly, very low extractionfficiency was achieved using 15 mL of MeOH:Ace (1:1) mixture forow concentration levels of diclofenac due to a possible increase inhe amount of organic interferences originating from the sludgeamples as a result of high solvent application.

.1.5. Effect of ultrasonication timeAfter optimization of solvent type, solvent volume and cycle

umber for high extraction recovery, the method was further

ptimized by altering the ultrasonication time with various con-entration levels. For high concentration levels of diclofenac, USErovided similar extraction recoveries for 5, 10 and 15 min (about0%) (Fig. 4). However, for low concentration levels, extraction

0

10

20

30

40

50

60

70

80

90

100

5 m L 10 mL 15 mL

Recovery

(%

)

100 ng/g

400 ng /g

ig. 3. Extraction recoveries for different solvent volumes tested (mean value ofhe three replicates). (Extraction conditions: sample amount, 0.5 g; centrifugationt 9000 rpm for 15 min followed by 15 min ultrasonication; cycle number, 2; recon-titution, 18 mL DI + 2 mL MeOH:Ace (1:1).)

replicates). (Extraction conditions: sample amount, 0.5 g; 10 mL of MeOH:Ace (1:1);centrifugation at 9000 rpm for 15 min followed by ultrasonication; cycle number, 2;reconstitution, 18 mL DI + 2 mL MeOH:Ace (1:1).)

recoveries were around 70% (which is the minimum acceptablelevel) and 15 min of ultrasonication resulted in recoveries greaterthan 90%. Therefore, a 5 min ultrasonication time will be effectivefor complete recovery of samples with high concentration levelsof diclofenac whereas the ultrasonication time should be increasedfor samples with low concentration levels of diclofenac. Althoughthe effect of extraction cycle time has not been discussed in liter-ature, Yu and Wu [45] established a method for the extraction ofpharmaceuticals and personal care products from sewage sludgewhich takes 60 min and uses only 12 mL of solvent.

3.1.6. Effect of sludge amountOptimization of the sludge amount is necessary to establish a

sensitive method since the amount of sludge is directly relatedto matrix effect. Most of the studies in literature used at least0.5 g of sludge [41,46] and in this study, 0.1 g, 0.2 g, 0.5 g, and 1 gsludge amounts were used. While extraction of 0.1 g sludge for both

low and high concentration levels of diclofenac resulted in accept-able recovery levels (about 80%), 0.5 g of sludge sample providedthe highest recovery for low concentration levels of diclofenac(Fig. 5). The presence of an excessive amount of interfering organic

0

20

40

60

80

100

120

0.1 g 0.2 g 0.5 g 1 g

Recovery

(%

)

100 ng /g

400 ng /g

Fig. 5. Extraction recoveries for different sludge amounts tested (mean value ofthe three replicates). (Extraction conditions: 10 mL of MeOH:Ace (1:1); centrifuga-tion at 9000 rpm for 15 min followed by 15 min ultrasonication; cycle number, 2;reconstitution, 18 mL DI + 2 mL MeOH:Ace (1:1).)

52 E. Topuz et al. / J. Chromatogr. B 958 (2014) 48–54

Table 2Extraction recoveries for different volume of dissolution solvent (mean value of thethree replicates).

Diclofenacconcentration (ng/gdry matter)

Dissolution solventMeOH:Ace (1:1) (mL)

Recovery(%)

Std dev(%)

100 2 94 24 66 12

400 2 84 74 70 8

Etr

cdlofu[efs

3

dtot(btrl

3

tqSWnrMArp

3

eavmvfi1dswst

Add 10 mL of MeOH:Ace (1:1) mi xture

Keep in ultrasonic bath for 15 minutes

Centri fuge at 90 00 rpm for 15 mi nutes

Supernatant

Filt er from 0.22 µm

Add 10 mL of MeOH:Ace (1:1)

mixture

Keep in ul tras onic bath for 15

minutes

Centrifuge at 9000 rpm for 15

minutes

Fil ter from 0.22 µm

Supern atant

Soli d phase

Evaporate the solve nt with rotary ev aporat or

Diss olve rem aining diclofenac in

18 mL DI + 2mL of MeOH:A ce (1:1)

SPE procedu re

Weigh 0. 1 g slu dge sampl e

Spi ke isotope la beled

internal standard diclofenac-d4

xtraction conditions: sample amount, 0.5 g; 10 mL of MeOH:Ace (1:1); centrifuga-ion at 9000 rpm for 15 min followed by 15 min ultrasonication; cycle number, 3;econstitution to 20 mL with DI and MeOH:Ace (1:1).

ompounds [8,48] resulted in problems for the measurement oficlofenac when 1 g of sludge sample was extracted, especially for

ow concentrations of diclofenac. Andreu and Pico [49] also pointedut the low efficiency of conventional phases of SPE cartridgesor the discrimination of pesticides and their transformation prod-cts/humic substances. To our knowledge, the study by Yu and Wu45] is the only study investigating the effect of sludge amount onxtraction efficiency, where they reported no significant differenceor 0.5 g and 1 g of sludge, however, SPE efficiency was low for 3 gludge sample.

