Depletion of selective serotonin reuptake inhibitors during sewage sludge composting

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Depletion of selective serotonin reuptake inhibitors during sewagesludge composting

Terje Vasskog a, Ove Bergersen b,*, Trude Anderssen a, Einar Jensen a, Trine Eggen b

a Deparment of Pharmacy, Faculty of Medicine, University of Tromsø, N-9037 Tromsø, Norwayb Bioforsk, Soil and Environment Division, N-1432 Ås, Norway

a r t i c l e i n f o

Article history:Accepted 9 June 2009

a b s t r a c t

Sewage and sewage sludge is known to contain pharmaceuticals, and since sewage sludge is often used asfertilizer within agriculture, the reduction of the selective serotonin reuptake inhibitors (SSRIs) Citalo-pram, Sertraline, Paroxetine, Fluvoxamine and Fluoxetine during composting has been investigated. Sew-age sludge was spiked with the SSRIs before the composting experiment started, and the concentration ofthe SSRIs in the sludge during a 21 day composting period was measured by liquid phase microextraction(LPME) and high-performance liquid chromatography–mass spectrometry. All the SSRIs had a significantdecrease in concentration during the composting process. The highest reduction rates were measured forFluoxetine and Paroxetine and the lowest for Citalopram. In addition three out of four known SSRI metab-olites were found in all the samples, and two of them showed a significant increase in concentration dur-ing the composting period.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years, the occurrence of pharmaceuticals and personalcare products (PPCPs) in the aquatic environment has been recog-nized as one of the emerging issues in environmental risk assess-ment (Daughton and Ternes, 1999; Heberer, 2002; Sandersonet al., 2004; Fent et al., 2006). Pharmaceuticals such as analgesics,anti-inflammatory drugs, antibiotics, antidepressants and antiepi-leptic drugs used in human medical care are not completely elim-inated in the human body and can enter the municipal sewagesystem as the pharmaceutical itself and as their biologically activemetabolites. Several investigators have examined the removal ofpharmaceutical compounds during the passage through municipalsewage treatment systems (e.g. Ternes, 1998; Simonich et al., 2002;Carballa et al., 2004; Vasskog et al., 2008). The results show thatinfluent and effluent concentrations and elimination rates for dif-ferent compounds vary a lot. Various pharmaceutical compoundshave been detected up to the lg/l level in sewage effluents, down-stream of sewage treatment plants and in surface and groundwater(Heberer, 2002; Jux et al., 2002; Kolpin et al., 2002). In a few casesdrugs have been detected in drinking water at trace-levels (Heber-er, 2002). Fate studies have also been performed of selected PPCPsthrough waste water treatment systems, and even during compost-ing (Buyuksonmez and Sekeroglu, 2005; Xia et al., 2005).

The major pathway for pharmaceuticals into the environment isby human and veterinary usage and in some cases as spill frompharmaceutical plants (Larsson et al., 2007). In addition to sewagetreatment plants, landfills where unused drugs and different per-sonal care products have been disposed may release these com-pounds into the environment. Nowadays, especially inindustrialized countries, regulations are supposed to prevent this.Norwegian regulations instruct all pharmacies to accept old/un-used medicines, and keep these safely stored for destruction orreturning them to the producer. These regulations are availablein Norwegian only in ‘‘Apotekboka 2007/2008”. The public are alsoencouraged to deliver unused medicines back to their local phar-macy. However, there are several old landfills which were estab-lished when the disposal of such compounds was tolerated anddrugs have been detected in leachates from such municipal land-fills (Holm et al., 1995). And even today everyone does not followthe request to deliver unused drugs back to their pharmacy, and in-stead disposes them through the toilet or garbage.

Soil may also be exposed to these chemicals when digestedsewage sludge is applied as fertilizer on agricultural fields or trea-ted wastewater is used for irrigation purposes or groundwaterrecharge.

Six selective serotonin reuptake inhibitors (SSRIs) are on theNorwegian market today, Fluoxetine, Fluvoxamine, Paroxetine,Sertraline, Citalopram and Escitalopram (the pure S-enantiomerof Citalopram). SSRIs are used to treat depressions and other men-tal disorders, and approximately 4% of the Norwegian populationare treated with one of the SSRIs. The total consumption is approx-imately 1350 kg/year (The Norwegian Institute of Public Health,

0956-053X/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.doi:10.1016/j.wasman.2009.06.010

* Corresponding author. Address: Bioforsk, Department for Soil and EnvironmentResearch, Fredrik A. Dahls vei 20, N-1432 Ås, Norway. Tel.: +47 926 03 105; fax: +47630 09 410.

E-mail address: [email protected] (O. Bergersen).

