ng

e12,arast

Keywords:

ever-indal sediuor p

terephthalate (PET), polystyrene and polyurethane with sizes of approximately 1 mm were between 91

were extracted from three sediment samples collected from the North Sea island Norderney. Using

st modnufactidly gron poly

plastics are entering our seas and oceans, posing a complex andmulti-dimensional challenge with signicant implications for the

e composition ofo be characterisednment (Europeanthere is still littlestics in sediments,loped. Several at-tics in sedimentsm chloride (NaCl)owne et al., 2010;

Claessens et al., 2011). Higher density salts such as zinc chloride(Imhof et al., 2012; Liebezeit and Dubaish, 2012) and polytungstate(Corcoran et al., 2009) were used to enable to higher densitypolymers to be extracted. One important issue in the extraction ofmicroplastics from sediments is the sediment sample mass appliedfor extraction. When NaCl was used for density separation, 1 kgsediments was usually extracted (Thompson et al., 2004; Reddyet al., 2006; Claessens et al., 2011), with the exception of Browneet al. (2010), who used a sediment volume of only 50 mL. When

* Corresponding author.E-mail address: e.fries@brgm.fr (E. Fries).

1 Present address: Direction Eau, Environnement et Ecotechnologies, Bureau deRecherches Gologiques et Minires (BRGM), 3, Avenue Claude Guillemin, B.P.

Contents lists availab

Environment

journal homepage: www.els

Environmental Pollution 184 (2014) 161e16936009, 45060 Orlans Cedex 02, France.in packaging, construction, medicine, electronics, automotive andaeroplane components (Goodship, 2007). In 2009, 230 million tonsof plastics were produced globally per year. Europe is the secondmajor region of plastic production (Plastics Europe, 2010). Germanyis the third major plastic producer after China and the USA (Abts,2010). According to sea-based sources such as shipping, shingand transport activities (Horsman, 1982; Pruter, 1987; Derraik,2002) and land-based sources such as tourism, adjacent in-dustries or river inputs (e.g. Reddy et al., 2006; Browne et al., 2010),

(European Parliament and the Council, 2008) thmicro-particles (in particular microplastics) has tin marine litter in the marine and coastal enviroParliament and the Council, 2010). However,monitoring data on the occurrence of microplasince analytical methods have yet to be devetempts have been made to monitor microplasbased on density separation in solutions of sodiu(Thompson et al., 2004; Reddy et al., 2006; Brin the 1930s and 1940s (Cole et al., 2011). Today, we cannot imaginelife without plastics, which are used in a vast array of elds for use

the global problem (Cole et al., 2011; Andrady, 2011).As an amendment to the Marine Strategy Framework DirectivePyrolysisGas chromatography

1. Introduction

Following the invention of the r1907, inexpensive high-volume madeveloped. These resulted in the rapof plastics, which for the most comm0269-7491/$ e see front matter 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.envpol.2013.07.027pyrolysis gas chromatography/mass spectrometry, these microplastics were identied as PP, PVC and PET. 2013 Elsevier Ltd. All rights reserved.

ern plastic Bakelit inuring techniques wereowing mass productionmer types commenced

marine and coastal environment and human activities all over theworld (UNEP, 2009). For example, in the German Bight (North Sea)anthropogenic debris was one of the three major otsam categories(32.4 items/km2); more than 70% of oating debris was made up ofplastic items (Thiel et al., 2011). It is assumed thatmicroplastics thatare no longer visible to the naked eye represent a major element ofMarine plastic debrisNorth SeaDensity separationand 99%. After being stored for one week in a 35% H2O2 solution, 92% of selected biogenic material haddissolved completely or had lost its colour, whereas the tested polymers were resistant. MicroplasticsA new analytical approach for monitorisediments

Marie-Theres Nuelle a,b, Jens H. Dekiff a,b, Dominiqua Institute of Environmental Systems Research, University of Osnabrueck, BarbarastraebDepartment of Biology/Chemistry, Division of Ecology, University of Osnabrueck, Barb

a r t i c l e i n f o

Article history:Received 11 November 2012Received in revised form15 July 2013Accepted 20 July 2013

a b s t r a c t

A two-step method was dpre-extracted using the ai(NaCl) solution. The originto reduce the volume of soof the whole procedure fAll rights reserved.microplastics in marine

