《环境分子生物学》课程教学课件（讲稿）案例3-微塑料 Microplastics in freshwater systems：A review on occurrence, environmental effects, and methods for microplastics detection

Water Research 137(2018)362374Contents listsavailableatScienceDirectKESTERCHWater Researchjournalhomepage:www.elsevier.com/locate/watresELSEVIERReviewMicroplastics in freshwater systems: A review on occurrence真environmentaleffects,andmethodsformicroplasticsdetectionChodhiorJingyi Li, Huihui Liu, J. Paul ChenDepartmentofCivilandEvironmentalEngineering,NationalUniversityofSingapore0KentRidgeCrescent,Singapore,119260,SingaporARTICLEEINFOABSTRACTArticle history:The continuous increase in synthetic plastic production and poor management in plastic waste have ledReceived 6 November 2017to a tremendous increase in the dumping into our aqueous environment. Consequently.microplasticsAccepted22December2017commonly defined as sizes less than 5 mm are produced and stay in both seawater and freshwaterAvailable online 28 December 2017environment.Thepresence of microplastics as a new type of emerging contaminant has become a greatissue of concerns from public and government authorities. The sources of microplastics to freshwaterKeywords:systems are many with the largest portion from wastewater treatment plants. The abundance ofAnalytical methodsmicroplastics varies with the location, from above 1 million pieces per cubic meter to less than 1 piece inFreshwaters100 cubic meters.MicroplasticsMicroplastics can cause several harmful physical effects on humans and living organisms through suchOccurrencemechanisms as entanglement and ingestion. The microplastics can act as carriers of various toxins suchas additives from industrial production processes and persistent contaminants by the sorption in waters.Those toxins may causegreat healthproblems to humans.Afew studies on thefishes demonstrated thatthe microplastics and the associated toxins are bio-accumulated and cause such problems as intestinaldamage and change in metabolic profiles.In studies of microplastics, fresh water is first sampled by the nets with typical mesh size of 330 μm forcollection of microplastics. After the volume reducing process, the samples will then go through thepurification process including density separation by such inorganic salts as sodium chloride and diges-tion process by oxidizing agents or enzymes. The sequence of these two processes (namely purificationand digestion) is dependent on the sample type. The purified samples can be studied by severalanalytical methods. The commonly used methods for the qualification studies are FTIR spectroscopy.Raman spectroscopy,pyrolysis-Gc/MS, and liquid chromatography.A tagging method can be used in thequantification study. Our literature study finds that there is still no universal accepted quantification andqualification tools of microplastics in fresh waters.More work isanticipated so as to obtain accurateinformation on microplastics in freshwater, which can then be used for the better assessment of theenvironmental risk.@ 2017 Elsevier Ltd. All rights reserved.Contents3631.Introduction3642.Microplastics in freshwater systems..3642.1.Occurrence....3652.2.Environmental impacts3.Microplastics sampling methods ..3673674.Sample extraction and purification3685.Microplastics identification and quantification.+.6.371Challenges and recommendations·Corresponding author.E-mail addresses: paulchen@nus.edu.sg, jpaulchen@gatech.edu (J. Paul Chen).https://doi.org/10.1016/j.watres.2017.12.0560043-1354/0 2017 Elsevier Ltd, All rights reserved

Review Microplastics in freshwater systems: A review on occurrence, environmental effects, and methods for microplastics detection Jingyi Li, Huihui Liu, J. Paul Chen* Department of Civil and Environmental Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore, 119260, Singapore article info Article history: Received 6 November 2017 Accepted 22 December 2017 Available online 28 December 2017 Keywords: Analytical methods Freshwaters Microplastics Occurrence abstract The continuous increase in synthetic plastic production and poor management in plastic waste have led to a tremendous increase in the dumping into our aqueous environment. Consequently, microplastics commonly defined as sizes less than 5 mm are produced and stay in both seawater and freshwater environment. The presence of microplastics as a new type of emerging contaminant has become a great issue of concerns from public and government authorities. The sources of microplastics to freshwater systems are many with the largest portion from wastewater treatment plants. The abundance of microplastics varies with the location, from above 1 million pieces per cubic meter to less than 1 piece in 100 cubic meters. Microplastics can cause several harmful physical effects on humans and living organisms through such mechanisms as entanglement and ingestion. The microplastics can act as carriers of various toxins such as additives from industrial production processes and persistent contaminants by the sorption in waters. Those toxins may cause great health problems to humans. A few studies on the fishes demonstrated that the microplastics and the associated toxins are bio-accumulated and cause such problems as intestinal damage and change in metabolic profiles. In studies of microplastics, fresh water is first sampled by the nets with typical mesh size of 330 mm for collection of microplastics. After the volume reducing process, the samples will then go through the purification process including density separation by such inorganic salts as sodium chloride and diges￾tion process by oxidizing agents or enzymes. The sequence of these two processes (namely purification and digestion) is dependent on the sample type. The purified samples can be studied by several analytical methods. The commonly used methods for the qualification studies are FTIR spectroscopy, Raman spectroscopy, pyrolysis-GC/MS, and liquid chromatography. A tagging method can be used in the quantification study. Our literature study finds that there is still no universal accepted quantification and qualification tools of microplastics in fresh waters. More work is anticipated so as to obtain accurate information on microplastics in freshwater, which can then be used for the better assessment of the environmental risk. © 2017 Elsevier Ltd. All rights reserved. Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 2. Microplastics in freshwater systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 2.1. Occurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 2.2. Environmental impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 3. Microplastics sampling methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 4. Sample extraction and purification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 5. Microplastics identification and quantification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 6. Challenges and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 * Corresponding author. E-mail addresses: paulchen@nus.edu.sg, jpaulchen@gatech.edu (J. Paul Chen). Contents lists available at ScienceDirect Water Research journal homepage: www.elsevier.com/locate/watres https://doi.org/10.1016/j.watres.2017.12.056 0043-1354/© 2017 Elsevier Ltd. All rights reserved. Water Research 137 (2018) 362e374

363J.Lietal. /WaterResearch137(2018)362-374371Conclusions7.Acknowledgements.371References.3711.Introductionhorizontalperspective,microplastics were reportedly found intropical areas (Ng and Obbard,2006;Norand Obbard,2014);theyPlastic products have such outstanding features as light-weight,were even seen in thepolar waters of Antarctica and Arctic(Barnesbeing durable and versatile, and production with low costet al.,2010: Bergmann et al.,2015).When one looks at vertical(Hammer et al.,2012; Ivleva et al.,2017).However,plasticdebrisdistribution,microplasticsexistinbenthiczoneofwater bodies,hasraisedglobalconcernsoveritswidedistributionandassociatedwatercolumns,surfacewatersandbeaches.Somereportshaveenvironmental consequences.The annual global production ofshown the concentrations in surface water vary from 10-5to105plastic product in 2016 alone was around 322 million tonnespieces/m3(LiebezeitandDubaish,2012:Desforges etal.,2014:Friasetal.,2014:Limaetal,2014;Auta etal.,2017)and 40to400pieces/(Europe,2016).Anestimationofupto10%ofplasticfragmentswould end up in marine environment as per suggested by Cole et al.Lin sediments (Zurcher,2009; Browne etal.,2011:Antunes et al.2013:Frias et al.,2014;Norand Obbard,2014).Mostrecently.