Waste Management 29 (2009) 2808–2815

Contents lists available at ScienceDirect

Waste Management

journal homepage: www.elsevier .com/ locate/wasman


2007). Citalopram/Escitalopram is most widely distributed andconstitutes approximately 500 kg of the total amount. SSRIs havebeen found in waste water in Norway (Vasskog et al., 2006; Vas-skog et al., 2008), and it is reasonable to assume that they are pres-ent in sewage sludge. This is also indicated by the generally lowerSSRI concentration in the effluent than in the influent of sewagetreatment plants with only mechanical sieving as the only sewagetreatment. A total concentration of SSRIs of up to 620 ng/l has beenmeasured in waste water effluents in Norway with Citalopram asthe dominating compound (ranging from 9.2 to 612 ng/l) reflectingits high sales amounts (Vasskog et al., 2006; Vasskog et al., 2008).Fluoxetine is the only SSRI that has been detected in sewage sludgein Norway so far, and has been found in two different sewage treat-ment plants at concentrations between 30 and 40 lg/kg dry matter(Norwegian Pollution Control Authority, report no. 949/2006 TA-2156/2006). The other SSRIs were not investigated in this work.The fate of some antibacterials in biosolids were recently investi-gated by Chenxi et al. (2008), showing that some were stable dur-ing a 77 day period, while others deteriorated within a few days.

SSRIs have been shown to be toxic toward aquatic species likeZebra Mussels and Fingernail Clams (Fong, 1998; Fong et al.,1998) and embryos of the gastropod Physa elliptica (Uhler et al.,2000). Brooks et al. (2003) determined the acute toxicity concen-trations of Fluoxetine toward Ceriodaphnia dubia, Daphnia magnaand Pimephales promelas to be in the high lg/l range. Additionallygrowth and fecundity were influenced at somewhat lower concen-trations. Henry et al. (2004) showed that all SSRIs have an impacton survival and reproduction of C. dubia, with Sertraline as themost toxic. None of these studies demonstrated harmful effectsof SSRIs toward non-target organisms at environmentally relevantconcentrations, however, nothing is known about long term expo-sure to low concentrations. To our best knowledge, no studies havebeen conducted on SSRI toxicity toward microorganisms present insewage sludge.

The aim of this work was to investigate the fate of five selectiveserotonin reuptake inhibitors (Citalopram, Fluoxetine, Paroxetine,Sertraline and Fluvoxamine) and four of their metabolites (Desm-ethylcitalopram, Didesmethylcitalopram, Norfluoxetine and Desm-ethylsertraline) during sewage sludge aerobic composting basedon a further development of an existing liquid phase microextrax-tion (LPME) method, followed by analysis with high performanceliquid chromatography (HPLC) coupled to mass spectrometrydetection (Ho et al., 2007). Composting experiments were per-formed in 2 l composting vessels to follow the dissipation of thepharmaceuticals during a mesophilic and thermophilic compostingprocess. Composted sewage sludge is used for several purposes,e.g. as fertilizer in agriculture. Therefore it is important to establishwhether different pollutants such as pharmaceuticals can be re-duced or are present in the sludge even after composting.

2. Materials and methods

2.1. Composting material

Iron sulphate precipitated sewage sludge from a waste watertreatment plant was mixed with a bulking agent consisting ofspruce bark and course reject of garden waste compost (size 2–3 cm), then mixed and inoculated with old mature sludge compostuntil the mixture had a proper porosity. The volumetric ratio ofsewage sludge:bulking agent:mature sludge compost was1:1:0.2. The 16 l sewage sludge/bulking agent/old mature sludgedescribed as compost mixture had an initial pH of 6.5. Dry matterconstituted 26% of the composting mixture and organic matter 72%(w/w) of the dry fraction. Four litre of compost mixture withoutSSRIs was used as control. The remaining 12 l of mixture was

spiked with a stock solution of SSRIs. The stock solution was madeof crushed tablets of each SSRI that was dissolved in 30 ml metha-nol and then diluted with 270 ml distilled water, stirred andheated for two hours at 35 �C. The initial theoretical concentrationof each SSRIs in the experiment was 14.30 mg/kg wet weight and196.5 mg/kg ash weight. After two days of composting it was obvi-ous that the process was inhibited and it was decided to dilute theSSRI-spiked compost mixture 50/50 with extra unspiked compostmixture. The compost mixture in the control vessel was not di-luted, but still it was evident that the trends for the diluted SSRIspiked samples followed the same trends as the control vessel,even though the control experiment was not diluted (Fig. 3).

2.2. Composting experiment

A small scale composting system was selected for the analysis.This is a stationary composting experiment run in 2 l Dewar ves-sels as illustrated in Fig. 1, originally designed for compost stabilitymeasurements. Aeration of the compost was ensured by free airexchange via the open top of the Dewar vessels. Composting ofsource-separated organic household waste has earlier been studiedin Dewar vessels system (Haarstad et al., 2006). The vessel simu-lated an inner cylinder of a large scale composting system.

