lable at ScienceDirect

Environmental Pollution 236 (2018) 916e925

Contents lists avai

Environmental Pollution

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

Evaluation of microplastic release caused by textile washing processesof synthetic fabrics*

Francesca De Falco a, 1, Maria Pia Gullo a, 1, Gennaro Gentile a, Emilia Di Pace a,Mariacristina Cocca a, *, Laura Gelabert b, Marolda Brouta-Agn�esa b, Angels Rovira b,Rosa Escudero b, Raquel Villalba b, Raffaella Mossotti c, Alessio Montarsolo c,Sara Gavignano c, Claudio Tonin c, Maurizio Avella a

a Institute for Polymers, Composites and Biomaterials, Italian National Research Council -Via Campi Flegrei 34, 80078 Pozzuoli, NA, Italyb Leitat Technological Center, C/de la Innovaci�o, 2, 08225 Terrassa, Barcelona, Spainc Institute for Macromolecular Studies, Italian National Research Council, Corso G. Pella 16, 13900 Biella, Italy

a r t i c l e i n f o

Article history:Received 7 March 2017Received in revised form15 September 2017Accepted 14 October 2017Available online 27 October 2017

Keywords:MicroplasticSynthetic fabricTextile washingCounting method

* This paper has been recommended for acceptanc* Corresponding author.

E-mail address: mariacristina.cocca@ipcb.cnr.it (M1 These authors contributed equally.

https://doi.org/10.1016/j.envpol.2017.10.0570269-7491/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t

A new and more alarming source of marine contamination has been recently identified in micro andnanosized plastic fragments. Microplastics are difficult to see with the naked eye and to biodegrade inmarine environment, representing a problem since they can be ingested by plankton or other marineorganisms, potentially entering the food web. An important source of microplastics appears to bethrough sewage contaminated by synthetic fibres from washing clothes. Since this phenomenon stilllacks of a comprehensive analysis, the objective of this contribution was to investigate the role ofwashing processes of synthetic textiles on microplastic release. In particular, an analytical protocol wasset up, based on the filtration of the washing water of synthetic fabrics and on the analysis of the filtersby scanning electron microscopy. The quantification of the microfibre shedding from three differentsynthetic fabric types, woven polyester, knitted polyester, and woven polypropylene, during washingtrials simulating domestic conditions, was achieved and statistically analysed. The highest release ofmicroplastics was recorded for the wash of woven polyester and this phenomenon was correlated to thefabric characteristics. Moreover, the extent of microfibre release from woven polyester fabrics due todifferent detergents, washing parameters and industrial washes was evaluated. The number of micro-fibres released from a typical 5 kg wash load of polyester fabrics was estimated to be over 6,000,000depending on the type of detergent used. The usage of a softener during washes reduces the number ofmicrofibres released of more than 35%. The amount and size of the released microfibres confirm that theycould not be totally retained by wastewater treatments plants, and potentially affect the aquaticenvironment.

© 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Marine contamination caused by plastics debris represents aglobal problem that has become particularly relevant in recentyears, due to the direct impact these pollutants have on the envi-ronment (Gall and Thompson, 2015), or to their potential effects onhuman health (Bouwmeester et al., 2015). Several scientific studies

e by Eddy Y. Zeng.

. Cocca).

have shown that plastics dominate the waste found in oceans andinland waters (Derraik, 2002; Barnes et al., 2009). The United Na-tions Environment Programme, UNEP, estimates that up to 18,000pieces of plastic debris are floating on every square kilometre ofocean (Eriksen et al., 2014). Plastic fragments can be found acrossthe Southwest Pacific in surprisingly high quantities, even inremote and non-industrialised places such as Tonga, Rarotonga andFiji (Gross, 2015; Gregory, 2009). The durability and slow rate ofdegradation allow these fragments, constituted by synthetic poly-mers, to withstand the ocean environment for years to decades orlonger (Sudhakar et al., 2007a,b; Shaw and Day, 1994). It isconsidered that (with the exception of materials that have been

Table 1Fabric type, code, weight and fibre length.

Type of Fabric Code Weight (g/m2) Fibre length (mm)

Plain weave polyester PEC 126 35Double knit jersey polyester PEP 200 -a

Plain weave polypropylene PP 170 50

a PEP yarns are made of continuous fibres.

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925 917

incinerated) all the conventional plastics that have ever beenintroduced into the environment do not degrade, becoming smallerin size as a result of abrasion, weathering, and fragmentation(Thompson et al., 2005). Moreover, many studies suggest that wind,wave action, and density of plastic influence the spread of thesefragments (Thompson et al., 2004; Browne et al., 2010).

Microplastic has been defined as particles smaller than 5 mm(Arthur et al., 2009; Costa et al., 2010). Microplastics have beendetected on beaches and in subtidal sediments worldwide (Browneet al., 2010, 2011; de Lucia et al., 2014; Song et al., 2014), andrepresent a threat for marine biota (Wright et al., 2013; Rochmanet al., 2013) since they can be ingested by plankton (Cole et al.,2013) or other marine organisms (Rochman et al., 2015), eventu-ally entering the human food web (Yang et al., 2015). Severalstudies report that plastics transfer contaminants such as plasti-cizers (Mathalon and Hill, 2014), dyes (Collard et al., 2015) andflame retardants (Schreder and La Guardia, 2014) to marine envi-ronment. Furthermore, these fragments can also adsorb andconcentrate organic pollutants that, once ingested bymarine fauna,could be transferred to the food chain and potentially reachhumans (Rochman et al., 2012; Bakir et al., 2012, 2014). Severalsources of microplastics have been identified. Microplastics derivefrom the deterioration of debris of large dimensions (bags, pack-aging), or are directly produced for a specific application such asabrasives (sandblasting) or additives for cosmetics (such asmicrobeads used for skin scrubs) (GESAMP, 2015; Napper et al.,2015). Another source of microplastics is the domestic and/or in-dustrial washing process of synthetic clothes (Zubris and Richards,2005; Habib et al., 1998; Thompson et al., 2004). In fact, micro-plastics found in marine sediments showed that the proportions ofpolyester and acrylic fibres used in clothing is similar to those foundin habitats that receive sewage-discharges and sewage-effluentsitself (Browne et al., 2011). The release of microplastics from syn-thetic clothes is caused by the mechanical and chemical stressesthat fabrics undergo during a washing process in a laundry ma-chine. Due to their dimensions, a majority of released microfibrescannot be blocked by wastewater treatment plants, reaching in thisway seas and oceans (Magnusson and Wahlberg, 2014).