.1.7. Effect of dissolution solvent volumeIncreasing the volume of solvent used for dissolving extracted

iclofenac before the SPE procedure can be expected to increase thearget compound recovery. However, in this study, the dissolutionf the extracted diclofenac in 4 mL of MeOH:Ace (1:1) + 16 mL DI ledo lower recoveries than those obtained with 2 mL of MeOH:Ace1:1) + 18 mL DI (Table 2). This result could be due to the possi-le preference of diclofenac for the dissolution solvent instead ofhe cartridge during the SPE. Thus, all the diclofenac may not beetained in the cartridge and in fact, be washed out with the disso-ution solvent.

.1.8. Solid phase extraction (SPE)Although several cartridges were tested for sample concen-

ration and clean-up, HLB was most effective for diclofenacuantification (data not shown). One of the most critical steps of thePE is the elution of target compounds with an appropriate solvent.hen applied as 4× 2 mL eluent, EtOAc and MTBE separately could

ot elute diclofenac from SPE cartridges and MeOH alone resulted inecoveries higher than %120 (data not shown). On the other hand,eOH–ACN and MeOH–Ace applied as 2× 2 mL MeOH + 2× 2 mL

CN and 2× 2 mL MeOH + 2× 2 mL Ace, respectively led to similarecoveries (approximately 90%) (data not shown). In the optimizedrocedure, 2 mL ACN + 2 mL MeOH was the selected eluent.

.2. Optimized method

Optimization of secondary factors affecting the extractionfficiency, such as cycle number, ultrasonication time, samplemount and SPE conditions, in addition to the main factors (sol-ent type and volume), led to a highly efficient and easy-to-applyethod in terms of not only extraction recovery but also solvent

olume used, time and sample amount. The optimized methodor the extraction of diclofenac from sludge samples is providedn Fig. 6. The LOD and LOQ of the method were determined as

ng/g and 5 ng/g, respectively. To summarize, 0.1 g of freezeried sludge sample was weighed and spiked with diclofenac-d4

tandard. Then 10 mL of MeOH:Ace (1:1) was added to the samples,hich were kept for 15 min in an ultrasonication bath and were

ubsequently centrifuged at 9000 rpm for 15 min. After repeatinghe extraction–ultrasonication–centrifugation steps twice, the

Fig. 6. Optimized method for the extraction of diclofenac from sludge samples.

liquid phases were collected and filtered through 0.22 �m PVDFmembrane filters. The filtered solvents were evaporated in therotary evaporator and then reconstituted with 2 mL MeOH:Ace(1:1) + 18 mL DI. OASIS HLB cartridges were pre-conditioned with10 mL ACN, 5 mL MeOH, and 5 mL DI consecutively, at an approx-imate flowrate of 2 mL/min. Reconstituted samples were elutedfrom cartridges at an approximate flowrate of 5 mL/min followedby washing of the cartridges with 5 mL water and subsequentlythe cartridges were dried under vacuum for 1.5 h to eliminateany remaining water in the cartridges. Diclofenac adsorbed ontocartridges were eluted by using 2 mL ACN and 2 mL MeOH at aflowrate of 2 mL/min. Eluates were blown down until drynessunder a gentle N2 stream (TurboVap II, Caliper Life Sciences, USA)and filtered through 0.22 �m PVDF membrane filters followingreconstitution with MeOH:DI (1:9) to 1 mL and were consequentlyquantified with LC–MS/MS (supplementary info Tables S1 and S2).

3.3. Application of the proposed method

Sludge samples were obtained from two municipal wastewatertreatment plants in Istanbul serving 1 million and 2.4 million peo-ple each (Table 3). Since studies reported that diclofenac removalmechanisms depended on redox conditions [50], samples were col-lected from different units of the treatment plants (i.e. primary

settling, aeration, anaerobic bio-P, anoxic, final settling, anaero-bic digestion, drying and centrifugation unit) in order to monitorthe possible sorption potential of diclofenac under different redoxconditions (aerobic, anoxic, anaerobic).