Six Dewar vessels were filled with compost mixtures and spikedwith SSRIs and started simultaneously. An extra Dewar vessel wasfilled with the compost mixture without SSRIs and run as a control.After two days of composting the composting process seemed tostop in the vessels containing SSRIs, observed as a slow decreasein temperature and CO2 production. This might have been due totoo high SSRI concentrations in the compost mixture, and the SSRIspiked compost mixture was therefore diluted to twice the initialvolume with the unspiked compost mixture. After the dilutionthe new theoretical concentration of each SSRI was 7.2 mg/kgwet weight, corresponding to a total SSRI concentration of35.8 mg/kg. The dry matter of the initial compost mixture was25.9%, and after dilution 29.2% (Table 1). The corresponding drymatter concentration was 24.5 mg/kg of each SSRI after dilution.The initial ash weight was 28.1% of the dry weight, and after dilu-tion 25.2%. The theoretical concentration after dilution was97.9 mg/kg of each SSRI, or a total of 489.5 mg SSRIs/kg ash weight(Table 1). A reduction of the SSRIs might have occurred during thefirst two days, and the concentration at day 2 might be lower than

Fig. 1. Small scale laboratory reactor (Dewar vessel) used in aerobic compostexperiments.

T. Vasskog et al. / Waste Management 29 (2009) 2808–2815 2809


the theoretical concentration calculated from day 0. This mighthave influenced the recovery measurements.

The compost mixture spiked with the SSRIs were pre-treated ina plastic bag twice before composting and possible binding of theSSRIs to the plastic surface could be part of the reason for the dif-ference between the theoretical and the analysed concentration atday 2.

The Dewar vessel containers and the composting process wasrun for a total of 21 days and the containers were harvested after2, 5, 7, 9, 11, 14 and 21 days, respectively. Samples were collectedas triplets from each Dewar vessel and taken out for chemical anal-ysis, dry matter measurement, pH measurement and loss on igni-tion (ash weight). The samples for chemical analysis were storedin a freezer at �18 �C to prevent further biological activity. Theday referred to as day 2 was the day the composting processstarted after dilution with the unspiked compost mixture.

Temperature was monitored in the centre of each Dewar vesselby means of copper–Constantan probes. Data was recorded everyhour and stored in a Delta-T Data logger. A plastic tube mountedadjacent to the temperature sensor enabled gas sampling. O2 andCO2 concentrations were measured by an infra-red spectropho-tometer (GA2000 Landfill Gas Analyser (Geotechnical InstrumentsLtd. (UK)) and read after 30 s sampling time with 300 cm3/minflows. Sufficient supply of air was present, as anoxic conditions(presence of CH4 or H2S) were never observed during theexperiments.

2.3. Chemicals and solvents

Standards of the SSRIs Citalopram (1-[3-(dimethylamino)propyl]-1-(4-fluorophenyl)-1,3-dihydro- 5-isobenzofurancarbonit-rile), Sertraline ((1S-cis)-4-(3,4-dichlorophenyl)-1,2,3,4-tetrahy-dro-N-methyl-1-naphtylamine), Paroxetine (trans-(-)-3-[(1,3-benzo-dioxol-5-yloxy)methyl]benzenepropanamine), Fluvox-amine (5-methoxy-1-[4-(trifluoromethyl)-phenyl]-1-pentanone-O-(2-aminoethyl)oxime) and Fluoxetine ((±)-N-methyl-c-[4(trifluoromethyl)phenoxy]benzenepropanamine), and the metabo-lites Desmethylcitalopram (1-(4-fluorophenyl)-1,3-dihydro-1-[3-(methylamino)propyl]-5-isobenzofurancarbonitrile), Didesmethyl-citalopram (1-(3-aminopropyl)-1-(4-fluorophenyl)-1,3-dihydro-5-isobenzofurancarbonitrile), Desmethylsertraline (4-(3,4-dichloro-phenyl)-1,2,3,4-tetrahydro-, (1S,4S)-1-naphthalenamine) andNorfluoxetine (c-[4-(trifluoromethyl)phenoxy]-benzenepropan-amine), the deuterated standard D3-sertraline and the methylatedstandard N-methyl-paroxetine were purchased from TorontoResearch Chemicals (Toronto, ON, Canada). D3-sertraline andN-methyl-paroxetine were used as internal standards (IS) forquantification. All standards were delivered with purity above98%. N-7084 (1-[3-(10, 11-dihydro-5H-dibenzo[a,d]cyclohepten-5-ylidene)propyl]-pyrrolidine) was purchased from Lundbeck

(Copenhagen, Denmark). The structures and pKa values are givenin Fig. 2.

The standards were dissolved in HPLC grade methanol (MeOH)from Merck (Darmstadt, Germany) in concentrations of 100 or200 lg/ml, stored at �18 �C in the dark, and used as stock solu-tions. From this stock solution the daily work solutions were madeby dilution with purified water.

Solvents for HPLC were HPLC grade acetonitrile from Merck,purified water obtained from a MilliQ purification unit from Milli-pore (Bedford, MA, USA), and pro analysis grade formic acid fromMerck. The same water and formic acid were used for LPME, whiledihexyl ether (purum P 97%) from FLUKA (Buchs, Switzerland)was used as organic solvent to fill the pores of the fibre.