(2o11),duetoextensiveusageandincreasingproductioninplasticproduct, and poor management (Rochman, 2015) While dimin-China Central Television website (cctv.com) reported a group ofishing aesthetic value of water environment, plastic debris is likelyChinese scientists have discovered the presence of microplastics into pose threats to public health and cause biodiversity lossAntarcticwaters(China.org.cn,2018).Furthermore,the distribution(Thompsonetal.2009:Gall andThompson,2015).shows clear geographical variations (Fossi et al.,2012;CollignonMicroplastics are widely defined as synthetic polymers with anetal.,2014:deLuciaetal,2014:Desforgesetal,2014)upper size limit of 5mm andwithout specifiedlower limitThefactors affectingthedistribution include suchlarge-scaleforces as currents driven by wind and geostrophic circulation(Thompson etal.,2009).Theycanbe categorized intoprimarymicroplasticsandsecondarymicroplastics.(Law et al.,2010),turbulence and oceanographic effects (BallentThe definition of primary microplastics is the microplastics,et al, 2012; Turra et al, 2014).As key factors, the inherent prop-which are originally manufactured to have a size less than 5 mmerties of microplastics such as density.shape and size of micro-and mainly found in textiles, medicines and such personal careplastics can affecttransportation and distribution patterns (Eerkes-productsasfacialandbodyscrubs(Coleetal.,2011:Browne2015)Medranoetal,2015)The aforementioned factors are more likely to play importantTheseprimary microplastics canbetransportedby rivers,dischargefrom water treatment plants, wind and surface run-off into eitherroles in a large freshwater environment like riverine systems;fresh water and seawater environments (Gall and Thompson,however, they become limited on smaller isolated fresh water2015)systems, where natural factors and long water residence timeSecondarymicroplastics arederived from fragmentation ordominantlyaffectquantity of microplastics(Freeetal.,2014)largeplasticdebrisduetosuchprocessesasphoto-degradation,Hence,microplasticsin open and dynamicfresh waters wouldphysical, chemical and biological interactions (Thompson et al.,eventually end up in marine environment, while microplastics in2009:Galganietal.,2013).Theoriginsofsecondarymicroplasticsisolatedandstaticwatersbodieswouldremainandaccumulateininclude fishing nets, industrial resin pellets, household items andthe waters.other discarded plastic debris (Eerkes-Medrano et al, 2015).Fresh waters may accumulate numerous microplastic particlesNotably.it was found that the majority of microplastics are sec-andfibers:however,lesseffortshavebeenmadetomonitortheondarymicroplastics(Eriksenetal.,2013)andthattheabundancemicroplasticsinfreshwatersthanthoseinseawaters.Suchfreshin waters would increase along with the increase in input of plasticwaters can be the sources (like waste water plants),transferringdebrisfrom different origins, leadingto continuous transformationmedia (like rivers) and sink (like isolated lakes) of microplastics,ofsecondarymicroplastics(Coleetal,2011).Whenmicroplasticswhichmaydifferfromthoseinseawatersbecauselargevariationsare exposed in the environment, there is a higher possibility ofin quantity can be expected (Klein et al.,2018),Meanwhile, thebreak-down of microplastics to nanoplastics that may have higherproperties of microplastics can be quite heterogeneous.Fonenvironmental risks due to the nature of nano-sizes.instance,microplastics in sewage are heavilycontaminated byMicroplastics can originate from both land- and ocean-basedorganic contents and exist as relatively large pieces; on the othersources (Hammer et al.,2o12).The ocean-based sources,due tohand in clean fresh waters are nearly free of organic contents andcommercial fishing, vessels and other activities in marine envi-hardly seen by naked eyes (Orb, 2017). In addition, some lakes orronment,onlycontribute20%of totalplasticdebris inmarinerivers with fresh water are close to the areas with high population.where higher microplastics abundance was detected (Eriksen et al.environment(Andrady.2011)Themicroplasticsfromterrestrialsourcescontributetheremaining80%.Terrestrialsourcesinclude2013).Anothersignificantcharacteristicofmicroplasticsstudiesindifferent origins that mainly are personal care products, air-blastingfreshwater systems is that the sample sizes are small. However,process,improperlydisposedplasticsandleachatesfromlandfilllargesamplingareasarenecessarytoadequatelyreducethelarge(Coleetal,2011),Onceterrestrialmicroplasticsarereleasedintovariationsdueto spatial andtemporal changes(Ryanetal,2009)the natural water systems, most of them would be transported toAs a result, we found that there was an urgent need to review theoceansbyrivers,whiletheremainingwouldresideinfreshwatercurrentresearchworkandmethodologies onmicroplasticsinfreshenvironment, including such isolated water systems as remotewaters in order that appropriate sampling, quantification andmountainlakes (Browneetal.,2010;Freeetal.,2014)identification approaches can be developed for the study in fresh-Microplasticsareofgreatpublicconcernsfortheubiquitouswater samples.presence and persistence in the aquatic environment. The globalThe objective of this review paper is to reveal the currentpresence of microplasticshas beenfound in recentyears.Fromknowledge about microplastics in fresh waters for a better

7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 1. Introduction Plastic products have such outstanding features as light-weight, being durable and versatile, and production with low cost (Hammer et al., 2012; Ivleva et al., 2017). However, plastic debris has raised global concerns over its wide distribution and associated environmental consequences. The annual global production of plastic product in 2016 alone was around 322 million tonnes (Europe, 2016). An estimation of up to 10% of plastic fragments would end up in marine environment as per suggested by Cole et al. (2011), due to extensive usage and increasing production in plastic product, and poor management (Rochman, 2015). While dimin￾ishing aesthetic value of water environment, plastic debris is likely to pose threats to public health and cause biodiversity loss (Thompson et al., 2009; Gall and Thompson, 2015). Microplastics are widely defined as synthetic polymers with an upper size limit of 5 mm and without specified lower limit (Thompson et al., 2009). They can be categorized into primary microplastics and secondary microplastics. The definition of primary microplastics is the microplastics, which are originally manufactured to have a size less than 5 mm and mainly found in textiles, medicines and such personal care products as facial and body scrubs (Cole et al., 2011; Browne, 2015). These primary microplastics can be transported by rivers, discharge from water treatment plants, wind and surface run-off into either fresh water and seawater environments (Gall and Thompson, 2015). Secondary microplastics are derived from fragmentation or large plastic debris due to such processes as photo-degradation, physical, chemical and biological interactions (Thompson et al., 2009; Galgani et al., 2013). The origins of secondary microplastics include fishing nets, industrial resin pellets, household items and other discarded plastic debris (Eerkes-Medrano et al., 2015). Notably, it was found that the majority of microplastics are sec￾ondary microplastics (Eriksen et al., 2013) and that the abundance in waters would increase along with the increase in input of plastic debris from different origins, leading to continuous transformation of secondary microplastics (Cole et al., 2011). When microplastics are exposed in the environment, there is a higher possibility of break-down of microplastics to nanoplastics that may have higher environmental risks due to the nature of nano-sizes. Microplastics can originate from both land- and ocean-based sources (Hammer et al., 2012). The ocean-based sources, due to commercial fishing, vessels and other activities in marine envi￾ronment, only contribute 20% of total plastic debris in marine environment (Andrady, 2011). The microplastics from terrestrial sources contribute the remaining 80%. Terrestrial sources include different origins that mainly are personal care products, air-blasting process, improperly disposed plastics and leachates from landfill (Cole et al., 2011). Once terrestrial microplastics are released into the natural water systems, most of them would be transported to oceans by rivers, while the remaining would reside in fresh water environment, including such isolated water systems as remote mountain lakes (Browne et al., 2010; Free et al., 2014). Microplastics are of great public concerns for the ubiquitous presence and persistence in the aquatic environment. The global presence of microplastics has been found in recent years. From horizontal perspective, microplastics were reportedly found in tropical areas (Ng and Obbard, 2006; Nor and Obbard, 2014); they were even seen in the polar waters of Antarctica and Arctic (Barnes et al., 2010; Bergmann et al., 2015). When one looks at vertical distribution, microplastics exist in benthic zone of water bodies, water columns, surface waters and beaches. Some reports have shown the concentrations in surface water vary from 10