E. Topuz et al. / J. Chromatogr. B 958 (2014) 48–54 53

Table 3Diclofenac concentrations and characterization of sludge samples in the WWTPs (mean value of the three replicates).

Sampling point WWTP I WWTP II

pH SS (mg/L) COD (mg/L) Diclofenac (ng/gdry matter)

pH SS (mg/L) COD (mg/L) Diclofenac (ng/gdry matter)

Primary settling unit –a –a –a –a 7.65 57,400 ± 495 33,545 ± 890 <5Anaerobic (bio-P) unit 7.08 3720 ± 122 2260 ± 390 <5 7.11 10,920 ± 55 7520 ± 110 <5Aeration unit (anoxictank)

6.94 7760 ± 183 5880 ± 770 <5 7.15 8660 ± 250 4510 ± 100 <5

Aeration unit (aerobictank)

6.88 6550 ± 481 3285 ± 580 <5 7.72 8780 ± 30 4895 ± 110 <5

Final settling unit 6.92 15,073 ± 450 13,415 ± 160 <5 7.01 36,800 ± 450 14,240 ± 2240 <5Anaerobic digestionunit

–a –a –a –a 7.39 37,575 ± 460 10,760 ± 895 <5

Centrifuge unit –b –b –b <5 –b –b –b <5b b b b b b

, Chem

dtIi[1ocsstii(dtdiddpit

4

mawlrdmilsbbmsucttfa

[

[[[

[

Drying unit – – – <5

a Wastewater treatment plant does not have this unit. SS, Suspended Solids; CODb Not measured.

Although the Kow value of diclofenac (4.5–4.8) [51] suggests thaticlofenac can be associated with the sludge phase, it was foundhat the concentration of diclofenac was <5 ng/g for all samples.n addition, diclofenac concentration in the sludge sample usedn the recovery experiments was <5 ng/g. Similarly, Martin et al.52] reported diclofenac concentrations below the LOD as 33.1 ng/g,9.3 ng/g and 1.2 ng/g in sludge samples taken from primary, sec-ndary and digested sludge, respectively. Additionally, diclofenaconcentrations of 5 ng/g were reported for both aerobic and anoxicludge samples [44]. Although the absence of diclofenac in sludgeamples may seem to suggest the inability of the optimized methodo extract diclofenac from sludge, experiments conducted by spik-ng a known amount of diclofenac that is completely homogenizednto the sludge sample (Section 2.2) resulted in high recoveriesapproximately 90%) and proved that the method is able to extracticlofenac from a sludge matrix. Since the wastewater that arriveso these treatment plants contains approximately 380–1005 ng/Liclofenac [53], the concentrations of diclofenac below the LOQ

ndicate that diclofenac does not accumulate on the sludge phaseuring wastewater treatment. Low concentrations of diclofenacetected in sludge samples suggest that the lack of a biodegradationrocess for this compound may allow for diclofenac to directly enter

nto a natural receiving body without any retention in wastewaterreatment plants.

. Conclusion

To evaluate the fate of a pollutant in the environment, experi-ents have to be conducted for different matrices such as aqueous

nd solid phases. Particularly when the compound is present inastewater, its quantification in the sludge phase may be prob-

ematic due to matrix effects and the difficulties associated with itsecovery from the sludge sample. The method for the extraction oficlofenac from wastewater sludge samples should consist of opti-ization for different factors such as type of extraction solvent and

ts volume, ultrasonication duration, amount of sludge to be ana-yzed and volume of the reconstitution solvent. The optimizationteps used in this study are provided in detail and hence it is possi-le to adapt this method for the extraction of other compoundsy applying a similar optimization strategy. Moreover, the USEethod, which is a widely available extraction technique, has been

hown to be effective for the extraction of diclofenac by use of anltrasonic bath. One of the main reasons for the scarcity of studiesoncerning the fate of different emerging pollutants in wastewater

reatment plants is the lack of an easy-to-use and effective extrac-ion method. Thus, this study can be an important reference pointor future studies regarding the quantification of emerging pollut-nts in wastewater sludge.

[

[

– – – <5

ical Oxygen Demand.

Although concentrations measured in the sludge samples arebelow the LOQ, these findings can be used to interpret liquid phasedata in a mass balance study of a WWTP and indicate that thedecrease in the concentration of diclofenac in the aqueous phase isdue to biodegradation and not sorption to the sludge phase. Oncethe lack of diclofenac or any other compound in sludge samplesfrom different units of the WWTP can be demonstrated by the appli-cation or adaptation of this method, more efforts can be focused onthe biodegradation of the pollutants in the aqueous phase in orderto eliminate uncontrolled discharge from WWTPs into receivingbodies.