2.4. Chemical analysis

Dry matter (105 �C over night) followed by loss on ignition(550 �C in 12 h) were determined in the initial sewage sludgeand structure mixtures before the pharmaceuticals were spikedinto the compost mixtures and throughout the experiment period.

pH was measured by mixing compost with deionised water (ra-tio 1:5 by volume). The pH was measured after 30 min with a Rosselectrode (Orion Instruments).

At each nominated sampling day, one vessel was removed andemptied to provide enough material for destructive analysisincluding dry matter, loss on ignition, pH and replicate analysesof the SSRIs and metabolites.

An already developed extraction method known as hollow fibresupported liquid phase microextraction (HF-LPME or just LPME)was employed to extract the pharmaceutical compounds fromthe samples. The method development is described elsewhere(Ho et al., 2007), and only a brief description will be given here.The method was originally designed for extracting pharmaceuti-cals from water, thus a procedure was developed to extract theSSRIs from the compost mixture into a water phase. Since the SSRIconcentration in the compost mixture was relatively high, and theLPME method has proved to be very sensitive (Vasskog et al.,2008), optimization of this extraction procedure was not consid-ered necessary.

From the samples from day 2, 5, 7, 9, 11, 14, and 21 of the com-posting process 1.0 g was weighed and added to 1.1 l deionisedwater, then 20 ll formic acid was added. Six replicates from eachday were analysed. The solution was stirred for 16 h at 900 roundsper minute (RPM) at room temperature using an IKA Labortechnikmagnetic stirrer from IKA� Werke GmbH & Co. KG (Staufen, Ger-many). After stirring the solution was filtered using Whatman(Maidstone, England) GF/F filters with 0.7 lm retention, and twodifferent dilutions were made (the initial filtrate gave too high con-centrations of SSRIs in the LPME extractions, disrupting the HPLC–MS/MS peaks). A 1:100 dilution was made for the analysis of theSSRIs, while a 1:20 dilution was made for the analysis of themetabolites.

The diluted samples were added internal standards (50 ng N-methylparoxetine and 100 ng D3-sertraline) and 10 ml 5 M NaOHbefore they were extracted by LPME. A 28 cm long plasmaphanpolypropylene hollow fibre (Membrana, Wuppertal, Germany)was used for the extraction, and the pores of the fibre were filledwith dihexyl ether by dipping the hollow fibre into the solventfor 10 s. The excess ether on the fibre surface was removed by dip-ping it in an ultra sonic bath for 2–3 s. Then the fibre lumen wasfilled with 20 ll acceptor phase (purified water/formic acid at pH2), and the ends of the fibre were sealed with a thin metal wire.The same kind of wire was used to keep the fibre hanging in themiddle of the sample bottle. The samples were stirred for 2 h at800 RPM. After extraction the extracts were transferred to Waters(Milford, MA, USA) total recovery HPLC-vials, and 50 ll of a 100 ng/

Table 1Sampling day, pH, ash weight calculated on dry weight and identified SSRIs of thecomposting mixture during the 21 days lab-scale composting vessel experiments.Standard deviations (n = 3) are given in brackets.

Day pH Dry weight (%) Ash weight (%) Sum SSRIs*

0 7.6 (0.15) 25.9 (0.20) 28.1 (0.76)2 (start) 6.6 (0.10) 29.2 (0.62) 25.2 (0.24) 268.65 7.7 (0.06) 27.4 (1.16) 29.4 (0.87) 253.37 7.7 (0.06) 28.8 (0.39) 29.8 (0.75) 212.59 7.1 (0.10) 26.8 (0.69) 31.8 (0.90) 201.111 7.4 (0.11) 31.2 (1.46) 30.1 (0.46) 146.214 7.5 (0.12) 29.5 (0.53) 30.7 (0.14) 175.921 7.5 (0.13) 31.2 (0.50) 30.7 (0.28) 154.5

* Total identified SSRIs (mg/kg ash weight) in the compost mixture. The initialtheoretical concentration calculated at day 2 was 489.5 (mg/kg ash weight).

2810 T. Vasskog et al. / Waste Management 29 (2009) 2808–2815


ml solution of N-7084 (Lundbeck, Copenhagen, Denmark) was thenadded to each extract as an external standard to ensure reliableLC–MS results. This yielded a total volume of approximately 70 ll.

The samples were then analysed using HPLC–MS/MS. The meth-od is described in detail earlier (Vasskog et al., 2008). The separa-tion was conducted on a Waters XBridge C18 column, and themobile phase consisted of water, acetonitrile and formic acid.The mass spectrometric detection was done on a triple quadrupoleinstrument (Micromass Quattro LC, Manchester, UK) in multiplereactions monitoring mode with positive electrospray ionisationfor optimal selectivity and sensitivity. The protonated molecularion [M+H]+ was used as precursor ion for all compounds, and themost intense product ion was selected for quantification.