Consequently, in the last years, a strong need has arisen ofevaluating and quantifying the effects of the release of microfibresduringwashings of synthetic clothes. Several approaches have beendeveloped to evaluate the amount of microfibres shed duringwashings. In particular, by using a gravimetric method, the micro-fibre release from polyester, acrylic and polyester-cotton jumperswas examined during domestic washing cycles carried out at twotemperatures (30 �C and 40 �C) and in presence/absence of adetergent and a fabric conditioner (Napper and Thompson, 2016). Agravimetric method was also applied to evaluate the release ofmicrofibres during washings of polyester jackets or sweaters, eithernew or mechanically aged. The release was discussed taking intoconsideration the type of washing machines (top-versus and front-load), the garment brand and age (Hartline et al., 2016). A similarapproach was also used to determine the amount of microfibresreleased from polyester fleece blankets during washings in do-mestic conditions, in presence of a detergent and a fabric softener(Pirc et al., 2016). In most of the cited works, a conversion formulawas used to transfer the gravimetric results into the number ofmicrofibres released.

Therefore, there is still a lack of information on the directquantification of the microfibres released from standard fabrics dueto laundering, and on the correlation of the release with fabricproperties. Moreover, the role of washing detergents, in liquid andpowder forms, as well as softener, oxidizing and bleaching agents,and parameters such as temperature, time, water hardness andmechanical action, have not been examined yet. The investigation

herein reported was performed to assess the influence of severalwashing parameters, such as those listed above, on microplasticrelease from different synthetic textiles. In order to reach this mainobjective, a new procedure was developed to evaluate the micro-fibre release during washings. Such procedure consists in thefiltration of the washing solutions and the analysis of the filters byscanning electron microscopy (SEM). In this way, a direct quanti-fication of the number and the dimension of the microfibresreleased was obtained. Compared to previous works (Hartline et al.,2016; Napper and Thompson, 2016; Pirc et al., 2016), in addition tothe different adopted approach, the present study also differsbecause it analyses microfibres with very low dimensions. In fact, afilter with a small pore size (5 mm)was used, allowing the detectionof microfibres that could escape through filters with a greater poresize (25 mm in Napper and Thompson, 2016; 20 and 330 mm inHartline et al., 2016; 200 mm in Pirc et al., 2016). Three differentsynthetic fabrics, woven polyester, knitted polyester and wovenpolypropylene, were investigated and quantitative informationwascollected about the amount and dimension of microplasticsreleased during washings simulating domestic conditions. Inaddition to the results related to fabric type, the effect onmicrofibrerelease of different detergents, washing parameters (i.e. tempera-ture, time, water hardness, etc.) and washing conditions (domesticand industrial), was evaluated.

2. Materials and methods

Materials. Three different commercial standard fabrics (Testfa-brics Inc. USA) were selected for the washing experiments: plainweave polyester, double knit jersey polyester and plain weavepolypropylene. The fabric type, code and the weight (g/m2) pro-vided by themanufacturer, alongwith the fibre length, are reportedin Table 1.

The identity of each fabric type was confirmed by FourierTransform Infrared (FTIR) spectroscopy. The spectra are reported inFigs. S1eS3 of the Supporting Information (SI). Untwisted yarns(both warp and weft for woven fabrics), removed from the selectedfabrics, were analysed by optical microscopy using a Stereo mi-croscope Lynx S115 (Vision Engineering, UK).

The detergents used in domestic and industrial washing ex-periments, are listed in Table 2.

Washing Process. Washing tests of synthetic standard fabricswere conducted in Linitest apparatus (URAI S.p.A., Assago, Italy), aslaboratory simulator of a real washing machine, operating in bothdomestic and industrial conditions, in order to correlate fabriccharacteristics and/or washing conditions/laundry products withthe extent of microfibres released. A detailed description of theLinitest apparatus is reported in the SI.

In particular, simulations of domestic washing tests were per-formed according to the ISO 105-C06:2010 standard method usedfor testing the colour fastness of textiles to domestic and com-mercial laundering, using the liquor ratio (liquor:specimen)150:1 vol/wt, corresponding to 150 mL of liquor per gram of fabric,where liquor means the solution constituted bywater plus the doseof detergent. One cycle of the employed washing process simulates

Table 2Laundry products tested during domestic and industrial washing tests.

Type of product Code Compositiona Dose (mL of liquid or gof powder/15 L water)b

Washing pH c Type ofwashed fabric

Domestic washingsDistilled Water R e 7.0 PEC, PEP, PPLight Duty Detergent (LDD, Liquid) DL Anionic and non ionic surfactants, fabric care additives,

enzymes60 mL 7.4 PEC, PEP, PP

Heavy-duty detergents (HDD, Powder) DP Anionic and non ionic surfactants, percabonate,tetraacetylethylenediamine, enzymes

73 g 10.7 PEC, PEP, PP

Oxy-product (liquid) OL Hydrogen peroxide, anionic and non ionic surfactants 85 mL 5.2 PECBleach (liquid) BL Hypochlorite 100 mL 9.7 PECSoftener (liquid) SL cationic surfactants, silicones 40 mL 4.6 PEC

Industrial washings

Distilled Water R e 7.0 PECStandard alkaline detergent solution DL2 Surfactant, sodium hydroxide (in accordance

with UNI EN ISO 105-C12)e 12e12.5 PEC

Sapo Igienbucato IB Nonionic detergent, anionic detergent, otherorganic components

22.5 mL 8.2 PEC

Oxitex OXI Whitening based O2 (Acid 6-phthalimido)-peroxyhexanoic.

15 mL 4.5 PEC

a The composition and brief description of the used detergent is detailed in the SI.b Dose is the amount of detergent as recommended by the manufacturer.c Washing pH is the pH of the liquor determined by using a pHmeter.

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925918

five domestic washing cycles (ISO 105-C06:2010). Fabric speci-mens, with a size of 9 � 9.30 cm2, were sewed with a cotton threadin order to avoid the release of fibres from the cut edges. Then, suchfabric specimens were placed in the steel containers of Linitest,containing 10 steel balls, and washed for 45 min at 40 �C usingdistilled water for the reference wash, R, and distilled water plusthe dose of detergent for the others (see Table 2). The sewing waseffective in preventing the release from the edges, as observed inpreliminary washings reported in the SI.

In order to evaluate the effect of other washing parameters onmicrofibre release, a set of experiments was performed changingtime, temperature, mechanical action and water hardness. For allthese washings, the same commercial liquid detergent was usedwith a dose of 65 mL/15 L water and washing pH of 8.1. In this case,the reference washing test C0 was performed with liquid detergentin distilled water as medium, while the other washing conditionswere obtained by changing the parameters mentioned above. InTable 3, the used washing conditions are summarized.