Remy b, Elke Fries a,*,1

D-49076 Osnabrueck, Germanyrae 13, D-49076 Osnabrueck, Germany

loped to extract microplastics from sediments. First, 1 kg sediments wasuced overow (AIO) method, based on uidisation in a sodium chloridediment mass was reduced by up to 80%. As a consequence, it was possiblem iodide (NaI) solution used for the subsequent otation step. Recoveriesolyethylene, polypropylene (PP), polyvinyl chloride (PVC), polyethylene

le at ScienceDirect

al Pollution

evier .com/locate/envpol

disappear and to determinewhether plastic particles would demonstrate resistance.In the second series of tests, 4 ml NaOH and HCl, respectively, were added to

different particles of biogenic organic matter such as parts of leafs and chitin cara-paces (four particles between 0.5 and 1 mm and 10 particles between 2 and 3 mm)stored in 10 ml vials. These experiments were performed on a small scale comparedto those involving H2O2 described above, because it sufced for an approximatecomparison to be made of the three reagents. HCl was only used in a concentrationof 20% because it has been reported that many polymers are not resistant to it in ahigher concentration (Brkle, 2010; Amsler and Frey, 2012). The experimental testconditions for HCl and NaOH were identical to those for H2O2 described above.

A third experiment was performed using 4mL of a 35% H2O2 solution in reactionwith various biogenic organic matter (13 particles between 0.5 and 1 mm and 10particles between 2 and 3 mm). Likewise, 12 plastic particles in two size classes(

ntalLDPE, PE, HDPE, PET, PUR, PS, PC, PA, ABS and EPS) were prepared as describedabove. All particles were photographed using a digital camera coupled to the ste-reomicroscope in a specied position. The surface area of each particle was deter-mined using Image J, an image-processing program. After seven days, particlesthat had not dissolved fully were extracted from the solutions using tweezers andphotographed again using the digital camera coupled to the stereomicroscope (WildM3Z Leica Microsystems, Wetzlar, Germany) in the same position. The surface areaof each particle was determined again and compared to the previous surface area.

2.3. Sediment sampling and sample preparation

Table 2Polymer names and types and supplier of plastic particles used for resistance ex-periments (PE: polyethylene, LDPE: low-density polyethylene, HDPE: high-densitypolyethylene, LLDPE: linear LDPE, PVC: polyvinyl chloride, PC: polycarbonate na-ture, PA: polyamide, ABS: acrylnitrile-butadiene-styrene, PET: polyethylene tere-phthalate, PP: polypropylene, PS: polystyrene).

Polymer name/product

Polymer type Supplier

PE laboratorybottle

PE Lab inventory

LDPE 1840D LDPE (pellet) LyondellBasell Industries AF S.C.A.,Rotterdam, the Netherlands

HDPE ACP9255plus

HDPE (pellet) LyondellBasell Industries AF S.C.A.Rotterdam, the Netherlands

LLDPE 2049E LLDPE (pellet) Dow Plastics, Midland, USARaw material PVC KTK Kunststofftechnik GmbH,

Germering, GermanyRaw material PC KTK Kunststofftechnik GmbH,

Germering, GermanyRaw material PUR KTK Kunststofftechnik GmbH,

Germering, GermanyRaw material PA KTK Kunststofftechnik GmbH,

Germering, GermanyRaw material ABS KTK Kunststofftechnik GmbH,

Germering, GermanyRaw material PET KTK Kunststofftechnik GmbH,

Germering, GermanyPPHC 904 CF PP (pellets) Borealis AG, Vienna, AustriaPS 143E PS (pellets) BASF SE, Ludwigshafen, Germany

M.-T. Nuelle et al. / EnvironmeOn 2 November 2011, three sediment samples were taken from a beach on thenorthern side of the island Norderney, which is about 14 km long and 2.5 km wide.Norderney is one of the eastern Frisian Islands, located on the German coast of theNorth Sea. Norderney is heavily inuenced by tidal streams and periodic stormsurges in spring and autumn, producing a high degree of sediment mobilisation. Thenorthern side of the island, which is part of Lower Saxonys Wadden Sea NationalPark, is characterised by beacheswith awidth ofmore than 250m, far away from theurban area, which is concentrated in the western part of the island. The randomlychosen sampling sites were located in the recreation zone (Zone III), containing themain tourist beach areas. The maximum distance between sampling sites wasapproximately 80 m. The coordinates were 53.722067/7.243967 for sample 1.6u,53.722383/7.247033 for sample 2.6u and 53.722383/7.247025 for sample 2.6 l.Approximately 3 kg sediments were taken randomly from a depth of about 3 cm andplaced in precleaned brown glass bottles using a stainless steel spoon. The spoonwas cleaned using sea water and lint-free paper between each sample. The bottleswere sealed and stored at room temperature in the laboratory until analysis. Next,2 kg wet sediments from each bulk sample was transferred to ceramic bowls, whichwere subsequently covered with aluminium foil. The bowls were placed in a dryingoven at 60 C until the sediments had dried out. A total of 1 kg dry sediments(approximately 600 ml) was weighed out and sieved through a 1 mm mesh. The