2.5. Calibration curves and recovery

Calibration curves were made in the concentration range 0.5–250 ng/l. The standard samples were set up and extracted the sameway as the filtered extracts. Standard samples were extracted andanalyzed throughout the experiment period to ensure that the cal-ibration curves were constant over time. Two internal standardswere used for quantification. Sertraline and Desmethylsertralinewere quantified with D3-sertraline as IS, while Paroxetine, Fluvox-amine, Fluoxetine, Norfluoxetine, Citalopram, Desmethylcitalop-ram and Didesmethylcitalopram were quantified with N-methyl-paroxetine as IS. The linearity (R2) of the method ranged between0.9998 (Sertraline) and 0.9851 (Desmethylsertraline) and was con-

Fig. 2. Structures, pKa and log P values of the selected pharmaceuticals, their metabolites and the standards used during the analyses. Log P values calculated by AdvancedChemistry Development (ACD/Labs) Software V8.14 for Solaris, obtained from SciFinder Scholar 2008.



sidered acceptable for all compounds in this concentration range.The day-to-day precision of all compounds except the Citaloprammetabolites were between 8.3% (Sertraline) and 24.6% (Paroxe-tine). The two Citalopram metabolites (Desmethylcitalopram andDidesmethylcitalopram) had day-to-day variations of 50% and 51%.

The recovery of the SSRIs from the compost mixture is calcu-lated as the measured concentration on day 2 divided by the the-oretical concentration on day 2. The compost mixture is not atotally homogenous sample, hence 1 g samples might not be to-tally representative for the whole sample. This is accounted forby doing six replicate measurements, and the results are summa-rized in Table 2.

3. Results


Traditional microbial activity, measured by increased tempera-ture, developed concentration of CO2 and decreased O2 concentra-tion, was observed in the control vessels and are shown in Fig. 3. Inthe spiked composting mixtures a clear inhibition of the compost-ing process was observed the first two days, possibly caused by atoo high SSRI concentration in the compost mixture. The microbial

activity measured as increased temperature, CO2 and O2 concen-tration stopped at 27 �C, 11% and 10% respectively, between day1 and 2 (Fig. 3).

In the control vessel the temperature, developed CO2 and con-sumed O2 increased in a regular manner to day 8. After dilutingthe spiked compost mixture to twice its initial volume the temper-ature started to increase again. During the composting experimentthe temperature increased from 20� C at day 2 to above 52 �C atday 8. Then the temperature decreased slowly throughout theexperiment to just below 30 �C at day 21 (Fig. 3). The microbialactivity increased rapidly the first day and reached a plateau thatlasted until day 8. The highest microbial activity measured insidethe composting mass measured by developed concentration ofCO2 and depletion of O2 was observed when the temperature readbetween 40 and 50 �C. However, microbial activity was observedthrough the whole period of 21 days of composting measured byO2 concentration and CO2 development. The highest microbialactivity was measured as an O2 concentration of 4.6% and as19.8% developed CO2. For comparison the activity in the controlvessel was measured to be 3.8% and 19.1%, respectively.

At day 2 dry matter constituted 29% of the material. This de-creased to below 27% at day 5, before it increased again to 31%on day 21, as illustrated in Table 1. The low increase in dry matter

Fig. 3. Temperature (A and D), concentration of CO2 (B and E), concentration O2 (C and F) measured inside the Dewar vessels. The values are measured through a period of 5,7, 9 days (A, B and C), and 11, 14 and 21 days (D, E and F) in 6 separate Dewar vessels compared with a control vessel without SSRIs.



during the composting period shows that little evaporation hasoccurred.

The ash weight increased during the composting from 25% atday 2 to 31% on day 21.The pH in the composting mixtures was6.6 at day 2, and increased to 7.7 at day 5. After day 11 and inthe rest of composting period the pH was measured at 7.5.

3.2. Pharmaceutical analysis

The five selective serotonin reuptake inhibitors Citalopram, Ser-traline, Paroxetine, Fluvoxamine and Fluoxetine were analyzedduring the composting period to investigate their fate during theprocess. Likewise, their metabolites Desmethylcitalopram,Didesmethylcitalopram, Desmethylsertraline and Norfluoxetinewere analyzed from the same samples to determine whether therewas an increase in their concentration. The compost mixture sam-ples ash weight is more stable through a composting process, andtherefore it was considered that calculating alteration of each SSRI/metabolite per kg ash weight would give the most correct data onthe depletion of SSRIs and increase of metabolites during the com-posting process. Student’s t-test was used to determine whetherthe reduction in SSRI concentration and increase in metabolite con-centration was statistically significant. A one-tailed t-test assum-ing unequal variances, and a confidence of 95% and 99% wasused in the calculation. The results are shown in Fig. 4. Didesmeth-ylcitalopram was not found in any samples. The concentrations atday 2, which was the day the composting reaction started afterdilution with unspiked compost mixture, were used to calculatethe recoveries in the sample extraction. Reduction rate per day inperiod of the first 1–10 days of composting showed that all reduc-tion were found in the beginning and that the last 10 days did notaffect the concentration of SSRIs (Table 2).