Another set of washing tests was carried out according to theUNI EN ISO 105-C12 standardmethod, which specify the operationsto simulate an industrial washing process. For industrial washings,fabric specimens with a size of 16 � 8 cm2 were sewed on the sideswith a cotton thread in order to obtain a “bag” of 8 � 8 cm2, asillustrated in Fig. S6 in the SI. This bag was filled with 25 steel ballsand placed in the steel containers, along with other 25 steel balls,and washed for 60 min at 75 �C using distilled water for thereference washes, and solutions containing detergent for the othertests (see Table 2).

All the washing tests performed are summarized in Tables 2 and3. Each washing test was repeated three times.

Table 3Wash trials performed changing the washing parameters.

Condition Temperature (�C) Time (min) N�

C0 40 45 10C1 60 45 10C2 40 45 0C3 40 90 10C4 40 45 10C5 40 45 20

Filtration. The washing effluents, obtained from the wash tests,were filtered by means of a peristaltic pump (Mettler Toledo, flowrate 100 mL/min) connected with Tygon tubes, throughout poly-vinylidene fluoride (PVDF) filters (Durapore®, Merck Millipore), seeFig. 1a, with an average pore width of 5 mm and a diameter of4.7 cm. Then, 400 mL of Milli-Q water at 70 �C were fluxed in thefiltration system, since such amount of water was found optimal toavoid an excess of detergent on the filter surface. The filters weredried at 105 �C for 30 min. The washing effluents coming from asingle washing test, whose volume was more than 150 mLdepending on the weight of the tested fabric, were filtered throughonly 1 filter that was analysed as described belowand never reused.Since each washing test was repeated three times, three filters pertype of wash were obtained. In total 68 filters were analysed.

To avoid cross contamination of fibres among the differentwashes, the Linitest apparatus and the filtration devices werecarefully rinsed with distilled water after each test. In detail, aftereach filtration, tygon tubes were cleaned fluxing about 2000 mL ofMilli-q water while the filter holder was rubbed with a toothbrushto remove any residues of detergent or microfibres, then rinsedwith Milli-q water. 1 filter per wash was used and never reused.Moreover, cotton lab coats and nitrile gloves were worn during allthe experimental work.

Counting Method. In order to determine the amount ofmicroplastics released during the washing tests, the filter surfaceswere analysed using a Scanning electron microscope, SEM, Quanta200 FEG (FEI, The Netherlands). SEM observations were performedin low vacuum mode (PH2O ¼ 0.7 torr), using a large field detector(LFD) and an accelerating voltage of 30 kV. The observations wereconducted on filters mounted on a circular sample stage (diameter

Steel Balls Medium Type of washed fabric

Distilled water PECDistilled water PECDistilled water PECDistilled water PECHard water (27 �d) PECDistilled water PEC

Fig. 1. a) Optical image of a PVDF filter; b) position of the acquired micrographs (Ar) along the filter used in the counting method; c) position of the acquired 121 micrographs of thefilter (extended counting method) used to validate the counting method.

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925 919

7 cm) by using adhesive tape. Due to the low vacuum conditionsthat prevent charging effects on non electrically conductive sam-ples, the filter surfaces were not modified, pretreated or coatedwith any kind of metal layer. The quantitative determination of theamount of microfibres released was performed using the proceduredescribed below, and named as “counting method” from now on.For each filter sample, 21 electronmicrographswere acquired alongtwo orthogonal diameters of the circular filter (see Fig. 1b). Thissampling was chosen since it permits to observe the filter from theborder to the centre of its surface. Every micrograph represents arectangular area (Ar) of the filter surface, equal to 7.8 mm2.

The amount of microfibres, ni, in each micrograph was deter-mined by a visual observation with the help of the public domainsoftware ImageJ (release 1.43u). The number of fibres per unit area,Ci, for each i-image, was calculated according to equation (1):

Ci ¼ ni/Ar (1)

where ni is the number of fibres of the i-image and Ar is the area of asingle rectangle, 7.8 mm2. The total number of fibres per filter, N,was determined by using equation (2)

N ¼ Ci,Atot (2)

where Ci is the average number of fibres per unit area calculated as

Ci ¼ðP21

i¼1CiÞ

21 , and Atot is the total area of the filter (1709.4 mm2).In order to validate the counting method, an extended counting

method was used. In detail, the counting procedure describedabove was applied on 2 filters obtained from different washings.Then, the resulting N value obtained for each filter was comparedwith the number of fibres determined by analysing a wider filtersurface. In particular, for these filter samples, 121 electron micro-graphs were acquired as schematized in Fig. 1c. These 121 micro-graphs cover an overall area of 943.8 mm2, that is 55% of the totalarea of the filter. Using this extended countingmethod, the numberof fibres per filter was calculated by using equation (3)

N ¼ Cj,Atot (3)

where Cj is the average number of fibres per unit area obtained as

Cj ¼ðP121

j¼1CjÞ

121 , and Atot is the total area of the filter (1709.4 mm2).The compatibility between the two methods was determinedthrough two-Sample t-test and non parametric Mann-Whitney Utest.

Several washing tests were conducted changing fabric type ordetergent or washing condition, and each test was replicated three

times. Three filters were obtained by the triplication of eachwashing test, and underwent the described counting method todetermine N per each filter, the average N value among the threefilters (Na) and the related standard deviation (SD). No accumula-tion of microfibres was observed on preferential zones of the filtersurfaces.

Since the tested fabrics differ for weight (g/m2) and fabricspecimens used for washing trials present the same size, thenumber of fibres released per each type of wash was normalized tothe weight of the washed fabrics.

Statistics. Statistical analysis of the number of fibres per unitarea, Ci, for each i-image, was carried out to compare the variouswashings by using OriginPro 8.5 software. The Ci values of the 3filters collected from the triplication of the samewashing test, wereaveraged before the statistical analysis, thus representing a me-dium distribution of the fibres, actually counted on the micro-graphs, along the two diagonals of a filter. The Kolmogorov-Smirnov normality test was used to determine whether data,from each filter, was drawn from a normally distributed population.The compared Ci values came from filters representing two or moredifferent type of wash. In order to assess the differences betweenthe washes per material type/detergent/condition, two-sample t-test and one-way analysis of variance (ANOVA) with Tukey'sposthoc test were used for normally distributed data. The non-parametric Mann-Whitney U (MWU) and Kruskal-Wallis (KW)tests were applied when the assumption of normality was not valid.All tests were applied to assess correlation between number of fi-bres released and type of fabric, used detergent, washing condi-tions and industrial and domestic washings. A 5% significance levelwas used for all statistical tests; p values < 0.05 indicate significantdifference among the data.

Microfibre sizing and weight estimation. SEM micrographs ofthe filter surfaces were analysed by ImageJ to measure the lengthand diameter of the microfibres released. For each washing trial,the average values of the length, L, and diameter, D, were evaluatedalong with the standard deviation, based on the measurements of25 microfibres per filter.