s, Bake

ntalmixed completely by shaking. The ask was shaken by hand for about 20 s andthen lled with NaI solution to the highest calibration mark. The ask was thenshaken again for approximately 10 s to limit the amount of minerals on the so-lution surface due to adhesion and surface tension. The ask was then lled withNaI solution to 1 cm below the rim. After a settling time of 10 min, the super-natant was decanted into a 200 mL glass beaker up to about 1 cm below the angleof the belly, enabling the suspension to be shaken again for 10 s. The procedure ofshaking, relling, shaking, relling, sedimentation and decantation was replicatedve times. The supernatant in the 200 mL glass beaker was subsequently poured

Fig. 2. Experimental setup for the AIO method (rst extraction step). A: laboratory standstorage tank for NaCl (25 25 cm), I: outer glass vessel (25 30 cm), J: inner glass be

M.-T. Nuelle et al. / Environme164into a vacuum ltration unit using a nitrocellulose lter (0.45 mm pore width, Mili,China). The glass beaker was rinsed thoroughly with about 250 ml distilled waterinto the vacuum ltration unit. The ltered NaI solution was collected for reuseand the lter was air-dried for further analysis. The lter in the vacuum ltrationunit was then washed with 750 ml more distilled water to remove any salt resi-dues from the lter.

If H2O2 was applied to eliminate biogenic organic particles from the lter cake, afurther ltration step involving passing the supernatant through a new lter wastested to remove NaI from the lter. To this end, the lter cakewas rinsed thoroughlywith about 75 ml distilled water into the 200 ml glass beaker. The content of theglass beaker was ltered again by vacuum ltration through a new lter, andwashedwith approximately 500 ml distilled water to remove any residues of NaI. The lterwas subsequently rinsed with about 20 ml of a 35% H2O2 solution into a glass beakerand stored in the solution at room temperature for seven days.

Since some of the particles on the lter had stuck to the protruding upper edgeof the top glass funnel of the vacuum ltration unit, these adhered particles wererinsed carefully with a small amount of distilled water into a glass petri dish. Finally,the lter was transferred to the second glass petri dish. The ltered NaI solutionwascollected for reuse, and the content of both petri dishes was air-dried for furtheranalysis and covered with aluminium foil.

In addition, an investigation was carried out to ascertain whether a separatoryfunnel (with volumes of 0.25, 1 and 2 L) is applicable as a suitable alternative tovolumetric asks. Sediments were taken from the 500 mL glass beaker and placedin the separatory funnel, and the NaI solution was added to roughly three-quartersof the funnels. After shaking for approximately 10 s and following a settling time of10 min, the deposed sediments were drained off the funnel by opening the stop-cock. Shortly before the sinking surface layer and the oating microplastics reachedthe zone above the stopcock, the stopcock was instantaneously closed. As aconsequence, only the surface layer of the solution with the oating plasticsremained in the funnel. The content of the funnel was then ltered as describedabove.

An optical analysis of the lters was performed using a stereomicroscope (WildM3Z, Leica Microsystems, Wetzlar, Germany) providing 6.5-fold up to 40-foldmagnications. Optical images of particles were taken using a digital camera,which was connected to the microscope via a phototube. Particles that were opti-cally identied as potential plastics were separated using tweezers.2.6. Pyr-GC/MS analysis

Extracted particles were analysed using Pyr-GC/MS by applying a MultipurposeSampler 2XL equipped with the thermal desorption system (TDS 3) (Gerstel, Muel-heim, Germany). Pyrolysis GC is used to obtain structural information about mac-romolecules by carrying out a GC/MS analysis of their thermal degradation products.Ten polymer types of the most common standard polymers (PVC, PC, PUR, PA, ABS,PET, LDPE/EVA, PP, PS and EVA, see Table 2) were also analysed using Pyr-GC/MS toobtain a pyrogram database. Details on the polymers, experimental conditions and

, C: clamps, D: glass tube, E: laboratory fume, F: glass tube, G: indoor fountain pump, H:r (2 L), K: double sockets with parallel metal rods.Pollution 184 (2014) 161e169resulting pyrograms are provided by Fries et al. (2013). Briey, the particle waspyrolysed at 700 C for 60 s. The temperature of the transfer linewas 350 C. In orderto separate and detect pyrolysis products, the TDS was interfaced to a gas chro-matograph (GC) 7890A coupled to amass selective detector (MS) 5975C (bothAgilentTechnologies, Santa Clara, USA). A 30 m HP-5MS capillary column (Agilent Technol-ogies, Santa Clara, USA) with an inner diameter of 250 mm and a lm thickness of0.25 mm was used to achieve chromatographic separation. The GC oven was pro-grammed from 40 to 180 C at 15 C/min and then to 300 C at 5 C/min, held for12 min. The carrier gas was helium (purity 5.0) with a vent ow of 60 mL/min. Massspectra of organic plastic additives (OPAs) and pyrolysis products were obtained byrunning the MS in full scan mode with a mass range between 10 and 600 amu. Py-rolysis products were identied by consulting the NIST05 mass spectra library.