4. Discussion


Composting experiments of pharmaceuticals like the SSRIshave, to our best knowledge, never been performed before. There-fore, it was not known whether the spiking of the compostingmaterial would affect the composting process. Unfortunately itseems like the initial concentration of 14.30 mg/kg wet weight ofeach SSRI was too high and inhibited the composting process com-pared with the control (Fig. 3). Which drug(s) that might have hadan inhibiting effect on the composting process, and at whichconcentration(s), are not known. After reducing the initial SSRIconcentration to 50%, the composting process showed a naturaldevelopment which resulted in increased microbial activity mea-sured as increased temperature, developed CO2 and consumed O2

(Fig. 3). During composting of household waste, the temperatureusually increases to 60–75 �C, however, in this study the tempera-ture never got higher than 52 �C both with and without SSRIsspiked into the compost mixture. The slower temperature develop-ment could be caused by several factors, such as a less amount ofprimary carbon source in the sewage sludge. The SSRI concentra-tion might also still be too high and have a reducing effect onthe process; however, the latter seems unlikely as the microbialactivity has not been affected by the SSRI or the formation of SSRImetabolites since the developed CO2 and consumed O2 follow thesame patterns as for the unspiked sludge (Fig. 3).

4.2. Reduction of selective serotonin reuptake inhibitors

As seen from Table 2 the recovery of the SSRIs are generally be-tween 55% and 72%. Fluvoxamine was an exception, and showed arecovery of only 26%. A reduced recovery might be caused by irre-versible binding to the sewage sludge or the other constituents ofthe composting material. The low recovery of Fluvoxamine couldresult in higher uncertainty in the measurements, however, atthe same time the relative standard deviation of the measurementsof Fluvoxamine is no higher than for the other compounds. Therecoveries are assumed to be constant throughout the experimentperiod which is essential to the concentration calculation. Thismight not be entirely true, but to establish the true recoverythroughout the composting period is difficult. One possibilitywould be to spike a blank compost mixture during composting,but the recovery of a freshly spiked sample is likely to differ fromone that is spiked at the start of the experiment and has gonethrough several days of composting. Hence, the use of a constantrecovery rate was seen as the only possibility to calculate theconcentrations.

As seen from Fig. 4 all the SSRIs have a significant decrease inconcentration during the composting experiment. The reductionrates for the SSRIs were all in the range of 0.88–1.31 mg/kg ashweight per day (Table 2). Highest removal rates were measuredfor Paroxetine (1.31) and Fluoxetine (1.23) and the lowest ratewas measured for Sertraline (0.93) and Citalopram (0.88). The finalconcentration of Fluvoxamine on the last day of the experimentwas significantly lower than for the other compounds. This couldhave indicated a high degree of reduction, but it is also evident thatthe concentration of Fluvoxamine is low throughout the entireexperiment. Whether this is due to a high reduction even beforethe composting process started resulting in a low concentration al-ready on day 2, or whether it is due to irreversible binding to thematrix giving generally lower extraction recoveries is not possibleto establish in this work.

However, the reduction rate of Fluvoxamine is average com-pared to the other SSRIs, indicating that irreversible binding to or-ganic matter might be the most probable reason for the low

Table 2Extraction recoveries of the SSRIs and SSRI reduction rates mg/(kg ash weight per days) during high (1–10 days), lower microbial activity (11–21 days) and the whole period.Standard deviations (n = 6) are given in brackets.

Pharmaceutical Identified conc. at day 2(mg/kg ash)

% Recovery Reduction rate day 1–10mg/(kg ash/day)

Reduction rateday 11–21mg/(kg ash/day)

Reduction rate day 1–21mg/(kg ash/day)

Fluvoxamine 25.6 26.2(7.6)

2.09 0.16* 1.07

Citalopram 69.9 71.4(12.7)

2.28 0 0.88

Fluoxetine 56.0 57.2(11.0)

2.69 0 1.23

Sertraline 57.2 58.4(14.4)

2.50 0 0.94

Paroxetine 17.5 61.4(11.6)

2.68 0.06* 1.31

* Not statistically significant.



concentrations measured for this compound. Different removalrates of pharmaceuticals as antibiotics caused by sorption to humicacids or organic matter have been reported by Gu et al., 2007 andPankai et al., 2004.

The reduction of the SSRIs might have been more effective if thecomposting process had followed a more normal trend with highertemperatures over a longer time period. This is indicated by thefact that the reduction rate is much higher the first ten days ofthe experiment, when the temperature, CO2-development andO2-consumption show high microbial activity, while the last10 days with lower microbial activity shows no significant concen-tration reduction of any of the compounds (Table 2). The reductionof SSRIs might be due to co-metabolism with the primary carbonsource in the sludge. If this is the case the reduction of SSRIs willstop when the primary carbon source is diminished, which seemsto occur after approximately 8–10 days.

When it comes to the metabolites, all except Didesmethylcita-lopram were present in the samples. The concentrations are gener-ally low, but with a slight increase during the experimental period.