The weight in gram of microfibre released per kg of fabricwashed, was estimated from the average number of microfibresreleased, Na, assuming the fibres were of cylindrical shape,following equation (4) (Napper and Thompson, 2016).

Grams of microfibre=kg fabrics ¼ 1000 ,Na,

�p ,

D2

4,L�,r

(4)

Where r is the density of the material.

Fig. 3. Number of fibres per gram of fabric (Na ± SD) released from woven and knittedpolyester (PEC and PEP, respectively) and woven polypropylene fabrics (PP), duringdomestic washing simulations performed with water (R), liquid detergent (DL) andpowder detergent (DP). In the upper part of the figure, SEM images of the filterscollected by simulating washings of PEC with water, liquid detergent and powderdetergent, are reported (false-colour SEM images).

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925920

3. Results and discussion

Firstly, the countingmethod set up was validated by applying anextended counting procedure, in which a wider filter surface wasanalysed. With this aim, two filters, collected from two differentwashing experiments on woven polyester, were analysed. Filter 1was collected from a washing with only water and filter 2 from awash with liquid detergent. Each filter was counted twice: usingthe counting method and the extended counting method. The re-sults from each method were statistically compared. This compar-ison was performed to confirm that the counting method(performed on 21 SEM images/filter) gives results comparable tothe extended counting method (performed on a larger area of eachfilter). In Fig. 2, the values obtained by applying the countingmethod and the extended counting method on the filters, aregraphed. The results, at a significant level of 0.05, indicated thatthere is no difference between the two methods, allowing toconclude that the counting method set up can be used to evaluatethe amount of microfibres released during washings (filter 1:p ¼ 0.53 - MWU; filter 2: p ¼ 0.40 - t-test).

Fig. 3 reports the results obtained from the three different fab-rics, PEC, PEP and PP, washed in Linitest apparatus under domesticconditions with water, R, as reference, and with liquid, DL, andpowder detergents, DP. In the same figure, an example of SEMmicrographs for each wash trial, is reported. In order to highlightthe presence of microfibres on the filters, the fibres in the reportedmicrographs were coloured in dark grey using a digital imageediting software. The digitally edited SEM images are identified asfalse colour SEM images from now on.

The amount of microfibres released ranged from hundreds tothousands microfibres per filter and the values depended on thekind of fabric tested and on the washing conditions/laundryproducts. In fact, taking into account PEC, the washings performedwith only water produced a release of 162 ± 52 microfibres pergram of fabric that increased to 1273 ± 177 using liquid detergent,and to 3538 ± 664 using powder detergents; a similar trend wasobtained for PEP and PP. These findings indicate that the use ofdetergents, both in liquid and powder form, induce an increase ofmicrofibre release. In particular, the powder product favours themicrofibre shedding more than the liquid one.

It is reported that washing products may significantly reducethe mechanical action during laundering. This tendency is ascribedto the presence of foam, generated by surfactant, and to the ab-sorption of surfactants on fibre surfaces. The first reduce thebeating and rubbing action, thus preventing fabric damage, whilethe surfactants reduce the friction among fibres (Bishop, 1995). In

0

500

1000

1500

2000

FILTER 1 FILTER 2

Fibe

rs/fi

lter

Counting methodExtended counting method

Fig. 2. Comparison of the number of fibres per filter obtained for two different washes,by using the counting method and the extended counting method.

the performed experiments these effects were not detected sincethe composition of the used detergent, in term of amount of sur-factant, was not modified. Moreover, it should be considered thatonly weakly or moderately foaming detergents are permissible inEurope, where the horizontal axis drum-type washing machinesare the most common, in order to avoid overfoaming that reducethe washing performance. For these reasons, foam regulatorsare commonly used to minimize detergent foaming tendencies(Smulders et al., 2001).

The higher release of microfibres caused by powder detergentcould be explained taking into account that it contains inorganiccompounds insoluble inwater, like zeolite, that could cause frictionwith the fabrics. Moreover, the increase in the amount of micro-fibres released could be also related to the higher pH of the powderdetergent. In fact, as reported in literature, though alkali-baseddetergents are effective in removing soil, there is some evidencethat they can induce chemical damage on polyester fabrics bymeans of slow surface hydrolysis (Bishop, 1995). In addition, it isimportant to note that the powder detergent can also induce asignificant error (underestimation) into microfibre determinationsince, as observable in Fig. 4 for PEC samples, the powder detergentinduced on the filter the formation of a thick layer in which themicrofibres were partially or completely embedded, thus makingdifficult their numerical determination.

Statistical analysis confirmed that the amount of microfibresreleased differs significantly depending on the detergent usedduring the washing (PEC: p ¼ 0.00 - ANOVA; PEP: p ¼ 0.00 - KW;PP: p ¼ 0.00 - ANOVA). Tukey post hoc test revealed that theaverage number of microfibres released from PEC samples, washedby using powder detergent, was significantly higher than all othervalues obtained by washing with water or liquid detergent(p ¼ 0.00 in both cases - ANOVA). Furthermore, the amounts ofmicrofibres released from PEP DL and PEP DP samples were

Fig. 4. False-colour SEM images of filter surfaces containing microfibres coming fromPEC washed under domestic condition with: water (a), liquid detergent (b) and powderdetergent (c).

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925 921

significantly higher than PEP R (p ¼ 0.00 in both cases - KW). In thecase of PP samples, the statistical analysis highlighted that theamount of microfibres released from PP DP samples was signifi-cantly higher than PP DL and PP R samples (p ¼ 0.00 in both cases -ANOVA). Usually, the main factor of loss of fibres from a textile ispilling, that consists in fibres entanglement on the textile surfaceand thus in the formation of fibre balls or pills during processes like

washing or wearing. As reported in literature, this phenomenon isrelevant particularly for knitted fabrics (Hussain et al., 2008). In thetests performed in the present work, however, this phenomenonwas not observed and, as a result, knitted polyester released lessmicrofibres than the woven one (see Fig. 3). In order to understandthe mechanism of microfibre release, an optical microscopy anal-ysis was carried out on untwisted yarns (both warp and weft forwoven fabrics), removed from the selected fabrics. As it can beobserved in Fig. 5a, the surface of the knitted polyester yarn ischaracterized by low hairiness, that consists of small fibres thatprotrude from themain yarn core (Haleem andWang, 2014). In fact,the yarn is made of continuous fibres with a very low twist. Fig. 5band c, show that the weft and the warp yarns of woven polyesterare characterized by a different structure: the warp is a doubledyarn and the weft is a single yarn. Both yarns present a high hair-iness. The weft and the warp yarns of woven polypropylene, Fig. 5dee, are both doubled yarns and with high hairiness. Since some ofthe analysed fabrics, PEC and PP, present similar hairiness butopposite trends in the release, see Fig. 3, this parameter could notbe directly related to the release. A textile parameter that couldinstead influence themicrofibre shedding, is the length of the fibresthat compose the yarn. PEP yarns are made of continuous fibers(see Fig. 5a), whereas PEC and PP yarns are made of short staplefibres with a length of 35 and 50 mm respectively (see Table 1).Such difference could affect the release of microfibres. In fact,shorter staple fibers could more easily slip away from the yarnduring the wash, leading to a higher microfibre release, as observedfor PEC and PP. Finally, the weight (g/m2) of the fabrics, reported inTable 1, gives an indication of the material mass per unit area. Thehighest is this value, the highest numbers of fibres are present perunit area. However, as observed before, the microfibres releasedcould not be related to the number of fibres present per unit area,since PEP, that has the greatest weight, is also the fabric thatreleased less microfibres.