3. Results

3.1. Removal of biogenic organic matter

First, the results for the 30% H2O2 solution are presented. Con-cerning biogenic organic particles >1 mm, 50% (most of which wasof animal origin) had completely dissolved after seven days. Theother 50% also demonstrated obvious reactions, i.e. they becamediscoloured or transparent, or had partly dissolved. Results forbiogenic particles 1mmin size also showed visible changes for PA, PC and PP, which weremore transparent, smaller and/or thinner after exposure. Deniteoptical changes were also determined for PET (brownish colour)and LLDPE (fragmented). For particles

polymers was slightly more transparent, thinner or smaller (PVC,PET, PA, PUR, PP, LDPE, LLDPE). It was observed that PC was de-nitely thinner than before.

Results of the experiments involving NaOH and HCl revealedthat overall optical changes to biogenic organic particles wereweaker than the reaction using 30% H2O2. None of the biogenicorganic particles had dissolved completely or become fully trans-parent. A strong reaction was only observed with 30% NaOH and20% HCl, respectively, with a beetle carapace.

The experiments involving the 35% H2O2 solution revealed thateach biogenic organic particle showed a visible, measurable reac-tion, as with the particles exposed to the 30% H2O2 solution. Theoutcome regarding biogenic organic particles >1 mmwas that fourparticles had dissolved completely, eight were transparent and one

of particle #3with that obtained fromthe Pyr-GC/MSanalysis of PVCrevealed a match of 13 major peaks, albeit with different relativepeak intensities. The following characteristic peaks were identied(RT are given in brackets): o-xylene (RT: 3.989 min), p-xylene (RT:4.223 min), napthalene (RT: 7.213 min), 1-methylnapthalene (RT:8.312min), 2-methylnapthalene (RT: 8.419min), acenaphthene (RT:9.976 min), uorene (RT: 10.925 min) and anthracene (RT:13.171 min). Hence, an original PVC-based polymer is likely. Acomparison of the pyrogram obtained from the pyrolysis of PP andthe transparent white particle #4 revealed that this particle was PP.Important evidence for the presence of PP was the rst peak, whichwas identied as 2,4 dimethyl-1-heptene e a typical pyrolysisproduct of PP (Tsuge et al., 2011). It was not possible to identifyparticle #1 because there was no comparable type of pyrogram in

M.-T. Nuelle et al. / Environmentalwas inconsistent. About 92% of particles had dissolved completelyor were becoming discoloured. Regardingmicro-particles (

All sediment samples collected fromNorderney contained bres substances whilst ensuring that plastics remain resistant to it. The

Table 4Recoveries for seven different polymer types from sediment samples (n 6 for two step-extraction, n 4 for two step-extraction including H2O2 oxidation of the material onthe lter) including mean recoveries (PE: polyethylene, PP: polypropylene, PVC: polyvinyl chloride, PET: polyethylene terephthalate, PS: polystyrene, EPS: expanded PS, PUR:polyurethane, RSD: relative standard deviation). The given sample mass is the mass of sediments resulting from the rst extraction step.

Run Reduced sample mass (g) Polymer type

PE PP PVC PET PS EPS PUR

Particles counted on the lter

1 74.1 10 8 10 10 9 9 102 78.1 10 10 10 10 10 3 103 87 10 10 10 9 10 8 104 43.9 10 9 10 8 10 5 105 74.9 10 10 10 10 8 10 106 122.7 10 10 10 9 7 7 10Mean recovery rate without oxidation

step n 6 (%)100 95 100 93.3 90 70 100

7 69.3 10 9 9 8 10 2 78 74 10 10 10 7 10 8 109 45.5 10 10 10 10 9 8 1010 56.7 9 10 8 10 9 8 9Mean recovery rate with oxidation

step n 4 (%)97.5 97.5 92.5 87.5 95 65 90

Mean recovery rate n 10 (%) (SD) 99 (3.0) 96 (6.6) 97 (6.4) 91 (10.4) 92 (9.8) 68 (24.8) 96 (9.2)

M.-T. Nuelle et al. / Environmental Pollution 184 (2014) 161e169166with a length from about 0.5 mm to a few centimetres, with adiameter of less than 100 mm. These samples were divided intocoloured and uncoloured groups. The latter also contained black,brown and beige bres in addition to mainly translucent bres todistinguish bres of a probable natural origin from obviously col-oured bres e mainly blue, red and green e where an articialorigin was more likely. The numbers of uncoloured and colouredbres were 213 and 21 in sample 1.6u, 55 and 14 in sample 2.6u,and 194 and 10 in sample 2.6.l.