5 to 105 pieces/m3 (Liebezeit and Dubaish, 2012; Desforges et al., 2014; Frias et al., 2014; Lima et al., 2014; Auta et al., 2017) and 40 to 400 pieces/ L in sediments (Zurcher, 2009; Browne et al., 2011; Antunes et al., 2013; Frias et al., 2014; Nor and Obbard, 2014). Most recently, China Central Television website (cctv.com) reported a group of Chinese scientists have discovered the presence of microplastics in Antarctic waters (China.org.cn, 2018). Furthermore, the distribution shows clear geographical variations (Fossi et al., 2012; Collignon et al., 2014; de Lucia et al., 2014; Desforges et al., 2014). The factors affecting the distribution include such large-scale forces as currents driven by wind and geostrophic circulation (Law et al., 2010), turbulence and oceanographic effects (Ballent et al., 2012; Turra et al., 2014). As key factors, the inherent prop￾erties of microplastics such as density, shape and size of micro￾plastics can affect transportation and distribution patterns (Eerkes￾Medrano et al., 2015). The aforementioned factors are more likely to play important roles in a large freshwater environment like riverine systems; however, they become limited on smaller isolated fresh water systems, where natural factors and long water residence time dominantly affect quantity of microplastics (Free et al., 2014). Hence, microplastics in open and dynamic fresh waters would eventually end up in marine environment, while microplastics in isolated and static waters bodies would remain and accumulate in the waters. Fresh waters may accumulate numerous microplastic particles and fibers; however, less efforts have been made to monitor the microplastics in fresh waters than those in seawaters. Such fresh waters can be the sources (like waste water plants), transferring media (like rivers) and sink (like isolated lakes) of microplastics, which may differ from those in seawaters because large variations in quantity can be expected (Klein et al., 2018). Meanwhile, the properties of microplastics can be quite heterogeneous. For instance, microplastics in sewage are heavily contaminated by organic contents and exist as relatively large pieces; on the other hand in clean fresh waters are nearly free of organic contents and hardly seen by naked eyes (Orb, 2017). In addition, some lakes or rivers with fresh water are close to the areas with high population, where higher microplastics abundance was detected (Eriksen et al., 2013). Another significant characteristic of microplastics studies in freshwater systems is that the sample sizes are small. However, large sampling areas are necessary to adequately reduce the large variations due to spatial and temporal changes (Ryan et al., 2009). As a result, we found that there was an urgent need to review the current research work and methodologies on microplastics in fresh waters in order that appropriate sampling, quantification and identification approaches can be developed for the study in fresh￾water samples. The objective of this review paper is to reveal the current knowledge about microplastics in fresh waters for a better J. Li et al. / Water Research 137 (2018) 362e374 363

364J.Lietal. / Water Research137(2018)362374understandingofmicroplasticscontaminationandpotentialrisks.abundanceinfreshwatersystemsvariedgreatlyfromalmostnoneSummaries of sampling methods and comparisons of differenttoseveralmillionpiecespercubicmeter.Thissignificantdifferencequantificationandidentificationapproachesarepresented.Severalresultsfromsuchkeyfactorsassamplinglocations,humanactiv-key challenges are discussed and suggestions are provided forities,inherent natural conditions and sampling approaches(Eerkes-Medrano et al., 2015).further research work.Many terrestrial sources contribute the microplastics. Amongthem, wastewater treatment is one of the dominant sources of2. Microplastics in freshwater systemsmicroplastics（MagnussonandNoren,2014;Talvitie,2014;EstahbanatiandFahrenfeld,2016:Murphyetal.,2016;2.1.OccurrenceDyachenkoetal.2017:Mintenigetal.,2017).Table2lists thekeyresults from the microplastics studies on several wastewaterMostofeffortsontheresearchofmicroplasticshavebeenplacedtreatmentplants(WWTPs)ThoughWWTPscanremoveupto95%onseawaterenvironment.Lessthan4%ofmicroplastics-relatedofmicroplastics(Talvitie,2014;Talvitieetal.,2017)andtertiarystudiesarereportedlyassociatedwithfreshwaters(Lambertandtreatmentcanhavea90%removalefficiencyoffineparticlesofsizeWagner,2018)Thelimitedinformation,however,revealedthatlarger than 10μm (Wardrop et al,2016),there is substantialtheabundance ofmicroplastics in freshwatersis comparableto thatamount of microplastics being discharged into natural waters viaof marine environment (Peng et al, 2017) and the distribution isWWTPs.highly heterogeneous (Klein et al.,2018).Table 1 summarizes someRochman et al. (2015) used the published data by Magnussonof therelevant studies on themicroplasticsabundance in fresh-andNoren(2014)andMartinandEizhvertina(2014)toestimatewater matrices. The mean/averaged values of microplasticsTable1Studies on microplastics contamination in natural fresh water systems.CollectionDepthSeparationMeanMaximumLocationCollectionPurificationIdentificationReferenceCut-off sizeabundancesubstrateabundance(μm)Austrian Danube,Stationary conical 5000.5m-Visualization-0.317p/m3141.7 p/m2(Lechner et al.Austriadriftnets2014)3008.93×10°p/3.9×10°p/(Manietal.Rhine river-Enzyme + H202 FTIRSieves Manta netkm?km22015)1×10° p/m2 1.87 ×105-FTIR0.7 μm glassDutch river delta andBulk water 2 L--(Leslie et al.Amsterdam canalsfiltersp/ma2017)800.130 p/m2106 p/m3Great ParisPlankton net-Visualization1.6 μm filter(Dris etal,2015)0.35m0.35 p/m2Great ParisManta Trawl33000.3m Visualization1.6 μm filter0.45 p/m3(Dris et al, 2015)300Lake Geneva Manta Trawl-.Visualization4.81×10°p/ (Alencastro,一km?2012)-1.36 ×107Three Gorges Dam,Trawl112-FTIR1.6mm8.47 ×10°p/(Zhang et al.stainless sieve km?p/km2China2015)1m4.70×10p/1.26×104Three Gorges Dam,Teflon pumpand 4830%H202Visualization + Raman 0.45 μm glass(Di and Wang.map/ma2017)Chinastainless steelmicrofiberfiltersieve1m4.14 ×10°p/ 1.02×104Yangtze EstuaryTeflon pump and 3230% H202Visualization1.2 μm(Zhao et.al, stainless steelcellulosem3p/m32014)sievenitrate fltersTeflon pump and 500.2m30% H202FTIR8.93×10°p/ (Wang etalLakes, Wuhan, China0.45 μm glassm3 stainless steelmicrofiber2017)sievefilter3330.3 m30% H202FTIR + SEM-EDS100μm6.8 × 10° p/ (Su et al., 2016)Taihu lake, China Plankton netkm2polycarbonatefilter-Bulk water 5 L-30% H2025μm2.58 ×104(Su et al., 2016)Taihu lake, ChinaFTIR + SEM-EDS-polycarbonatep/m3filter-2.03 ×10° p/ 4.44 ×10433330% H202(Free et al.Lake Hovsgol, Mongolia Manta trawlVisualizationTyler sieveskm?P/km22014)333-Lake Winnipeg. Canada Manta trawl30% H202SEM-EDS250 μm sieve1.93×10°p/7.48 ×10°p/ (Anderson et al.km2km22017)-Los Angeles river, SanHand net, Manta800,500, 333 VisualizationTyler sieves1.29×10°p/(Moore etal.4Gabriel river, Coyote trawlm32011)Creek3330.24.2 p/m332 p/m329 Great Lakes Neuston net30% H202+FebVisualization125 μm sieve(Baidwin etal)tributaries, USA0.35m2016)3332M HCISEM-EDS4.30 ×10° p/ 4.66 ×105(Eriksen et al.,Laurentian Great Lakes, Manta trawl-Tyler sievesUSAkm2p/km?2013)153-30% H202+FebRaritan River, USAPlankton netVisualizationSieves一(Estahbanati andFahrenfeld,2016)3.1×10~4-0.19 p/m2Goiana Estuary, BrazilConical plankton 300-Visualization45 μm mesh(Lima et al.)-2.6 × 103net2014)p/m3Data werestandardizedfor consistencyb Wet peroxide oxidation

understanding of microplastics contamination and potential risks. Summaries of sampling methods and comparisons of different quantification and identification approaches are presented. Several key challenges are discussed and suggestions are provided for further research work. 2. Microplastics in freshwater systems 2.1. Occurrence Most of efforts on the research of microplastics have been placed on seawater environment. Less than 4% of microplastics-related studies are reportedly associated with freshwaters (Lambert and Wagner, 2018). The limited information, however, revealed that the abundance of microplastics in freshwaters is comparable to that of marine environment (Peng et al., 2017) and the distribution is highly heterogeneous (Klein et al., 2018). Table 1 summarizes some of the relevant studies on the microplastics abundance in fresh￾water matrices. The mean/averaged values of microplastics abundance in fresh water systems varied greatly from almost none to several million pieces per cubic meter. This significant difference results from such key factors as sampling locations, human activ￾ities, inherent natural conditions and sampling approaches (Eerkes-Medrano et al., 2015). Many terrestrial sources contribute the microplastics. Among them, wastewater treatment is one of the dominant sources of microplastics (Magnusson and Noren, 2014; Talvitie, 2014;