Besides the fibres counting, SEM micrographs were also ana-lysed to determine the average dimensions (length and diameter)of the microfibres released. The results indicated that PEC micro-fibres were 340 ± 292 mm in length and 14 ± 3 mm in diameter.Similarly, PEP microfibre length was 478 ± 408 mm and the diam-eter was 20 ± 6 mm. PP microfibres showed a length of339 ± 247 mm and a diameter of 19 ± 6 mm. The microfibredimension was found independent from the detergent used. Theweight in grams ofmicrofibres released per kg of fabric washedwasestimated using equation (4). Such approximation was necessarysince the weight of microfibres released per filter was not deter-minable by gravimetric method. The grams of microfibres releasedper kg of fabric in the case of PEC fabrics were 0.012, 0.092, 0.255 inthe washings with water, liquid detergent and powder detergentrespectively; 0.013, 0.235 and 0.399 g/kg of PEP microfibres werereleased during the washings with R, DL and DP. Finally 0.017, 0.057and 0.146 g/kg of fabrics were released in the case of PEP washedwith R, DL and DP.

Moreover, the analysis of quantitative results obtained byapplying the counting method indicate that, passing from lab-scaleto household washings, a typical 5 kg wash load of polyester fabricscould release an impressive number of microfibres, in the range of6,000,000e17,700,000, corresponding to 0.43e1.27 g of micro-fibres, depending on the type of detergent used.

Since woven polyester produced the greatest release of micro-plastics, further investigations were carried out on this type offabric using the detergents OL, SL and BL (see Table 2), as describedin the experimental section. The results are shown in Fig. 6a,indicating that the washings performed with the softener, SL, andwith the bleaching agent, BL, induced a reduction of fibre losscompared to PEC DL and PEC OL. In particular, the amount of

Fig. 5. Optical microscope images of a) continuous polyester yarn from PEP; b) a staple polyester weft yarn, c) a staple polyester warp yarn from PEC; d) a staple polypropylene weftyarn, e) a staple polypropylene warp yarn from PP.

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925922

microfibres released from PEC DL and PEC OL were significantlydifferent from PEC R (p ¼ 0.01 and p ¼ 0.00 respectively - ANOVA).This trend could not be correlated to the pH of the different de-tergents since, for instance, OL and SL have a similar pH but affectedthe release in opposite ways. These results indicate a mitigatingeffect of the softening and bleaching agents on the number ofmicrofibres released by the fabrics. Concerning the softener, itseffect can be explained by its ability to reduce the friction betweenfibres (Habereder and Bereck, 2002; Yangxin et al., 2008), allowingmicrofibrils to lay parallel to the fibre bundle (Levinson, 1999) andthus decreasing damaging and breaking phenomena. The extensionof such explanation also to the bleach liquid, should be carefullyconsidered since it appears to be in contradiction with precedentstudies performed on cotton fabrics (Kothari et al., 1991; Nair et al.,2013). For such reason, further experiments would be needed toestablish the role of bleaching agents on microfibre release. Arelevant observation is that this trend is different with respect tothat reported in other studies where the use of softeners resulted inan increase of the fibre release from fabrics (Smith and Block, 1982;Chiweshe and Crews, 2000). Nevertheless, it is to be noted that inthe cited works, the main mechanism of fabric deterioration duringwashing is pilling, not evidenced in our study.

Finally, it was estimated that the use of a softener during ahousehold washing of a 5 kg wash load of polyester fabrics, couldreduce the release of microfibresmore than 35% (total release about4,000,000 microfibres) with respect to the amount released duringthe washing under the same conditions but only with a liquiddetergent (about 6,000,000 microfibres).

The amount of microplastics released during the washes per-formed by changing the washing parameters (temperature, time,mechanical action, water hardness), as described in Table 3, is re-ported in Fig. 6b. The obtained results indicate that higher tem-perature (C1), washing time (C3) and mechanical action (C5)produced an increase of microplastics release, even if the recordeddifferences were not very significant. In fact, ANOVA analysisindicated no substantial difference among the washes (p ¼ 0.30).These outcomes could be explained considering a synergistic effectbetween the detergent and the washing parameter. The highertemperature could increase the surface hydrolysis of polyesterfabrics caused by the alkaline detergent, as well as a longer washing

time could extend the fabric exposure to the chemical damageinduced by the alkaline detergent. Moreover, the increased waterhardness could induce fabric abrasion during the test. In fact, asreported in literature for cotton fabrics, the use of hard water inlaundering accelerated the rate of abrasive damage (Morris andPrato, 1976).

Finally, a last set of experiments was performed to simulate theimpact of industrial laundry facilities on the environment, anaspect of the microfibre problem that has never been consideredbefore. In this respect, industrial washings in Linitest were per-formed on woven polyester, as it showed the worst results in thedomestic trials, upon washings with water, as reference, and withthree laundry products. The results are graphed in Fig. 6c. As ex-pected, due to the more aggressive washing conditions, in all casesthe release of microfibres was greater than that obtained underdomestic washing conditions. The presence of liquid detergentssuch as DL2 and OXI induced an increase ofmicrofibre loss, whereasthe release obtained by using IB was closer to R. However, also inthis case, no significant difference among all these washes wasdetected by ANOVA analysis (p ¼ 0.28). The pH of the detergentsseemed to not affect the release, since DL2 and OXI have anopposite pH but a similar influence on the release.

Several studies have identified microplastics in wastewater ef-fluents (Talvitie et al., 2015; Sutton et al., 2016; Ziajahromi et al.,2017; Mintenig et al., 2017) highlighting that they may be anentrance route for microplastics to the aquatic environment.Considering the different municipal wastewater treatment plants(WWTP), with different efficiencies values and effluents concen-trations in terms of particles/liter, a variable percentage of micro-plastics passes through the filtration systems ofWWTPs reaching inthis way the marine environment (Ziajahromi et al., 2017). More-over, some countries with lower infrastructure do not collect andtreat most part of their wastewater (UNEP, 2006). In such scenario,the evaluation of the size fraction of the fibres released during theperformed washing experiments, is a very complex point. Never-theless, a recent study (Boucher and Friot, 2017) has showed thatthe release due to the laundry of synthetic textiles contributes ofabout 35% to the global release of primary microplastics to theworld oceans.