4. Discussion

4.1. Removing biogenic organic matter

The task was to dene an appropriate reagent and reactionconditions that cause the effective deletion of biogenic organicTable 5Densities of polymer types, and European demand for commonly used polymers, densitiused in this study (w/w: weight/weight).

Polymerabbr.

Polymer name European demand[%] (Plastics Europe,2010)

Density [g/cm3](US EPA, 1992)

DensityAdditive(US EPA,

HDPE High-densitypolyethylene

12 0.94e0.97 1.18e1.2

LDPE Low-densitypolyethylene,

17 0.89e0.94

PP Polypropylene 19 0.89e0.91 1.04e1.1PVC Polyvinylchloride 11 1.3e1.58 1.3e1.7PET Polyethylene

terephthalate8 1.29e1.40

PS Polystyrene 8 (incl. PS-E) 1.04e1.08 1.2e1.5EPS Expanded

polystyrene0.015e0.03a

PA Polyamide 2.3 1.07e1.08 1.13e1.6PC Polycarbonate 1.2 1.20PMMA

(acrylic)Polymethyl-methacrylate

0.9 1.17e1.20

SAN Styrolacrylnitrile 1.9 (incl. ABS) 1.02e1.08ABS Acrylonitrile/

butadiene/styrene

1.01e1.08 1.18e1.6

PUR Polyurethane 7 1.17e1.28

a BASF SE (2011).rst step taken before testing a reagent was to verify and comparechemical resistance tables issued by plastics manufacturers. Somepolymers such as PET, PUR, PA and PC are reported to be non-inertor only constrictively inert to H2O2, NaOH and HCl (Brkle GmbH,2010; Polydraack GmbH, 2006). However, tests under respectivetest conditions are recommended. Generally, concentrated acidsand alkalis are known to destroy biological tissue by essentiallycleaving proteins, carbohydrates and fats. However, biogenicorganic particles were dissolved to a lower extent following the useof NaOH and HCl thanwhen H2O2 was applied. Consequently, H2O2was used for further processing. A 30% H2O2 solution was found tobe the ideal reagent for deleting about 50% of biogenic organicmatter, However, a number of visible changes in some polymerswere also observed, including the development of gas bubbles,indicating a chemical reaction between H2O2 and the polymers. Theoptical analysis of polymers before and after exposure to H2O2 wases of minerals mainly present in sediments, and densities of saturated salt solutions

with[g/cm3]1992)

Mineral Density [g/cm3](Matthes, 2001)

Salt solution(saturated)

Density [g/cm3](Schfer andSynowietz, 1984)

8 (PE) Quartz 2.65 NaCl (26% w/w) 1.2

PotassiumFeldspar

2.56 NaI (60% w/w) 1.8

7 Light Mica 2.80Magnetite 5.2Calcite 2.7

2

1

found to be insufcient for evaluating the resistance of polymers toH2O2. In contrast, it emerged that calculating the reduction insurface areas of polymers before and after exposure was anappropriate tool for determining their resistance to a solution of35% H2O2. However, it must bementioned that applying the surface

area calculations apparently caused signicant measurement in-accuracy. This could be seen in the frequent positive values for thechange in surface area. Since an increase in particle size due toreaction with H2O2 is unlikely, the positive values act as an indi-cator of the magnitude of measuring error. Thus an absolute

Fig. 4. Fibres extracted from procedural blank 1 (a, b) and procedural blank 2 (c, d).