Estahbanati and Fahrenfeld, 2016; Murphy et al., 2016; Dyachenko et al., 2017; Mintenig et al., 2017). Table 2 lists the key results from the microplastics studies on several wastewater treatment plants (WWTPs). Though WWTPs can remove up to 95% of microplastics (Talvitie, 2014; Talvitie et al., 2017) and tertiary treatment can have a 90% removal efficiency of fine particles of size larger than 10 mm (Wardrop et al., 2016), there is substantial amount of microplastics being discharged into natural waters via WWTPs. Rochman et al. (2015) used the published data by Magnusson and Noren (2014)
and Martin and Eizhvertina (2014) to estimate Table 1 Studies on microplastics contamination in natural fresh water systems. Location Collection Collection Cut-off size (mm) Depth Purification Identification Separation substrate Mean abundancea Maximum abundancea Reference Austrian Danube, Austria Stationary conical driftnets 500 0.5 m e Visualization e 0.317p/m3 141.7 p/m3 (Lechner et al., 2014) Rhine river Manta net 300 e Enzyme þ H2O2 FTIR Sieves 8.93 105 p/ km2 3.9 106 p/ km2 (Mani et al., 2015) Dutch river delta and Amsterdam canals Bulk water 2 L e ee FTIR 0.7 mm glass filters 1 105 p/m3 1.87 105 p/m3 (Leslie et al., 2017) Great Paris Plankton net 80 0.1 e0.35 m e Visualization 1.6 mm filter 30 p/m3 106 p/m3 (Dris et al., 2015) Great Paris Manta Trawl 330 0e0.3 m e Visualization 1.6 mm filter 0.35 p/m3 0.45 p/m3 (Dris et al., 2015) Lake Geneva Manta Trawl 300 e e Visualization e 4.81 104 p/ km2 e (Alencastro, 2012) Three Gorges Dam, China Trawl 112 e e FTIR 1.6 mm stainless sieve 8.47 106 p/ km2 1.36 107 p/km2 (Zhang et al., 2015) Three Gorges Dam, China Teflon pump and stainless steel sieve 48 1 m 30% H2O2 Visualization þ Raman 0.45 mm glass microfiber filter 4.70 103 p/ m3 1.26 104 p/m3 (Di and Wang, 2017) Yangtze Estuary Teflon pump and stainless steel sieve 32 1 m 30% H2O2 Visualization 1.2 mm cellulose nitrate filters 4.14 103 p/ m3 1.02 104 p/m3 (Zhao et al., 2014) Lakes, Wuhan, China Teflon pump and stainless steel sieve 50 0.2 m 30% H2O2 FTIR 0.45 mm glass microfiber filter e 8.93 103 p/ m3 (Wang et al., 2017) Taihu lake, China Plankton net 333 0.3 m 30% H2O2 FTIR þ SEM-EDS 100 mm polycarbonate filter e 6.8 106 p/ km2 (Su et al., 2016) Taihu lake, China Bulk water 5 L e e 30% H2O2 FTIR þ SEM-EDS 5 mm polycarbonate filter e 2.58 104 p/m3 (Su et al., 2016) Lake Hovsgol, Mongolia Manta trawl 333 e 30% H2O2 Visualization Tyler sieves 2.03 104 p/ km2 4.44 104 p/km2 (Free et al., 2014) Lake Winnipeg, Canada Manta trawl 333 e 30% H2O2 SEM-EDS 250 mm sieve 1.93 105 p/ km2 7.48 105 p/ km2 (Anderson et al., 2017) Los Angeles river, San Gabriel river, Coyote Creek Hand net, Manta trawl 800, 500, 333 e e Visualization Tyler sieves e 1.29 104 p/ m3 (Moore et al., 2011) 29 Great Lakes tributaries, USA Neuston net 333 0.2 e0.35 m 30% H2O2þFeb Visualization 125 mm sieve 4.2 p/m3 32 p/m3 (Baldwin et al., 2016) Laurentian Great Lakes, USA Manta trawl 333 e 2 M HCl SEM-EDS Tyler sieves 4.30 104 p/ km2 4.66 105 p/km2 (Eriksen et al., 2013) Raritan River, USA Plankton net 153 e 30% H2O2þFeb Visualization Sieves e e (Estahbanati and Fahrenfeld, 2016) Goiana Estuary, Brazil Conical plankton net 300 e e Visualization 45 mm mesh 3.1 10
4 -2.6 10
3 p/m3 0.19 p/m3 (Lima et al., 2014) a Data were standardized for consistency. b Wet peroxide oxidation 364 J. Li et al. / Water Research 137 (2018) 362e374

365J. Li et al. / Water Research 137 (2018) 362374Table 2Microplastics detected in waste water treatment plants.LocationCollectionCollection Cut-PurificationIdentification Mean abundanceMaximumReferenceabundanceoff size (μm)Efluents of 17 WWTPs, USAExtraction pump + Tyler 355, 12530 %H202Visualization 50 p/m3195 p/m3(Mason et al,2016)sievesEast Bay Municipal Utility355, 12530%H202FTIRSieves-169p/m3(Dyachenko et al.,District's wWTP, USA2017)-FTIR-Influent:(Talvitie et al.Viikinmaki WWTP, FinlandPumping through filters300, 100, 209 ×10° p/m32017)or metallic beakerEffluent:3.5 × 103 p/m365-FTIRInfluent: 1.57 × 10° p/m2WWTP, River Clyde, GlasgowSteel buckets(Murphy et al.,2016)Effluent:250 p/m2Effluents of 12 WWTPs inA custom made pumping 10Enzymes + H202 FTIR9 ×10° p/m2(Mintenig et al.)2017)Lower Saxony. GermanydeviceVisualization Influent: 1.6× 10° p/m3b300, 100, 20-WWTP, St. PetersburgA specific filter device(Talvitie, 2014)Effluent:7×103p/m3b300-FTIRLangeviksverket, Lysekil.Ruttner samplerInfluent: 1.5 × 10° p/m*:(Magnusson andSwedenEfluent: 8.25 p/m3Noren,2014)-7 wWTPs, Netherland-FTIRGlass jars (2 L)Influent: 7.3×10*p/m3Influent:(Leslie et al,Efluent: 5.2 ×10° p/m5.66 × 10°p/2017)maEffluent:9.1 × 10° p/m2--Ljubljana, SloveniaEffluent: 13.9 mg/m’ PE(Kaicikova et al.,2017)(estimated)Data was standardizedforconsistencyb Data represents the number of synthetic particles.the amount of microbeads discharged into the waterways in theThough to what extent of physical effects would affect organismsUSA They came to a conservative conclusion that 8 billionremain uncertain,entanglement effect that is often associated withmicrobeads (pieces) released from the municipal WWTPs per day.comparativelylargeanimalsisvisiblewhenwecompareitwithMason et al. (2016) studied the effluent samples of 17 WWTPs andingestion. Entanglement could cause severe impacts on aquaticpredicted that the average discharge of microbeads from USspecies; they can even be fatal by the means of drowning, suffo-municipal WWTPs was 13 billion pieces per day, similar to thatcating, strangulating or starving (Allsopp et al., 2006). The vulner-from Rochman et al. (2015). Given the fact that high volume ofable species includesea turtles,mammals,seabirds,andtreated and untreated wastewater is released globally and only 60%crustaceans(Gilardietal,2010).Whentheseanimalsdrown inof municipal wastewater is treated (Mateo-Sagastaetal,2015).aghost nets, they may suffer suffocation and starvation; when theirhuge amount of microplastics would enter the environment via thepredators appear, they are bound to die (Derraik, 2002).However, there is no report on entanglement incidence indischarges from WWTPs. In addition, other sources such as surfacerun-off, atmospheric fallout (Driset al,2016)and direct wastefreshwaterbodies.Nevertheless,theoccurrenceofentanglementindisposal contributethe increasein the microplasticsflow intomarineorganismshasprovidedaclearindicationtothesituationinaqueous environment.freshwaterenvironment.In September 2017,Orb Media,a nonprofit journalism organi-Ingestion does not directly impose fatal effects on organisms,zation, published a report that claims the presence of microplasticseven thoughit has well been observed.Thechroniceffects howeverin drinking water. This cross-border research tested 159 drinkingbecome akey issue (Wright etal.,2013a).Ryan(1987)conducted awater samples from five continents found out that 83% of themsurvey on the potential effects of plastic ingestion on domesticwere contaminated with tiny plastic debris (Orb, 2017).Sincechicks, in order to simulate the biological behavior of ingestedmicroplastics can directly enter human bodies, it may be a long-plastics on seabirds. It was found that there was a positive rela-term exposure if people drink microplastics-containing water.tionship between reduced food consumption andfeed of plastics.AThese findings would trigger public concerns over the safety ofresearch team suggested that a negative correlation between thedrinking water and food.fitness of seabirds and theingested plasticdebris (Spear et al.1995).The negative correlation was also observed on the fish(Lusher et al.,2014).The pathways of microplastics entering the2.2.Environmentalimpactstract include direct and indirect ingestion. Fish mainly ingestsmicroplastics via predation activities; the accumulation effects canThe concerns over microplastics are about the potential harmsbe observed in higher trophic levels, such as seabirds, seals and seathat can impose on organisms and humans. The environmentallions(McMahonetal.,1999;Eriksson and Burton,2003:Romeoimpactscanbecataloguedtophysical,chemical andbiologicalet al.,2015).Theconcentration factorofmicroplasticsfrom sur-impacts,as describedbelow.Thefindingsontheimpactsaremainlyrounding waters to seals was reported to be as high as 160 timesbased on marine environment, but can be used for fresh water(ErikssonandBurton,2003;Wrightetal.,2013b;Eerkes-Medranoenvironment.et al.,2015).Physical impacts mainly include entanglement and ingestionIt was found that entanglementhappenedmorefrequently thanbased on the work on the microplastics in sweater. The studyingestion.55% of marine organisms incidences are associated withconductedbyLaist(1997)hasshownover200marinespeciesentanglement; ingestion contributes to 31% of incidences (Gall andsuffered from the entanglement and ingestion of plastic debris