The size and material type of fibres encountered in marine

Fig. 6. Counting results (Na ± SD) related to domestic washing simulations on woven polyester (PEC) with: a) different types of detergents, in the upper part SEM images of thefilters collected by washing with DL, BL and SL are reported; b) different washing conditions, in the upper part SEM images of the filters collected by simulated washing withconditions C0 and C5 are reported; c) industrial washing simulations, in the upper part SEM images of the filters collected by washing with DL2 are reported (false-colour SEMimages).

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925 923

sediments and fauna have been the focus of different researches(Cesa et al., 2017; Rochman et al., 2015). In agreement with the datareported, the size of the microfibres released during our washingtests ranges from 20 to 2000 mm in length. As known, microfibrescan be ingested by marine organisms. In fact, microfibres of about1 mm in diameter, and 15 mm in length, were ingested by planktonicspecies (Frias et al., 2016) and polyethylene terephthalate textilemicrofibres (length range: 62e1400 mm, width 31e528 mm, thick-ness 1e21.5 mm) were ingested by crustacean Daphnia magna,causing an increased mortality of the specie (Jemec et al., 2016).Moreover, textile fibres were also found in fishes and shellfish onsale for human consumption, sampled from markets in Makassar,

Indonesia, and from California, USA (Rochman et al., 2015). On thebasis of such data, the size of the microfibres evaluated in this workmatches the size range with potential negative effects on aquaticorganisms.

4. Conclusion

In this work, an analytical protocol based on the filtration of thewashing effluents of synthetic fabrics and on the analysis of thefilters by scanning electron microscopy, was developed. Such pro-tocol differs from others reported in literature (Hartline et al., 2016;Napper and Thompson, 2016; Pirc et al., 2016) because it is based on

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925924

the direct quantification of low dimension microfibres (filter poresize 5 mm) released during washing trials. The adopted protocolproved to be a useful tool for the evaluation of the extent of therelease from textiles, allowing the identification of specific trendsin the microplastic release, as a function of the textile nature andgeometry, different detergents and washing conditions.

The results showed that woven polyester released the highestnumber of microfibres with respect to knitted polyester and wovenpolypropylene during washing under domestic conditions, inde-pendently of the used detergent. Additional trials performed onwoven polyester pointed out that the lowest release of microfibreswas obtained by using a softener, due to its ability of reducing thefriction among the fibres. Further studies are needed to betterunderstand the role of bleach liquid in the decrease of the numberof microfibres released. Regardless the type of fabric, the resultsindicated that powder detergent, higher temperature, higher waterhardness and mechanical action increased the microplasticsrelease. Finally, as expected, industrial washings produced a sig-nificant release of microfibres.

The approximate number of microfibres released from a typical5 kg wash load of polyester fabrics was calculated to be more than6,000,000 and it is influenced by the type of detergent used.Considering the different efficiency of WWTPs and the amount andthe dimensions of the microfibres collected in this work, a signifi-cant part of them could potentially reach marine environment withnegative effects on aquatic organisms. These results clarify keyfactors (fabric and detergent types, wash conditions and parame-ters) involved in the microfibre release caused by washing pro-cesses of synthetic textiles, which should be taken into account forthe development of mitigation strategies to reduce microfibrepollution. Further experiment to examine the effect of laundry inreal conditions will be performed to corroborate the results ob-tained in lab scale on standard textiles.

Funding

This work was supported by LIFE13 ENV/IT/001069 projectMERMAIDS - Mitigation of microplastics impact caused by textilewashing processes - co-founded by the Lifeþ 2013 programme.

Acknowledgment

The authors acknowledge Mrs. Maria Cristina Del Barone ofLAMeST laboratory of the Institute for Polymers, Composites andBiomaterials, for her technical support in SEM analysis.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttps://doi.org/10.1016/j.envpol.2017.10.057.

References

Arthur, C., Baker, J., Bamford, H., 2009. In: Proceedings of the International ResearchWorkshop on the Occurrence, Effects and Fate of Microplastic Marine Debris,49. NOAA Technical Memorandum NOS-OR&R-30.

Bakir, A., Rowland, S.J., Thompson, R.C., 2012. Competitive sorption of persistentorganic pollutants onto microplastics in the marine environment. Mar. Poll.Bull. 64, 782e2789. https://doi.org/10.1016/j.marpolbul.2012.09.010.

Bakir, A., Rowland, S.J., Thompson, R.C., 2014. Enhanced desorption of persistentorganic pollutants frommicroplastics under simulated physiological conditions.Environ. Pollut. 185, 16e23. DOI: 1016/j.envpol.2013.10.007.

Barnes, D.K.A., Galgani, F., Thompson, R.C., Barlaz, M., 2009. Accumulation andfragmentation of plastic debris in global environments. Philos. Trans. R. Soc.Lond. B Biol. Sci. 364, 1985e1998. https://doi.org/10.1098/rstb.2008.0205.

Bishop, D., 1995. Physical and chemical effects of domestic laundering processes. In:Carr, C.M. (Ed.), Chemistry of the Textiles Industry. Springer, New York,pp. 125e172.

Boucher, J., Friot, D., 2017. Primary Microplastics in the Oceans: a Global Evaluationof Sources. IUCN, Gland, Switzerland. https://doi.org/10.2305/IUCN.CH.2017.01.en, 43pp.

Bouwmeester, H., Hollman, P.C.H., Peters, R.J.B., 2015. Potential health impact ofenvironmentally released micro- and nanoplastics in the human food produc-tion chain: experiences from nanotoxicology. Environ. Sci. Technol. 49,8932e8947. https://doi.org/10.1021/acs.est.5b01090.

Browne, M.A., Galloway, T.S., Thompson, R.C., 2010. Spatial patterns of plastic debrisalong estuarine shorelines. Environ. Sci. Technol. 44, 3404e3409. https://doi.org/10.1021/es903784e.

Browne, M.A., Crump, P., Niven, S.J., Teuten, E.L., Tonkin, A., Galloway, T.,Thompson, R.C., 2011. Accumulations of microplastic on shorelines worldwide:sources and sinks. Environ. Sci. Technol. 45, 9175e9179. https://doi.org/10.1021/es2031505.

Cesa, F.S., Turra, A., Baruque-Ramos, J., 2017. Synthetic fibers as microplastics in themarine environment: a review from textile perspective with a focus on do-mestic washings. Sci. Total Environ. 598, 1116e1129. https://doi.org/10.1016/j.scitotenv.2017.04.172.