M.-T. Nuelle et al. / Environmental Pollution 184 (2014) 161e169 167Fig. 5. Microplastics extracted from sediments collected at Norderney (a: sample 1.6u, b: sam(the length between elongated markings is 1 mm).ple 2.6u, c: sample 2.6 l). The scale bar at the bottom is divided into 100 mmmarkings

ntalmeasurement error was estimated from themean of positive valuesfor surface change, amounting to 6.2% for particles of1 mm in size. When taking this into ac-count, the reduction in size of PP and PE particles 1 mm size class, these re-sults could be interpreted as a minimal change in PP and PE incontact with H2O2. In summary, a week-long treatment using a 35%H2O2 solution is suitable for removing biogenic organic particleswhilst ensuring that all tested polymer types are resistant to it. Theresults of the present study show that the remaining particlesbecame discoloured, i.e. white or transparent, after H2O2 treatment.Since other studies report that microplastics found in marine sed-iments are often naturally coloured white or transparent (reviewedin Hidalgo-Ruz et al., 2012), it must be mentioned that H2O2 couldalso complicate optical analysis rather than facilitate it, since nat-ural particulates can then no longer be distinguished from plasticsthat are whitish or transparent. To conclude, the oxidation stepduring analysis is only recommended if there is a large amount oforganic matter on the lter comprising almost a layer of particles,making visual selection impossible. It is therefore recommendedthat this oxidation step is not applied routinely in the extractionprocedure, but should be determined on a case-by-case basis foreach sample, depending on the quantity of biogenic matter ob-tained in the lter following extraction. In contrast to Liebezeit andDubaish (2012), it is also recommended to expose lters to H2O2instead of the whole original sediment sample, since it may bedifcult to treat 1 kg sediments. Since iodide ions act as catalysts,causing an undesirable decomposition of H2O2 into water and ox-ygen, a reaction of NaI and H2O2 must be avoided by all means. Thereaction is extremely exothermic, and causes strong gas develop-ment. After having conducted a few trials, it became evident thatresidues of NaI can only be eliminated fully by additional ltrationand rinsing through a new lter.

4.2. Separating microplastics from sediments

Our results demonstrate the importance of performing blanksamples to obviate overestimations of microplastic pollution ofmarine sediments when taking microplastic bres into account.From our experience gained whilst developing the method, weassume that extraction methods for microplastics in general arehighly susceptible, particularly to background contamination bybres from the working environment. Since bres contained insamples were optically very similar to those found in proceduralblanks (Fig. 4), contamination of samples with bres must beavoided by all means during sampling and analysis. The sources ofbre background contaminationwere not proven. It was found thatrecycling the NaCl solution had no signicant impact on back-ground contamination. The problem of bre background contami-nation has been touched upon in the past (Fries et al., 2013), and acomparison of background contamination levels among labora-tories was urgently recommended by Hidalgo-Ruz et al. (2012).However, it is likely to be difcult to nd appropriate sedimentsthat are free of bres for running the procedural blank.

By applying the two-step separation method, it was possible toefciently extract common polymers of about 1 mm in size from1 kg sediments. Recovery rates can be fed as correction factors intoevaluations of environmental sample analyses, as suggested byClaessens et al. (2011). Since EPS had the lowest density of thosepolymers tested, its lower recovery rate may be attributed to lossprocesses other than otation. A loss of EPS was observed due to itsstrong adherence behaviour to glass surfaces. Up to now, recoveryrates for extracting microplastics from sediments have only rarely

M.-T. Nuelle et al. / Environme168been reported. The recovery rates of bres and granules/spheresreported in Claessens et al. (2011), using the method of Thompsonet al. (2004) with a number of minor modications, ranged from68.8% to 97.5% (no information is given on recoveries for particularpolymer types). Imhof et al. (2012) achieved a mean recovery rateof 95.5 1.8% for 0.1 g microplastic particles from seven environ-mentally relevant plastic types (PA, PE, PVC, PC, HDPE, PET, and PP)with a diameter

particular advantage is that a lower solution volume of high-

ntaldensity salt (NaI) was required thanks to the reduction in mass ofthe sample achieved in the rst step. Compared to previous at-tempts based on otation in higher density salts, this method iseco-friendly, it incurs low material costs, and the equipment iseasily obtainable. The device for applying the AIO/otation methodcan easily be established in laboratories to monitor the occurrenceof microplastics in sediments, even in countries with only standardlaboratory equipment at their disposal. Since marine plasticpollution is a global problem, this would be a great advantage. Dueto the high contamination potential for bres, additional research isrequired to avoid contamination of samples by bres from theworking environment. Inter-laboratory comparisons and stand-ardised protocols are necessary in the future as a prerequisite toimplementing MSFD in Member States. In this way, quantitativedata on microplastics will be less speculative, due to possibleoverestimations or underestimations. The results of applying thenew method to environmental sediments indicate the occurrenceof microplastics in sediments in Lower Saxonys Wadden Sea Na-tional Park, which contains sensitive habitats that deserve partic-ular protection. Further studies are required to quantify and analysethe spatial distribution patterns of contamination by microplasticson Norderney and other islands along the North Sea coast.