Desmethylcitalopram, one of the Citalopram metabolites, was themetabolite found in the highest concentration, ranging between0.9 and 3.5 mg/kg ash weight (Fig. 3). The concentration of Nor-fluoxetine, the Fluoxetine metabolite, also increased during thecomposting period; however, the concentration never exceeded170 lg/kg ash weight. The metabolite of Sertraline, Desmethyl-sertraline, was observed in all samples at medium concentrations(0.5–1.0 mg/kg ash weight), but no increase was observed duringcomposting. This might be due to a further transformation of thismetabolite into other products. The metabolites of Fluoxetineand Citalopram, Norfluoxetine and Desmethylcitalopram, respec-tively, had a significant concentration increase during composting.There is also a connection between when the concentration ofthese metabolites become significantly higher than on day 2, andwhen the corresponding pharmaceutical is significantly reducedin concentration compared to day 2. However, the increased con-centration of the known metabolites can not account for the totalreduction of the SSRIs, indicating that other decomposition prod-ucts are formed as well. Moreover, the metabolite concentration

Fig. 4. SSRIs and metabolites concentrations throughout the composting period. Concentrations significantly higher (metabolites) or lower (SSRIs) than day 2 are markedwith letters, (a) denote 95% confidence, while (b) denote 99% confidence.



is not only dependent on the reduction of the SSRIs, but also onfurther decomposition of the metabolites themselves to otherunknown products. This might explain the low increase in metab-olite concentration during the experiment. It would be beneficial toinvestigate other metabolites of the SSRIs during composting, how-ever, at present analytical standards of these compounds are hardto obtain.

In this experiment the recovery measurement might give a low-er value than what was actually the true value, since a certainreduction might have occurred between day 0 and 2 even thoughthe composting process did not start. However, this does not influ-ence the measured reduction rates of the compounds as the recov-ery is assumed to be constant throughout the experiment period.Further it is assumed that the non-extractable phase does not de-grade during composting. If the opposite is true, the reductionrates are in this case underestimated. The presence of metabolitesat day 2 might also indicate that the reduction of SSRIs is not en-tirely biological, but also chemical, since metabolites had beenformed before the composting process started. Whether this wasdue to chemical- and/or biological degradation can not be estab-lished in this work.

The lack of microbial activity in the first attempt to start thecomposting process might be due to a negative effect of the SSRIson the microbes when present in high concentrations. The amountof available SSRIs will probably also affect the reduction. If toomuch is present (without reaching the limit of inhibition of micro-bial growth) the amount degraded by the microbes might be toosmall to be recognized as a significant decrease in concentration,while a too low concentration might lead to no reduction at all ifthe microbes do not adapt to using the SSRIs as an energy source.The SSRIs are not volatile and most likely the measured reductionof the compounds is due to biological and/or chemical degradationand humification. There is no obvious reason why Fluoxetine andParoxetine have higher reduction rates than the other SSRIs, butthey are obviously more susceptible for chemical and/or biologicaldecomposition (Fig. 4, Table 2).

5. Conclusion

The present work shows that selective serotonin reuptakeinhibitors can be reduced in concentration in a composting pro-cess. The highest reduction rates were measured for Fluoxetineand Paroxetine and the lowest for Citalopram. Based on the highsales numbers of Citalopram/Escitalopram and the fact that it isthe predominating SSRI, further studies should be conducted toinvestigate its toxicity toward microorganisms.

It is important to emphasize that the initial test concentrationwas high and based on measured concentrations in STP influentand effluents, higher than expected to be found in sewage sludge.As a consequence of this the experiment might not give a true esti-mate of what will happen during composting of unspiked sludge.Moreover, little is not known about which degradation productsare formed from the SSRIs during composting, and whether thesedecomposition products are more or less toxic to the environmentthan its parent compounds. However, this study shows that minoramounts of some known human metabolites are formed duringcomposting. Further investigations on decomposition productsand reduction rates with environmentally relevant concentrationsshould be conducted.

Acknowledgement

Financial support for this project was provided by Bioforsk andthe Norwegian Research Council through the ‘‘Pharmafate” project(Project No. 171738/S30).

References

Brooks, B.W., Turner, P.K., Stanley, J.K., Weston, J.J., Glidewell, E.A., Foran, C.M.,Slattery, M., Point, T.W.L., Huggett, D.B., 2003. Waterborne and sedimenttoxicity of fluoxetine to select organisms. Chemosphere 52, 135–142.

Buyuksonmez, F., Sekeroglu, S., 2005. Presence of pharmaceuticals and personalcare products (PPCPs) in biosolids and their degradation during composting.Journal of Residuals Science Technology 2, 31–40.

Carballa, M., Omil, F., Lema, J.M., Llompart, M., Garcia-Jares, C., Rodriguez, I., Gomez,M., Ternes, T., 2004. Behavior of pharmaceuticals, cosmetics and hormones in asewage treatment plant. Water Research 38, 2918–2926.

Chenxi, W., Spongberg, A.L., Witter, J.D., 2008. Determination of the persistence ofpharmaceuticals in biosolids using liquid-chromatography tandem massspectrometry. Chemosphere 73, 511–518.

Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products inthe environment: agents of subtle change. Environmental Health Perspectives107, 907–938.

Fent, K., Weston, A.A., Caminda, D., 2006. Ecotoxicology of human pharmaceuticals.Aquatic Toxicology 76, 122–159.