the amount of microbeads discharged into the waterways in the USA. They came to a conservative conclusion that 8 billion microbeads (pieces) released from the municipal WWTPs per day. Mason et al. (2016) studied the effluent samples of 17 WWTPs and predicted that the average discharge of microbeads from US municipal WWTPs was 13 billion pieces per day, similar to that from Rochman et al. (2015). Given the fact that high volume of treated and untreated wastewater is released globally and only 60% of municipal wastewater is treated (Mateo-Sagasta et al., 2015), a huge amount of microplastics would enter the environment via the discharges from WWTPs. In addition, other sources such as surface run-off, atmospheric fallout (Dris et al., 2016) and direct waste disposal contribute the increase in the microplastics flow into aqueous environment. In September 2017, Orb Media, a nonprofit journalism organi￾zation, published a report that claims the presence of microplastics in drinking water. This cross-border research tested 159 drinking water samples from five continents found out that 83% of them were contaminated with tiny plastic debris (Orb, 2017). Since microplastics can directly enter human bodies, it may be a long￾term exposure if people drink microplastics-containing water. These findings would trigger public concerns over the safety of drinking water and food. 2.2. Environmental impacts The concerns over microplastics are about the potential harms that can impose on organisms and humans. The environmental impacts can be catalogued to physical, chemical and biological impacts, as described below. The findings on the impacts are mainly based on marine environment, but can be used for fresh water environment. Physical impacts mainly include entanglement and ingestion based on the work on the microplastics in sweater. The study conducted by Laist (1997) has shown over 200 marine species suffered from the entanglement and ingestion of plastic debris. Though to what extent of physical effects would affect organisms remain uncertain, entanglement effect that is often associated with comparatively large animals is visible when we compare it with ingestion. Entanglement could cause severe impacts on aquatic species; they can even be fatal by the means of drowning, suffo￾cating, strangulating or starving (Allsopp et al., 2006). The vulner￾able species include sea turtles, mammals, seabirds, and crustaceans (Gilardi et al., 2010). When these animals drown in ghost nets, they may suffer suffocation and starvation; when their predators appear, they are bound to die (Derraik, 2002). However, there is no report on entanglement incidence in freshwater bodies. Nevertheless, the occurrence of entanglement in marine organisms has provided a clear indication to the situation in freshwater environment. Ingestion does not directly impose fatal effects on organisms, even though it has well been observed. The chronic effects however become a key issue (Wright et al., 2013a). Ryan (1987) conducted a survey on the potential effects of plastic ingestion on domestic chicks, in order to simulate the biological behavior of ingested plastics on seabirds. It was found that there was a positive rela￾tionship between reduced food consumption and feed of plastics. A research team suggested that a negative correlation between the fitness of seabirds and the ingested plastic debris (Spear et al., 1995). The negative correlation was also observed on the fish (Lusher et al., 2014). The pathways of microplastics entering the tract include direct and indirect ingestion. Fish mainly ingests microplastics via predation activities; the accumulation effects can be observed in higher trophic levels, such as seabirds, seals and sea lions (McMahon et al., 1999; Eriksson and Burton, 2003; Romeo et al., 2015). The concentration factor of microplastics from sur￾rounding waters to seals was reported to be as high as 160 times (Eriksson and Burton, 2003; Wright et al., 2013b; Eerkes-Medrano et al., 2015). It was found that entanglement happened more frequently than ingestion. 55% of marine organisms incidences are associated with entanglement; ingestion contributes to 31% of incidences (Gall and Table 2 Microplastics detected in waste water treatment plants. Location Collection Collection Cut￾off size (mm) Purification Identification Mean abundancea Maximum abundancea Reference Effluents of 17 WWTPs, USA Extraction pump þ Tyler sieves 355, 125 30 %H2O2 Visualization 50 p/m3 195 p/m3 (Mason et al., 2016) East Bay Municipal Utility District's WWTP, USA Sieves 355, 125 30 %H2O2 FTIR e 169 p/m3 (Dyachenko et al., 2017) Viikinm€ aki WWTP, Finland Pumping through filters or metallic beaker 300, 100, 20 e FTIR e Influent: 9 105 p/m3 Effluent: 3.5 103 p/m3 (Talvitie et al., 2017) WWTP, River Clyde, Glasgow Steel buckets 65 e FTIR Influent: 1.57 104 p/m3 Effluent: 250 p/m3 e (Murphy et al., 2016) Effluents of 12 WWTPs in Lower Saxony, Germany A custom made pumping device 10 Enzymes þ H2O2 FTIR e 9 103 p/m3 (Mintenig et al., 2017) WWTP, St. Petersburg A specific filter device 300, 100, 20 e Visualization Influent: 1.6 105 p/m3 b Effluent:7 103 p/m3 b e (Talvitie, 2014) Långeviksverket, Lysekil, Sweden Ruttner sampler 300 e FTIR Influent: 1.5 104 p/m3 ; Effluent: 8.25 p/m3 e (Magnusson and Noren, 2014
) 7 WWTPs, Netherland Glass jars (2 L) e e FTIR Influent: 7.3 104 p/m3 Effluent: 5.2 104 p/m3 Influent: 5.66 105 p/ m3 Effluent: 9.1 104 p/m3 (Leslie et al., 2017) Ljubljana, Slovenia e ee Effluent: 13.9 mg/m3 PE (estimated) e (Kalcíkov
a et al., 2017) a Data was standardized for consistency. b Data represents the number of synthetic particles. J. Li et al. / Water Research 137 (2018) 362e374 365

366J. Li et al. / Water Research 137 (2018)362-374Thompson, 2015). Other solid matters contribute the rest (14%).microplastics can be released when microplastics are ingested andEntanglement ofmicroplasticsmainlyhappensto compara-stayinsideofhumanbody.wherepH isratherlow,temperatureistively large marine organism. On the other hand, ingestion ofcomparatively high and digestive liquids are present. A few studiesmicroplastics can befound throughout almost all the trophiclevels,confirmedthatthedesorptionrateof sorbedcontaminants in or-such as zooplankton taxa (Cole et al, 2013), marine lugwormganisms was accelerated, substantially faster than that in seawater(Wright et al..2013a).mussel (von Moos et al.,2012),oyster(Teuten etal.,2007:Bakir etal,2014).(Sussarellu et al., 2016).fish (Rochman et al.,2013). sea turtlesHowever, very limited information is available about the real(Bugoni et al, 2001),dolphins (Denuncio et al, 2011), whalessorption behavior for water-borne pollutants in fresh water on the(Walkerand Coe,1989)and seabirds (Derraik,2002)microplastics. The toxicity of microplastics in fresh waters is notThe chemical and biological impacts play key roles. Afterwell understood.ingestion, microplastics cause toxicity effects to humans and livingAfewlab-scalestudieshaveprovidedsomeimplicationsontheorganisms through several pathways and mechanisms.The poly-potential biological hazards from microplastics. A research con-meric compounds used in production of plastics, the additives suchductedbyMaetal.