References

Abts, G., 2010. Kunststoff-Wissen fr Einsteiger. Hanser Verlag, Mnchen (inGerman).present study, the use of separatory funnels with different vol-umes for otation was compared to the use of volumetric asks.Flotation in 1 and 2 L separatory funnels entailed a relatively largeamount of minerals in the oating fraction due to the surfacetension of the relatively large solutions surface. In addition,numerous rinsing steps were required due to the large innersurface of the separatory funnels. Decantation of the solution in avolumetric ask turned out to be more effective than draining offfrom a separatory funnel with high volumes (1 and 2 L) in order tominimise rinsing steps to save NaI. It was impossible to use asmaller separatory funnel (250 mL) because the outlet clockedaccording to the low bore width (a larger bore width was notcommercially available).

The two-step extraction method enabled microplastics to beextracted from environmental samples collected from a NationalPark of an East Frisian Island (Norderney) in the North Sea. Theresults of the Pyr-GC/MS analysis revealed the presence of higherdensity polymers such as PETand PVC, demonstrating the necessityto use higher density salts than NaCl. The occurrence of micro-plastics in sediments from a National Park indicates that marineplastic debris can potentially result in the contamination of pro-tected areas with sensitive habitats by microplastics, which oughtto be of special concern.

5. Conclusions

The AIO/otation method based on two-step extraction usingNaCl for pre-extraction to decrease the original sediment samplemass and NaI for the subsequent otation of microplastics wasfound to be efcient for extracting common polymer types,including higher density polymers, from marine sediments. Oneapparatus needs to be constructed, which is the case for theapparatus invented by Imhof et al. (2012).

In previous studies, separatory funnels were often used forotation (e.g. Thompson et al., 2004; Browne et al., 2010). In the

M.-T. Nuelle et al. / EnvironmeAmsler, Frey, A.G., 2012. Chemical Resistance e Summary Table (in German).Schinznach Dorf, Switzerland.Andrady, A.L., 2011. Microplastics in the marine environment. Mar. Pollut. Bull. 62,1596e1605.

BASF SE, 2011. Technical data sheet e StyroporF 295-N (in German). Ludwigshafen,Germany.

Braun, D., 2003. Kunststofftechnik fr Einsteiger. Carl Hanser, Mnchen (in German).Browne, M.A., Galloway, T.S., Thompson, R.C., 2010. Spatial patterns of plastic debris

along estuarine shorelines. Environ. Sci. Technol. 44, 3404e3409.Brkle GmbH, 2010. Nachschlagewerk: Chemische Bestndigkeit von Kunststoffen

(Germany, Bad Bellingen).Claessens, M., Meester, S.D., Landuyt, L.V., Clerck, K.D., Janssen, C.R., 2011. Occur-

rence and distribution of microplastics in marine sediments along the Belgiancoast. Mar. Pollut. Bull. 62, 2199e2204.

Cole, M., Lindeque, P., Halsband, C., Galloway, T.S., 2011. Microplastics as contami-nants in the marine environment: a review. Mar. Pollut. Bull. 62, 2588e2597.

Corcoran, P.L., Biesinger, M.C., Gri, M., 2009. Plastics and beaches: a degradingrelationship. Mar. Pollut. Bull. 58, 80e84.

Derraik, J.G.B., 2002. The pollution of the marine environment by plastic debris: areview. Mar. Pollut. Bull. 44, 842e852.

European Parliament and the Council, 2008. Directive 2008/56/EC of the Europeanparliament and of the Council of 17 June 2008 establishing a framework forcommunity action in the eld of marine environmental policy (Marine StrategyFramework Directive). Off. J. Eur. Union 51, 19e40. L 164.

European Parliament and the Council, 2010. Commission Decision 2010/477/EUof 1 September 2010 on criteria and methodological standards on goodenvironmental status of marine waters (notied under document C(2010)5956). Off. J. Eur. Union 53, 14e24. L 232.

Fries, E., Dekiff, J.H., Willmeyer, J., Nlle, M.-T., Ebert, M., Remy, D., 2013. Identi-cation of polymer types and additives in marine microplastic particles usingpyrolysis-GC/MS and scanning electron microscopy. Environ. Sci. ProcessImpact. http://dx.doi.org/10.1039/c3em00214d (in press).

Goodship, V., 2007. Introduction to Plastics Recycling, second ed. Smithers RapraTechnology Limited.

Hidalgo-Ruz, V., Gutow, L., Thompson, R.C., Thiel, M., 2012. Microplastics in themarine environment: a review of the methods used for identication andquantication. Environ. Sci. Technol. 46, 3060e3075.