Chiweshe, A., Crews, P., 2000. Influence of household fabric softeners and laundryenzymes on pilling ratings and breaking strength. Text. Chem. Color Am. Dyest.Rep. 9, 41e47.

Cole, M., Lindeque, P., Fileman, E., Halsband, C., Goodhead, R., Moger, J.,Galloway, T.S., 2013. Microplastic ingestion by zooplankton. Environ. Sci.Technol. 47, 6646e6655. https://doi.org/10.1021/es400663f.

Collard, F., Gilbert, B., Eppe, G., Parmentier, E., Das, K., 2015. Detection of anthro-pogenic particles in fish stomachs: an isolation method adapted to identifica-tion by raman spectroscopy. Arch. Environ. Contam. Toxicol. 69, 331e339.https://doi.org/10.1007/s00244-015-0221-0.

Costa, M.F., Ivar do Sul, J.A., Silva-Cavalcanti, J.S., Araújo, M.C.B., Spengler, A.,Tourinho, P.S., 2010. On the importance of size of plastic fragments and pelletson the strandline: a snapshot of a Brazilian beach. Environ. Monit. Assess. 168,299e304. https://doi.org/10.1007/s10661-009-1113-4.

de Lucia, G.A., Caliani, I., Marra, S., Camedda, A., Coppa, S., Alcaro, L., Campani, T.,Giannetti, M., Coppola, D., Cicero, A.M., Panti, C., Baini, M., Guerranti, C.,Marsili, L., Massaro, G., Fossi, M.C., Matiddi, M., 2014. Amount and distributionof neustonic micro-plastic off the western Sardinian coast (Central-WesternMediterranean Sea). Mar. Environ. Res. 100, 10e16. https://doi.org/10.1016/j.marenvres.2014.03.017.

Derraik, J.G.B., 2002. The pollution of the marine environment by plastic debris: areview. Mar. Pollut. Bull. 44, 842e852. https://doi.org/10.1016/S0025-326X(02)00220-5.

Eriksen, M., Lebreton, L.C.M., Carson, H.S., Thiel, M., Moore, C.J., Borerro, J.C.,Galgani, F., Ryan, P.G., Reisser, J., 2014. Plastic pollution in the World's oceans:more than 5 trillion plastic pieces weighing over 250,000 tons afloat at sea.PLoS One 9, 1e15. https://doi.org/10.1371/journal.pone.0111913.

Frias, J.P.G.L., Gago, J., Otero, V., Sobral, P., 2016. Microplastics in coastal sedimentsfrom Southern Portuguese shelf waters. Mar. Environ. Res. 114, 24e30. https://doi.org/10.1016/j.marenvres.2015.12.006.

Gall, S.C., Thompson, R.C., 2015. The impact of debris on marine life. Mar. Poll. Bull.92, 170e179. https://doi.org/10.1016/j.marpolbul.2014.12.041.

GESAMP, 2015. Sources, fate and effects of microplastics in the marine environ-ment: a global assessment. In: Kershaw, P.J. (Ed.), IMO/FAO/UNESCO-IOC/UNIDO/WMO/IAEA/UN/UNEP/UNDP Joint Group of Experts on the ScientificAspects of Marine Environmental Protection. Rep Stud GESAMP No. 90, 96 p.

Gregory, M.R., 2009. Environmental implications of plastic debris in marine set-tingsdentanglement, ingestion, smothering, hangers-on, hitch-hiking andalien invasions. Philos. Trans. R. Soc. Lond. B Biol. Sci. 364, 2013e2025. https://doi.org/10.1098/rstb.2008.0265.

Gross, M., 2015. Oceans of plastic waste. Curr. Biol. 25, 93e96. https://doi.org/10.1016/j.cub.2015.01.038.

Habereder, P., Bereck, A., 2002. Part 2: silicone softeners. Rev. Prog. Color 32,125e137. https://doi.org/10.1111/j.1478-4408.2002.tb00256.x.

Habib, D., Locke, D.C., Cannone, L.J., 1998. Synthetic fibers as indicators of municipalsewage sludge, sludge products, and sewage treatment plant effluents. WaterAir & Soil Pollut 103, 1e8. https://doi.org/10.1023/A:1004908110793.

Haleem, N., Wang, X., 2014. Recent research and developments on yarn hairiness.Text. Res. J. 85, 211e224. https://doi.org/10.1177/0040517514538692.

Hartline, N.L., Bruce, N.J., Karba, S.N., Ruff, E.O., Sonarand, S.U., Holden, P.A., 2016.Microfibre masses recovered from conventional machine washing of new oraged garments. Environ. Sci. Technol. 50, 11532e11538. https://doi.org/10.1021/acs.est.6b03045.

Hussain, T., Ahmed, S., Qayum, A., 2008. Effect of different softeners and sanforisingtreatment on pilling performance of polyester/viscose blended fabrics. ColorTechnol. 124, 375e378. https://doi.org/10.1111/j.1478-4408.2008.00166.x.

Jemec, A., Horvat, P., Kunej, U., Bele, M., Kr�zan, A., 2016. Uptake and effects ofmicroplastic textile fibers on freshwater crustacean Daphnia magna. Environ.Pollut. 219, 201e209. https://doi.org/10.1016/j.envpol.2016.10.037.

Kothari, V.K., Swani, N.M., Gangal, M., 1991. Frictional properties of woven fabrics.Indian J. Fibre Text. 16, 251e256.

Levinson, M.I., 1999. Rinse-added fabric softener technology at the close of thetwentieth century. J. Surfact Deterg. 2, 223e235. https://doi.org/10.1007/s11743-999-0077-4.

Magnusson, K., Wahlberg, C., 2014. Screening of Microplastic Particles in and Down-Stream of a Wastewater Treatment Plant. Technical Report published for IVLSwedish Environmental Research Institute. Swedish Environmental Research

F. De Falco et al. / Environmental Pollution 236 (2018) 916e925 925

Institute, Stockholm, Sweden.Mathalon, A., Hill, P., 2014. Microplastic fibers in the intertidal ecosystem sur-

rounding Halifax Harbor, Nova Scotia. Mar. Pollu Bull. 81, 69e79. https://doi.org/10.1016/j.marpolbul.2014.02.018.

Mintenig, S.M., Int-Veen, I., L€oder, M.G., Primpke, S., Gerdts, G., 2017. Identificationof microplastic in effluents of waste water treatment plants using focal planearray-based micro-Fourier-transform infrared imaging. Water Res. 108,365e372. https://doi.org/10.1016/j.watres.2016.11.015.

Morris, M.A., Prato, H.H., 1976. Fabric Damage during Laundering, vol. 1. Californiaagriculture.