(2016)selectedDaphniamagnaasamodelfreshas coper ions used during production of plastics are toxic. Morewaterorganismtostudythetoxicityofmicroplasticstogetherwithimportantly,various toxins in waters that are initially sorbed ontosorbedphenanthrene.Theyconcludedthatnano-sizedPSexhibitmicroplastics may subsequently be desorbed inside of human andhigh toxicity and physical damage to Daphnia magna and toxicityanimal bodies.was enhanced by sorbed phenanthrene.Microplastics are made of polymeric compounds that can causeA recent study demonstrated that the mixture of antimicrobialflorfenicol and microplastics caused higher inhibition level ofcertain health effects.For example,polystyrene (PS),resistant tobiodegradation can accumulate in the stomach of fish (Carpentercholinesterase activity on freshwater bivalve Corbicula flunineaet al.,1972)and can translocate through blood circulation (Chenthan thatofflorfenicol or microplastics(Guilhermino et al.,2018).Itet al.,2006).For oysters after being experienced a two-monthwas found that microplastics caused certain levels of biologicalexposure to Ps microplastics, the decreases in oocyte number,effects on Corbicula flumineaand and their predator Acipenserdiameter and sperm velocity were reported; reproductive disrup-transmontanus(Rochmanetal..2017)A few studies on larval and adult zebrafish showed thattion for marine filter feeders was expected (Sussarellu et al.,2016).Avarietyofadditivesareaddedduringplasticsproduction,tomicroplasticswerefirstingested,becameaccumulated,andimprove physical properties, such as color, flame resistance, andconsequently caused alternations in locomotion, intestinal damage,hardness.They can be low molecular or polymeric,inorganic orand changeinmetabolicprofiles(Luetal,2016:Chen etal.,2017:Sleight et al., 2017: Lei et al., 2018). With regards to other freshorganic substances. The most common additive is plasticizer that isfor improvement of plasticityor viscosity.For example,polyvinylwater organisms, the toxicity studies on marine organisms maychloride(PvC)musthaveplasticizerslikephthalatesandbisphenolprovide suggestive information.Hence,itisof importancetoA in order that thermal and photo-degradation can be minimizedinvestigate the interaction between microplastics and key com-(Hammeretal.,20i2).Otheradditivesincludecolorantsandflamepounds in lab-scale experiments, which will be helpful in the riskretardants. It is anticipated that these chemicals would accumulateassessmentofmicroplasticsin humanbodies throughbioaccumulation process,some ofwhichThe biological effects include the potential to geographicallyare well known as endocrine disrupting compounds. Some studiestransfer microorganisms(Oberbeckmann et al.,2015)Microor-confirmed that such additives as bisphenol A, polybrominatedganisms can quickly colonize the surface ofmicroplastics and bediphenyl ethers, tetrabromobisphenol A and phthalates are presenttransported with the movement of microplastics, as plastics areinhumans (Talsnessetal.,2009)usuallydurableand persistentthan othermedia(Thiel and Gutow,In addition, some compounds with heavy metals such as chro-2005).While this interaction iscommonlyknown and possiblemium, cadmium and lead are often used in production of colorants,consequences are raised, such as the introduction of pathogens to astabilizers and plasticizers(Ernst et al.,2000;Murphy,2001).Theyclean environment,limited literatures are available to reveal thecan be released from plastic debris into water systems and furtherdiverse biofilms communities, and even less for freshwater envi-enter the food chain to cause bioaccumulation of toxins inronment.(McCormick et al.,2014)conducted high-throughputorganismssequencinganalysis on themicroplasticscollected in an urbanMicroplastics can be a vector for water-borne hydrophobicriver in Chicago, llinois. They found out some attached taxa werepollutants (Teuten et al.,2009;Lee etal.,2014:Bakiret al.,2016:plastic decomposing organisms and pathogens: the findings sug-O'Connoretal.,2016;Ziccardi etal,2016).Examplesincludepol-gest that microplastics can transport bacterial assemblages inychlorinated biphenyls (PCBs)(Velzeboer et al.,2014).andfreshwater ecosystems. Their study also emphasized that patho-genic wastewater-associated organisms could be discharged intodichlorodiphenyltrichloroethane(DDT)(Rios etal.,2010).whichare well known for their high toxicity and persistence in thewaterways by means of the microplastics with the attachment ofenvironment. Due to the large specific surface areas and intrinsicthe organisms. A study on microplastics-associated bacteria inhydrophobicity, the potential of hydrophobic chemical adsorptionYangtze Estuaryalso confirmed thepresence ofpotential pathogensontothesurfaceofmicroplasticshascausedgreatconcernsoveronmicroplastics(liangetal,2018)microplastics (Horton et al., 2017)Another biological effect is the change in the plastic physicalPCBs are well known to be carcinogenic, mutagenic, and/orpropertiesbecauseofbiofilmsthatattachontosurfaceofmicro-2012)DDT can lead to adverseteratogenic (Hammer etplastics.Firstly.thedensity ofmicroplasticscanbeincreased (Car)aneurological effectsandimmunodeficiency(Mansourietal.,2017).etal.,2016),enablinglightmicroplasticstosinkinthewatercol-The partition studies show the large partitioning coefficients forumnandbenthiczones.Furthermore,thebiofilmscanaltertheorganic compounds, in the range of 10410° (Andrady, 2011).A fewsurface nature of microplastics and make the surface less hydro-studies suggested that microplastics could sorb high amounts ofphobic (Lobelle and Cunliffe,2011:Zettler et al,2013).The findingsPCBs form the surrounding seawater in coastal areas of USA andmay provide some insights on studies associated with sorption ofJapan (Hammer et al,2012).Microplastics can sorb lubrication oilspersistent organic pollutants. Whether the attached biofilms canand heavy metals (Angiolillo et al.,2015;Hu et al.,2017).enhance or weakentheinteractionsbetweenmicroplasticsandTheaforementionedpollutantsinitiallysorbedonthesurfaceofwater-borne pollutants remains unknown to us

Thompson, 2015). Other solid matters contribute the rest (14%). Entanglement of microplastics mainly happens to compara￾tively large marine organism. On the other hand, ingestion of microplastics can be found throughout almost all the trophic levels, such as zooplankton taxa (Cole et al., 2013), marine lugworm (Wright et al., 2013a), mussel (von Moos et al., 2012), oyster (Sussarellu et al., 2016), fish (Rochman et al., 2013), sea turtles (Bugoni et al., 2001), dolphins (Denuncio et al., 2011), whales (Walker and Coe, 1989) and seabirds (Derraik, 2002). The chemical and biological impacts play key roles. After ingestion, microplastics cause toxicity effects to humans and living organisms through several pathways and mechanisms. The poly￾meric compounds used in production of plastics, the additives such as coper ions used during production of plastics are toxic. More importantly, various toxins in waters that are initially sorbed onto microplastics may subsequently be desorbed inside of human and animal bodies. Microplastics are made of polymeric compounds that can cause certain health effects. For example, polystyrene (PS), resistant to biodegradation can accumulate in the stomach of fish (Carpenter et al., 1972) and can translocate through blood circulation (Chen et al., 2006). For oysters after being experienced a two-month exposure to PS microplastics, the decreases in oocyte number, diameter and sperm velocity were reported; reproductive disrup￾tion for marine filter feeders was expected (Sussarellu et al., 2016). A variety of additives are added during plastics production, to improve physical properties, such as color, flame resistance, and hardness. They can be low molecular or polymeric, inorganic or organic substances. The most common additive is plasticizer that is for improvement of plasticity or viscosity. For example, polyvinyl chloride (PVC) must have plasticizers like phthalates and bisphenol A in order that thermal and photo-degradation can be minimized (Hammer et al., 2012). Other additives include colorants and flame retardants. It is anticipated that these chemicals would accumulate in human bodies through bioaccumulation process, some of which are well known as endocrine disrupting compounds. Some studies confirmed that such additives as bisphenol A, polybrominated diphenyl ethers, tetrabromobisphenol A and phthalates are present in humans (Talsness et al., 2009). In addition, some compounds with heavy metals such as chro￾mium, cadmium and lead are often used in production of colorants, stabilizers and plasticizers (Ernst et al., 2000; Murphy, 2001). They can be released from plastic debris into water systems and further enter the food chain to cause bioaccumulation of toxins in organisms. Microplastics can be a vector for water-borne hydrophobic pollutants (Teuten et al., 2009; Lee et al., 2014; Bakir et al., 2016; O'Connor et al., 2016; Ziccardi et al., 2016). Examples include pol￾ychlorinated biphenyls (PCBs) (Velzeboer et al., 2014), and dichlorodiphenyltrichloroethane (DDT) (Rios et al., 2010), which are well known for their high toxicity and persistence in the environment. Due to the large specific surface areas and intrinsic hydrophobicity, the potential of hydrophobic chemical adsorption onto the surface of microplastics has caused great concerns over microplastics (Horton et al., 2017). PCBs are well known to be carcinogenic, mutagenic, and/or teratogenic (Hammer et al., 2012). DDT can lead