Horsman, P.V., 1982. The amount of garbage pollution from merchant ships. Mar.Pollut. Bull. 13, 167e169.

Imhof, H.K., Schmid, J., Niessner, R., Ivleva, N.P., Laforsch, C., 2012. A novel,highly efcient method for the separation and quantication of plasticparticles in sediments of aquatic environments. Limnol. Oceanogr. Methods10, 524e537.

JRC (Joint Research Centre) MSFD GES Technical Subgroup on Marine Litter, 2011.Marine Litter e Technical Recommendations for the Implementation of MSFDRequirements (Luxembourg).

Liebezeit, G., 2008. Marine litter on the Kachelotplate, lower Saxonian Wadden Sea.Senckenberg. Marit. 38, 147e151.

Liebezeit, G., Dubaish, F., 2012. Microplastics in beaches of the East Frisian islandsSpiekeroog and Kachelotplate. Bull. Environ. Contam. Toxicol. 89, 213e217.

Matthes, S., 2001. Mineralogie. Springer-Verlag.Moret-Ferguson, S., Law, K.L., Proskurowski, G., Murphy, E.K., Peacook, E.E.,

Reddy, C.M., 2010. The size, mass, and composition of plastic debris in thewestern North Atlantic Ocean. Mar. Pollut. Bull. 60, 1873e1878.

OSPAR, 2009. Marine Litter in the North-east Atlantic Region: Assessment andPriorities for Response (London).

Plastics Europe, 2010. Plastics e the Facts 2010 An analysis of European plasticsproduction, demand and recovery for 2011. Brussels, Belgium.

Polydraack GmbH, 2006. Nachschlagewerk: Chemische Bestndigkeit von Kunst-stoffen (Germany, Nettetal).

Pruter, A.T., 1987. Sources, quantities and distribution of persistent plastics in themarine-environment. Mar. Pollut. Bull. 18, 305e310.

Reddy, M.S., Basha, S., Adimurthy, S., Ramachandraiah, G., 2006. Description of thesmall plastics fragments in marine sediments along the Alang-Sosiya ship-breaking yard, India. Est. Coast. Shelf Sci. 68, 656e660.

Schfer, K., Synowietz, C., 1984. Chemiker Kalender, third ed. Springer, Berlin,Heidelberg, New York, Tokio.

Shaw, D.G., Day, R.H., 1994. Color-dependent and form-dependent loss of plasticmicro-debris from the North Pacic Ocean. Mar. Pollut. Bull. 28, 39e43.

Thiel, M., Hinojosa, I.A., Joschko, T., Gutow, L., 2011. Spatio-temporal distribution ofoating objects in the German Bight (North Sea). J. Sea Res. 65, 368e379.

Thompson, R.C., Olsen, Y., Mitchell, R.P., Davis, A., Rowland, S.J., John, A.W.G.,McGonigle, D., Russel, A.E., 2004. Lost at sea: where is all the plastic? Science304, 838.

Thompson, R.C., Moore, C.J., Vom Saal, F.S., Swan, S.H., 2009. Plastics, the environ-ment and human health: current consensus and future trends. Phil. Trans. R.Soc. B 364, 2153e2166.

Tsuge, S., Ohtani, H., Watanabe, C., 2011. Pyrolysis-GC/MS Data Book of SyntheticPolymers. Pyrograms, Thermograms andMSof Pyrolyzates. Elsevier, Amsterdam.

UNEP (United Nations Environmental Program), 2009. Marine Litter: a GlobalChallenge. UNEP, Nairobi (Kenya).

US EPA (Environmental Protection Agency), 1992. Plastic Pellets in the AquaticEnvironment: Sources and Recommendations. Final Report 842/B-92/010,Washington DC.

Zurcher, N., 2009. Small Plastic Debris on Beaches in Hong Kong: an Initial Inves-

Pollution 184 (2014) 161e169 169tigation (thesis for the degree of Master of Science in Environmental Manage-ment). University of Hong Kong.

A new analytical approach for monitoring microplastics in marine sediments1 Introduction2 Materials and methods2.1 Chemicals and materials2.2 Tests for eliminating biogenic organic matter2.3 Sediment sampling and sample preparation2.4 Fluidisation (first extraction step)2.5 Flotation (second extraction step)2.6 Pyr-GC/MS analysis

3 Results3.1 Removal of biogenic organic matter3.2 Separating microplastics from sediments3.3 Background contamination3.4 Occurrence of microplastics in sediments from Norderney

4 Discussion4.1 Removing biogenic organic matter4.2 Separating microplastics from sediments

5 ConclusionsReferences