Nair, A.U., Patwardhan, B.A., Nachane, R.P., 2013. Studies on friction in cotton tex-tiles: Part IIda study on the relationship between physical properties andfrictional characteristics of chemically treated cotton fabrics. Indian J. Fibre Text.38, 366e374.

Napper, E.I., Bakir, A., Rowland, S.J., Thompson, R.C., 2015. Characterisation, quantityand sorptive properties of microplastics extracted from cosmetics. Mar. PolluBull. 99, 178e185. https://doi.org/10.1016/j.marpolbul.2015.07.029.

Napper, I.E., Thompson, R.C., 2016. Release of synthetic microplastic plastic fibresfrom domestic washing machines: effects of fabric type and washing condi-tions. Mar. Pollu Bull. 112, 39e45. https://doi.org/10.1016/j.marpolbul.2016.09.025.

Pirc, U., Vidmar, M., Mozer, A., Kr�zan, A., 2016. Emissions of microplastic fibers frommicrofibre fleece during domestic washing. Environ. Sci. Pollut. Res. 23,22206e222211. https://doi.org/10.1007/s11356-016-7703-0.

Rochman, C.M., Hoh, E., Hentschel, B.T., Kaye, S., 2012. Long-term field measure-ment of sorption of organic contaminants to five types of plastic pellets: im-plications for plastic marine debris. Environ. Sci. Technol. 47, 1646e1654.https://doi.org/10.1021/es303700s.

Rochman, C.M., Hoh, E., Kurobe, T., Teh, S.J., 2013. Ingested plastic transfers haz-ardous chemicals to fish and induces hepatic stress. Sci. Rep. 3, 1e7. https://doi.org/10.1038/srep03263.

Rochman, C.M., Tahir, A., Williams, S.L., Baxa, D.V., Lam, R., Miller, J.T., Teh, F.C.,Werorilangi, S., Teh, S.J., 2015. Anthropogenic debris in seafood: plastic debrisand fibres from textiles in fish and bivalves sold for human consumption. Sci.Rep. 5, 1e10. https://doi.org/10.1038/srep14340.

Schreder, E.D., La Guardia, M.J., 2014. Flame retardant transfers from U.S. House-holds (dust and laundry wastewater) to the aquatic environment. Environ. Sci.Technol. 48, 11575e11583. https://doi.org/10.1021/es502227h.

Shaw, D.G., Day, R.H., 1994. Colour- and form-dependent loss of plastic micro-debrisfrom the North Pacific Ocean. Mar. Pollu Bull. 28, 39e43. https://doi.org/10.1016/0025-326X(94)90184-8.

Smith, B., Block, I., 1982. Textiles in Perspective. Prentice Hall, New Jersey, U.S.A.Smulders, E., R€ahse, W., von Rybinski, W., Steber, J., Sung, E., Wiebel, F., 2001.

Detergent ingredients. In: Smulders, E. (Ed.), Laundry Detergents. Wiley-VCHVerlag GmbH & Co. KGaA, Weinheim, FRG. https://doi.org/10.1002/

3527600450.ch3.Song, Y.Y., Sang, H.H., Jang, M., Kang, J.H., Kwon, O.Y., Han, G.M., Shim, W.J., 2014.

Large accumulation of micro-sized synthetic polymer particles in the sea sur-face microlayer. Environ. Sci. Technol. 48, 9014e9021. https://doi.org/10.1021/es501757s.

Sudhakar, M., Trishul, A., Doble, M., Kumar, K.S., Jahan, S.S., Inbakandan, D.,Viduthalai, R.R., Umadevi, V.R., Murthy, P.S., Venkatesan, R., 2007a. Biofoulingand biodegradation of polyolefins in ocean waters. Polym. Degrad. Stabil. 92,1743e1752. https://doi.org/10.1016/j.polymdegradstab.2007.03.029.

Sudhakar, M., Priyadarshini, C., Doble, M., Murthy, P.S., Venkatesan, R., 2007b.Marine bacteria mediated degradation of nylon 66 and 6. Int. Biodeterior.Biodegrad. 60, 144e151. https://doi.org/10.1016/j.ibiod.2007.02.002.

Sutton, R., Mason, S.A., Stanek, S.K., Willis-Norton, E., Wren, I.F., Box, C., 2016.Microplastic contamination in San Francisco Bay, California. USA. Mar. Pollut.Bull. 109, 230e235. https://doi.org/10.1016/j.marpolbul.2016.05.077.

Talvitie, J., Heinonen, M., P€a€akk€onen, J., Vahtera, E., Mikola, A., Set€al€a, O., Vahala, R.,2015. Do wastewater treatment plants act as a potential point source ofmicroplastics? Preliminary study in the coastal Gulf of Finland, Baltic Sea. WaterSci. Technol. 72, 1495e1504. https://doi.org/10.2166/wst.2015.360.

Thompson, R.C., Olsen, Y., Mitchell, R.P., Davis, A., Rowland, S.J., John, A.W.G.,McGonigle, D., Russell, A.E., 2004. Lost at sea: where is all the plastic? Science304, 838. https://doi.org/10.1126/science.1094559.

Thompson, R., Moore, C., Andrady, A., Gregory, M., Takada, H., Weisberg, S., 2005.New directions in plastic debris. Science 310, 1117e1119. https://doi.org/10.1126/science.310.5751.1117b.

UNEP, 2006. Annual Report. Publishing Agency: United Nations Environment Pro-gramme (UNEP). UNEP/Earthprint, 2007.

Wright, S.L., Thompson, R.C., Galloway, T.S., 2013. The physical impacts of micro-plastics on marine organisms: a review. Environ. Pollut. 178, 483e492. https://doi.org/10.1016/j.envpol.2013.02.031.

Yang, D.Q., Shi, H.H., Li, L., Li, J.N., Jabeen, K., Kolandhasamy, P., 2015. Microplasticpollution in table salts from China. Environ. Sci. Technol. 49, 13622e13627.https://doi.org/10.1021/acs.est.5b03163.

Yangxin, Y.U., Jin, Z., Bayly, A.E., 2008. Development of surfactants and builders indetergent formulations. Chin. J. Chem. Eng. 16, 517e527. https://doi.org/10.1016/S1004-9541(08)60115-9.

Ziajahromi, S., Neale, P.A., Rintoul, L., Leusch, F.D.L., 2017. Wastewater treatmentplants as a pathway for microplastics: development of a new approach tosample wastewater-based microplastics. Water Res. 112, 93e99 (Wastewatertreatment plants as a pathway for microplastics: Development of a newapproach to sample wastewater-based microplastics).

Zubris, K.A.V., Richards, B.K., 2005. Synthetic fibers as an indicator of land appli-cation of sludge. Environ. Pollut. 138, 201e211. https://doi.org/10.1016/j.envpol.2005.04.013.