to adverse neurological effects and immunodeficiency (Mansouri et al., 2017). The partition studies show the large partitioning coefficients for organic compounds, in the range of 104 e106 (Andrady, 2011). A few studies suggested that microplastics could sorb high amounts of PCBs form the surrounding seawater in coastal areas of USA and Japan (Hammer et al., 2012). Microplastics can sorb lubrication oils and heavy metals (Angiolillo et al., 2015; Hu et al., 2017). The aforementioned pollutants initially sorbed on the surface of microplastics can be released when microplastics are ingested and stay inside of human body, where pH is rather low, temperature is comparatively high and digestive liquids are present. A few studies confirmed that the desorption rate of sorbed contaminants in or￾ganisms was accelerated, substantially faster than that in seawater (Teuten et al., 2007; Bakir et al., 2014). However, very limited information is available about the real sorption behavior for water-borne pollutants in fresh water on the microplastics. The toxicity of microplastics in fresh waters is not well understood. A few lab-scale studies have provided some implications on the potential biological hazards from microplastics. A research con￾ducted by Ma et al. (2016) selected Daphnia magna as a model fresh water organism to study the toxicity of microplastics together with sorbed phenanthrene. They concluded that nano-sized PS exhibit high toxicity and physical damage to Daphnia magna and toxicity was enhanced by sorbed phenanthrene. A recent study demonstrated that the mixture of antimicrobial florfenicol and microplastics caused higher inhibition level of cholinesterase activity on freshwater bivalve Corbicula fluninea than that of florfenicol or microplastics (Guilhermino et al., 2018). It was found that microplastics caused certain levels of biological effects on Corbicula flumineaand and their predator Acipenser transmontanus (Rochman et al., 2017). A few studies on larval and adult zebrafish showed that microplastics were first ingested, became accumulated, and consequently caused alternations in locomotion, intestinal damage, and change in metabolic profiles (Lu et al., 2016; Chen et al., 2017; Sleight et al., 2017; Lei et al., 2018). With regards to other fresh water organisms, the toxicity studies on marine organisms may provide suggestive information. Hence, it is of importance to investigate the interaction between microplastics and key com￾pounds in lab-scale experiments, which will be helpful in the risk assessment of microplastics. The biological effects include the potential to geographically transfer microorganisms (Oberbeckmann et al., 2015). Microor￾ganisms can quickly colonize the surface of microplastics and be transported with the movement of microplastics, as plastics are usually durable and persistent than other media (Thiel and Gutow, 2005). While this interaction is commonly known and possible consequences are raised, such as the introduction of pathogens to a clean environment, limited literatures are available to reveal the diverse biofilms communities, and even less for freshwater envi￾ronment. (McCormick et al., 2014) conducted high-throughput sequencing analysis on the microplastics collected in an urban river in Chicago, Illinois. They found out some attached taxa were plastic decomposing organisms and pathogens; the findings sug￾gest that microplastics can transport bacterial assemblages in freshwater ecosystems. Their study also emphasized that patho￾genic wastewater-associated organisms could be discharged into waterways by means of the microplastics with the attachment of the organisms. A study on microplastics-associated bacteria in Yangtze Estuary also confirmed the presence of potential pathogens on microplastics (Jiang et al., 2018). Another biological effect is the change in the plastic physical properties because of biofilms that attach onto surface of micro￾plastics. Firstly, the density of microplastics can be increased (Carr et al., 2016), enabling light microplastics to sink in the water col￾umn and benthic zones. Furthermore, the biofilms can alter the surface nature of microplastics and make the surface less hydro￾phobic (Lobelle and Cunliffe, 2011; Zettler et al., 2013). The findings may provide some insights on studies associated with sorption of persistent organic pollutants. Whether the attached biofilms can enhance or weaken the interactions between microplastics and water-borne pollutants remains unknown to us. 366 J. Li et al. / Water Research 137 (2018) 362e374

J. Li et al. / Water Research137 (2018)3623743673.Microplastics sampling methodsThe size distribution and particle compositions were also depen-dent on sampling methods. For instance, while the PS particlesDespite the fact that studies on microplastics have been per-showed 96% of total particles in manta trawl samples, they seldomformedforyears,themethodologieson samplecollection,sampleappeared in surface microlayer samples and only accounted for0.2% of total particles.pre-treatment, quantification and identification are not standard-ized. The findings from reported studies vary significantly andThe size of microplastics can be as small as 1.6 μm to the uppercannot be easily compared.boundary of 5 mm (Cole et al., 2011). A study has shown plasticThe sampling methods can be volume-reduced and bulk sam-fragments of size less than 50 μm were underestimated (Song et al.pling (Hidalgo-Ruz et al, 2012). In the volume-reduced sampling.2015).Additionally,fordynamicfreshwaterbodies,likerivers,thethe volume of sample is reduced during the sampling period. In themicroplastics can quickly transfer to the aquatic environment.bulk sampling, no water volume is reduced. In the volume-reducedHowever,inisolatedandremotefreshwaterbodies,duetothelowsampling,neuston plankton net and manta trawl are twohuman activity, microplastics are trapped in the water and wouldcommonly adopted approaches, which aim at collecting surfacegradually fragment into tiny microplastics,which have the sizes ofwatersamples.Onlyafewstudiesusedthewatercollectedbythesub-micro meters. Hence, underestimation can be expected in thebulksamplingapproach.results,when a330-ummeshisused tocollectsamples.ItisBecause of its lower density than water, microplastics tend toinappropriate to apply the marine microplastics sampling methodsfloat on the surface of water. Therefore, a trawl along a transect canfor the work in fresh water environment.It is thereforean urgentbe applied to collect surface water samples (Rocha-Santos andneed to develop optimized and standardized methods for waterDuarte, 2015).During sampling,a flow meter is used to calculatesampling.the entire volume filtered through the mesh, by which the totalvolumeofwatercanbedetermined (Freeetal,2014)4. Sample extraction and purificationAs shown in Table 1, the researchers typically use net with330μm mesh to concentrate top water samples(Ryan etal.,2009:Themicroplasticsanalysismainlyconsistsof two steps:(1)Hidalgo-Ruz etal.,2012:Eriksen etal.,2013;Zhao etal.,2014;Maniextraction/purification, and (2) quantification and/or identification.et al.,2015:Baldwin et al.,2016:Su et al.,2016; Anderson etal..The microplastics must first be separated from the initial matrix in2017).Suchaselectionismainlyduetothelowersizeboundaryofordertoimproveandsimplifythesubsequentprocessforthemicroplasticsof333μmsuggested byNational Oceanic and At-quantificationand/oridentification.mospheric Administration of USA (Arthur et al., 2009). The ad-For the initial separation, density separation is the most oftenvantages of using the trawl are that it can cover large samplingused approach, which involves the mixing of the sample with aareas and reduce sample volume.liquid that has thedefined density (normally a saturated salt so-Only one study adopts 500-μm mesh in Austrian, leading to alution).Thesuspensionwasthenshakenandstirredforadefinedmuch smaller content of microplastics (Lechner et al., 2014).time, followed by the settling of the mixture. This process canHowever, as the mesh size is comparatively large, it may neglectenablethelow-densityparticlessuchasthemicroplasticstofloattosmall particles, which may be more important in fresh waters duethe upper layer of water and the high-density particles such astotheenvironmentalimpacts(Eerkes-Medranoetal,2015)inorganic clay to sink to the bottom.Thus, microplastics could beAdditionally. when the mesh of different sizes is selected,recoveredfromthesupernatantThe saturated NaCl solution with a density of1.2kg/L is themostgreater difference in abundanceof microplastics can always beobserved(LozanoandMouat,2009:Drisetal,2015;Zhang etal.commonly used to achieve this separating process due to its low2015:Wang et al.,2017).Lozano and Mouat (2009)found outcostandnotoxicitytohumans(Browneetal2o11:Claessensetal.that when an 80-μm mesh was used, the results were 100,0002011).Filtered seawater and distilled water are also used as sepa-timeshigherthanthatbya450-ummesh.Thesimilarobservationrating solutions(Hanveyetal.,2017).However,forthosedenserwas reported in a study of the waters in Great Paris (Dris et al.,microplastics that contain polyvinyl chloride or polyoxymethylene2015).Therefore, the concern over comparability and data credi-with densities of 1.16-1.58 kg/L and 1.41-1.61 kg/L, they cannot bebility from studies is being raised.recovered by this approach.In addition to common trawl methods,other approaches areOther salt solutions were also reportedly used.They are sodiumused. They include surface microlayer method, hand-net collectionpolytungstate (SPT) solution with a density of 1.4-1.5kg/Landbulkwatersampling.