Acknowledgements

This work was funded by The Scientific and TechnologicalResearch Council of Turkey (TUBITAK project # 110Y319) and ITUBAP Project # 36258.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.jchromb.2014.02.047.

References

[1] K.K. Barnes, D.W. Kolpin, E.T. Furlong, S.D. Zaugg, M.T. Meyer, L.B. Barber, Sci.Total Environ. 402 (2008) 192–200.

[2] M.J. Focazioa, D.W. Kolpin, K.K. Barnes, E.T. Furlong, M.T. Meyer, S.D. Zaugg, L.B.Barbere, M.E. Thurman, Sci. Total Environ. 402 (2008) 201–216.

[3] U. Borgmann, D.T. Bennie, A.L. Ball, V. Palabrica, Chemosphere 66 (2007)1278–1283.

[4] F. Gagne, C. Blaise, C. Andre, Ecotoxicol. Environ. Saf. 64 (2006) 329–336.[5] S.K. Khetan, T.J. Collins, Chem. Rev. 107 (2007) 2319–2364.[6] E. Zuccato, S. Castiglioni, R. Fanelli, G. Reitano, R. Bagnati, C. Chiabrando, F.

Pomati, C. Rossetti, D. Calamari, Environ. Sci. Pollut. Res. 13 (2006) 15–21.[7] C. Zwiener, F.H. Frimmel, Sci. Total Environ. 309 (2003) 201–211.[8] P.K. Jemba, Pharma-Ecology: The Occurrence and Fate of Pharmaceuticals and

Personal Care Products in the Environment, John Wiley & Sons, New Jersey,2008.

[9] J.M. Brozinski, M. Lahti, A. Meierjohann, A. Oikari, L. Kronberg, Environ. Sci.Technol. 47 (2013) 342–348.

10] R. Bort, X. Ponsoda, R. Jover, M.J. Gomez-Lechon, J.V. Castell, J. Pharmacol. Exp.Ther. 288 (1999) 65–72.

11] M. Gros, M. Petrovic, A. Ginebreda, D. Barcelo, Environ. Inter. 36 (2010) 15–26.12] T. Haap, R. Triebskorn, H.R. Koehler, Chemosphere 73 (2008) 353–359.13] J. Hartmann, P. Bartels, U. Mau, M. Witter, W. von Tuempling, J. Hofmann, E.

Nietzschmann, Chemosphere 70 (2008) 453–461.14] K. Kümmerer, J. Environ. Manage. 90 (8) (2009) 2354–2366.

15] K. Kümmerer, Pharmaceuticals in the Environment, Sources, Fate, Effects and

Risks, Springer-Verlag, Berlin/Heidelberg, 2008.16] Global Water Research Coalition Research Agenda, 2008. http://

globalwaterresearchcoalition.net/resagenda.htm (accessed on February2013).

5 matog

[[

[[[[[[

[[

[[

[

[[

[[

[[[

[

[

[

[

[

[[[

[[

[[

[[[

[

[53] S. Sari, G. Özdemir, C. Yangin-Gomec, G.E. Zengin, E. Topuz, E. Aydin,

4 E. Topuz et al. / J. Chro

17] P. Verlicchi, A. Galletti, M. Petrovic, D. Barcelo, J. Hydrol. 389 (2010) 416–428.18] A. Jelic, M. Gros, A. Ginebreda, R. Cespedes-Sanchez, F. Ventura, M. Petrovic, D.

Barcelo, Water Res. 45 (2011) 1165–1176.19] J. Eriksson, J. Svanfelt, L.A. Kronberg, Photochem. Photobiol. 86 (2010) 528–532.20] E. Aydin, I. Talinli, Chemosphere 90 (2013) 2004–2012.21] H.R. Buser, T. Poiger, M.D. Muller, Environ. Sci. Technol. 32 (1998) 3449–3456.22] T. Heberer, Toxicol. Lett. 131 (2002) 5–17.23] K. Lin, J. Gan, Chemosphere 83 (2011) 240–246.24] M. Stumpf, T.A. Ternes, R.D. Wilken, S.V. Rodrigues, W. Baumann, Sci. Total

Environ. 225 (1999) 135–141.25] Y. Zhang, S.U. Geissen, C. Gal, Chemosphere 73 (2008) 1151–1161.26] M. Gardner, S. Comber, M.D. Scrimscaw, E. Cartmell, J. Lester, B. Ellor, Sci. Total

Environ. 437 (2012) 363–372.27] M. Carballa, F. Omil, T. Ternes, J.M. Lema, Water Res. 41 (2007) 2139–2150.28] J. Martin, M. Dolores Camacho-Munoz, J. Luis Santos, I. Aparicio, E. Alonso, J.