Fong, P.P., 1998. Zebra mussel spawning is induced in low concentrations ofputative serotonin reuptake inhibitors. Biology Bulletin 194, 143–149.

Fong, P.P., Huminski, P.T., D’Urso, L.M., 1998. Induction and potentiation ofparturition in fingernail clams (Sphaerium striatinum) by selective serotoninreuptake inhibitors (SSRIs). The Journal of Experimental Zoology 280, 260–264.

Gu, C., Karthikeyan, K.G., Sibley, S.D., Pedersen, J.A., 2007. Complexation of theantibiotic tetracycline with humic acid. Chemosphere 66 (8), 1494–1501.

Haarstad, K., Bergersen, O., Sørheim, R., 2006. Occurrence of carbon monoxideduring organic waste degradation. Journal of the Air and Waste ManagementAssociation 56, 575–580.

Heberer, T., 2002. Occurrence, fate, and removal of pharmaceutical residues in theaquatic environment: a review of recent research data. Toxicology Letters 131,5–17.

Henry, T.B., Kwon, J.-W., Armbrust, K.L., Black, M.C., 2004. Acute and chronic toxicityof five selective serotonin reuptake inhibitors in Ceriodaphnia dubia.Environmental Toxicology and Chemistry 23 (9), 2229–2233.

Ho, T.S., Vasskog, T., Anderssen, T., Jensen, E., Rasmussen, K.E., Pedersen-Bjergaard,S., 2007. 25,000-fold pre-concentration in a single step with liquid-phasemicroextraction. Analytica Chimica Acta 592, 1–8.

Holm, J.V., Rügge, K., Bjerg, P.L., Christensen, T.H., 1995. Occurrence and distributionof pharmaceutical organic compounds in the groundwater downgradient of alandfill (Grindsted, Denmark). Environmental Science and Technology 29,1415–1420.

Jux, U., Baginski, R.M., Arnold, H.-G., Krönke, M., Seng, P.N., 2002. Detection ofpharmaceutical contaminations of river, pond, and tap water from Cologne(Germany) and surroundings. International Journal of Hygiene andEnvironmental Health 205, 393–398.

Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B.,Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewatercontaminants in U.S. streams, 1999–2000: a national reconnaissance.Environmental Science and Technology 36, 1202–1211.

Larsson, D.G.J., de Pedro, C., Paxeus, N., 2007. Effluent from drug manufacturescontains extremely high levels of pharmaceuticals. Journal of HazardousMaterials 148, 751–755.

The Norwegian Institute of Public Health, 2007. Personal Communication/ØysteinKarlstad.

The Norwegian Pollution Control Authority, 2006. Report No. 949/2006 TA-2156/2006. Karlegging av utvalgte legemidler og kosmetikk. Study performed by DetNorske Veritas and Eurofins on behalf of the Norwegian Pollution ControlAuthority.

Pankai, K., Rossman, F.G.J., Diana, S.A., 2004. Investigating the molecularinteractions of oxytetracycline in clay and organic matter: insights on factorsaffecting its mobility in soil. Environmental Science and Technology 38, 4097–4105.

Sanderson, H., Johnson, D.J., Reitsma, T., Brain, R.A., Wilson, C.J., Solomon, K.R., 2004.Ranking and prioritization of environmental risks of pharmaceuticals in surfacewaters. Regulatory Toxicology and Pharmacology 39, 158–183.

Simonich, S.L., Federle, T.W., Eckhoff, W.S., Rottiers, A., Webb, S., Sabaliunas, D.,DeWolf, W., 2002. Removal of fragrance materials during US and Europeanwastewater treatment. Environmental Science and Technology 36, 2839–2847.

Ternes, T.A., 1998. Occurrence of drugs in German sewage treatment plants andrivers. Water Research 32, 3345–3360.

Uhler, G.C., Huminski, P.T., Les, F.T., Fong, P.P., 2000. Cilia-driven rotational behaviorin gastropod (Physa elliptica) embryos induced by serotonin and putativeserotonin reuptake inhibitors (SSRIs). Journal of Experimental Zoology 286,414–421.

Vasskog, T., Berger, U., Samuelsen, P.-J., Kallenborn, R., Jensen, E., 2006. Selectiveserotonin reuptake inhibitors in sewage influents and effluents from Tromsø,Norway. Journal of Chromatography A 1115, 187–195.

Vasskog, T., Anderssen, T., Pedersen-Bjergaard, S., Kallenborn, R., Jensen, E., 2008.Occurrence of selective serotonin reuptake inhibitors in sewage and receivingwaters at Spitsbergen and in Norway. Journal of Chromatography A 1185, 194–205.

Xia, K., Bhandari, A., Kas, K., Pilar, G., 2005. Occurrence and fate of pharmaceuticalsand personal care products (PPCPs) in biosolids. Journal of EnvironmentalQuality 34, 91–104.


Depletion of selective serotonin reuptake inhibitors during sewage sludge composting

Documents

Transcript of Depletion of selective serotonin reuptake inhibitors during sewage sludge composting