Environ. Manage. 102 (2012) 18–25.29] A. Hospido, M. Carballa, M. Moreira, F. Omil, J.M. Lema, G. Feijoo, Water Res. 44

(2010) 3225–3233.30] A.S. Stasinakis, Bioresour. Technol. 121 (2012) 432–440.31] C.G. Daughton, B.W. Brooks, in: W.N. Beyer, J.P. Meador (Eds.), Environmental

Contaminants in Biota: Interpreting Tissue Concentrations, 2nd ed., CRC, Press,Taylor and Francis, Boca Raton, 2011, pp. 286–347.

32] M. Bernhard, J. Mueller, T.R. Knepper, Water Res. 40 (2006) 3419–3428.33] M. Clara, N. Kreuzinger, B. Strenn, O. Gans, H. Kroiss, Water Res. 39 (2005)

97–106.34] J.B. Quintana, S. Weiss, T. Reemtsma, Water Res. 39 (2005) 2654–2664.

35] W.W. Buchberger, J. Chromatogr. A 1218 (2011) 603–618.36] H. Matsuo, H. Sakamoto, K. Arizono, R. Shinohara, Bull. Environ. Contam. Toxi-

col. 87 (2011) 31–35.37] J.L. Tadeo, C. Sanchez-Brunete, B. Albero, A.I. Garcia-Valcarcel, R.A. Perez, Cent.

Eur. J. Chem. 10 (2012) 480–520.

r. B 958 (2014) 48–54

38] O. Zuloaga, P. Navarro, E. Bizkarguenaga, A. Iparraguirre, A. Vallejo, M. Olivares,et al., Anal. Chim. Acta 736 (2012) 7–29.

39] P. Vazquez-Roig, V. Andreu, M. Onghena, C. Blasco, Y. Pico, Anal. Bioanal. Chem.400 (2011) 1287–1301.

40] J. Martin, J. Luis Santos, I. Aparicio, E. Alonso, J. Sep. Sci. 33 (2010) 1760–1766.

41] A. Gobel, A. Thomsen, C.S. McArdell, A.C. Alder, W. Giger, N. Theiss, D. Löffler,T.A. Ternes, J. Chromatogr. A 1085 (2005) 179–189.

42] D. Loffler, T.A. Ternes, J. Chromatogr. A 1021 (2003) 133–144.43] J. Xu, L. Wu, W. Chen, A.C. Chang, J. Chromatogr. A 1202 (2008) 189–195.44] W. Xue, C. Wu, K. Xiao, X. Huang, H. Zhou, H. Tsuno, H. Tanaka, Water Res. 44

(2010) 5999–6010.45] Y. Yu, L. Wu, Talanta 89 (2012) 258–263.46] T.A. Ternes, M. Bonerz, N. Herrmann, D. Loffler, E. Keller, B.B. Lacida, A.C. Alder,

J. Chromatogr. A 1067 (2005) 213–223.47] J. Antonic, E. Heath, Anal. Bioanal. Chem. 387 (2007) 1337–1342.48] A.M. Botero-Coy, M. Ibanez, J.V. Sancho, F. Hernandez, J. Chromatogr. 292 (2013)

132–141.49] V. Andreu, Y. Pico, Trends Analyt. Chem. 23 (10–11) (2004) 772–789.50] C. Avila, C. Reyes, J.M. Bayona, J. Garcia, Water Res. 47 (1) (2013) 315–325.51] Syracuse Research Corporation Physprop database, PhysProp Database,

Environmental Science Center, Syracuse, NY, 2010 http://www.syrres.com(accessed on November 2011).

52] J. Martin, M.D. Camacho-Munoz, J.L. Santos, I. Aparicio, E. Alonso, J. Hazard.Mater. 239–240 (2012) 40–47.

E. Pehlivanoglu-Mantas, D.O. Tas, Seasonal variation of diclofenac concen-tration and its relation with wastewater characteristics at two munic-ipal wastewater treatment plants in Turkey, J. Hazard. Mater. (2014),http://dx.doi.org/10.1016/j.jhazmat.2014.03.015.