(Corcoran etal.,2009:Corcoran,2015),calcium chloride(CaCl2)Surface microlayer method enables researchers to collect upper-with a density of1.3kg/L(Stolte et al,2015)and sodium iodidelayer water. This process is performed manually by using a 2-mm(Nal) solutions with a density of 1.8 kg/L (Nuelle et al., 2014). Thesieve to get top 1-mm seawater (Song et al., 2014), while rotatingselection is based on the separation efficiency and the cost ofglass drum can automatically collect surface water at upper level ofmaterials.50-60 μm (Ng and Obbard,2006).Asanon-toxicseparatingsalt,theSPT wasappliedforliquidseparation for years(Munsterman and Kerstholt,1996).However, itHand-net sampling is another common method to collectmicroplastics samples.When one uses the method, hand bucket isis rather expensive when it is compared with other salts. A study byrequired to collect certain volume of surface water at top 20-cmStolte (2015) demonstrated that CaCl2 was not suitable as it causedlevel followed by thehand-net filtration (Moore et al.,2011;Songinterferenceinthemeasurementet.al,2014)Nal was used for the microplastics separation in several studies.For bulk water samples,the volume varies among studies.SongAlthough the density of Nal solution is as high as 1.8 kg/L, it isetal.(2014)reportedlycollectedashighas100-Lwatersamples,expensiveandmustbehandledwithcare(Nuelleetal,2014)while the volume of 100 mL and 2L was only collected for theZinc chloride (ZnCl2) solution with a dentistry of 1.6 kg/L wasstudies by Dubaish and Liebezeit (2013) and Leslie et al. (2017).also used forthe microplastics separation (lmhof et al,2012,2016):respectively. The results by different sampling approaches weremost types of microplastics can be recovered. However, it is haz-compared by Song et al. (2014). It was concluded that the meanardous compared to other substances reported. Thus, the subse-abundanceofmicroplasticspresentinsurfaceseawaterwasinthequent recycling and reuse of ZnCl2 with care are necessary in orderorder ofsurfacemicrolayer,handnet,bulkwaterand mantatrawl.to minimize environmental pollution

3. Microplastics sampling methods Despite the fact that studies on microplastics have been per￾formed for years, the methodologies on sample collection, sample pre-treatment, quantification and identification are not standard￾ized. The findings from reported studies vary significantly and cannot be easily compared. The sampling methods can be volume-reduced and bulk sam￾pling (Hidalgo-Ruz et al., 2012). In the volume-reduced sampling, the volume of sample is reduced during the sampling period. In the bulk sampling, no water volume is reduced. In the volume-reduced sampling, neuston plankton net and manta trawl are two commonly adopted approaches, which aim at collecting surface water samples. Only a few studies used the water collected by the bulk sampling approach. Because of its lower density than water, microplastics tend to float on the surface of water. Therefore, a trawl along a transect can be applied to collect surface water samples (Rocha-Santos and Duarte, 2015). During sampling, a flow meter is used to calculate the entire volume filtered through the mesh, by which the total volume of water can be determined (Free et al., 2014). As shown in Table 1, the researchers typically use net with 330 mm mesh to concentrate top water samples (Ryan et al., 2009; Hidalgo-Ruz et al., 2012; Eriksen et al., 2013; Zhao et al., 2014; Mani et al., 2015; Baldwin et al., 2016; Su et al., 2016; Anderson et al., 2017). Such a selection is mainly due to the lower size boundary of microplastics of 333 mm suggested by National Oceanic and At￾mospheric Administration of USA (Arthur et al., 2009). The ad￾vantages of using the trawl are that it can cover large sampling areas and reduce sample volume. Only one study adopts 500-mm mesh in Austrian, leading to a much smaller content of microplastics (Lechner et al., 2014). However, as the mesh size is comparatively large, it may neglect small particles, which may be more important in fresh waters due to the environmental impacts (Eerkes-Medrano et al., 2015). Additionally, when the mesh of different sizes is selected, greater difference in abundance of microplastics can always be observed (Lozano and Mouat, 2009; Dris et al., 2015; Zhang et al., 2015; Wang et al., 2017). Lozano and Mouat (2009) found out that when an 80-mm mesh was used, the results were 100,000 times higher than that by a 450-mm mesh. The similar observation was reported in a study of the waters in Great Paris (Dris et al., 2015). Therefore, the concern over comparability and data credi￾bility from studies is being raised. In addition to common trawl methods, other approaches are used. They include surface microlayer method, hand-net collection and bulk water sampling. Surface microlayer method enables researchers to collect upper￾layer water. This process is performed manually by using a 2-mm sieve to get top 1-mm seawater (Song et al., 2014), while rotating glass drum can automatically collect surface water at upper level of 50e60 mm (Ng and Obbard, 2006). Hand-net sampling is another common method to collect microplastics samples. When one uses the method, hand bucket is required to collect certain volume of surface water at top 20-cm level followed by the hand-net filtration (Moore et al., 2011; Song et al., 2014). For bulk water samples, the volume varies among studies. Song et al. (2014) reportedly collected as high as 100-L water samples, while the volume of 100 mL and 2 L was only collected for the studies by Dubaish and Liebezeit (2013) and Leslie et al. (2017), respectively. The results by different sampling approaches were compared by Song et al. (2014). It was concluded that the mean abundance of microplastics present in surface seawater was in the order of surface microlayer, hand net, bulk water and manta trawl. The size distribution and particle compositions were also depen￾dent on sampling methods. For instance, while the PS particles showed 96% of total particles in manta trawl samples, they seldom appeared in surface microlayer samples and only accounted for 0.2% of total particles. The size of microplastics can be as small as 1.6 mm to the upper boundary of 5 mm (Cole et al., 2011). A study has shown plastic fragments of size less than 50 mm were underestimated (Song et al., 2015). Additionally, for dynamic freshwater bodies, like rivers, the microplastics can quickly transfer to the aquatic environment. However, in isolated and remote freshwater bodies, due to the low human activity, microplastics are trapped in the water and would gradually fragment into tiny microplastics, which have the sizes of sub-micro meters. Hence, underestimation can be expected in the results, when a 330-mm mesh is used to collect samples. It is inappropriate to apply the marine microplastics sampling methods for the work in fresh water environment. It is therefore an urgent need to develop optimized and standardized methods for water sampling. 4. Sample extraction and purification The microplastics analysis mainly consists of two steps: (1) extraction/purification, and (2) quantification and/or identification. The microplastics must first be separated from the initial matrix in order to improve and simplify the subsequent process for the quantification and/or identification. For the initial separation, density separation is the most often used approach, which involves the mixing of the sample with a liquid that has the defined density (normally a saturated salt so￾lution). The suspension was then shaken and stirred for a defined time, followed by the settling of the mixture. This process can enable the low-density particles such as the microplastics to float to the upper layer of water and the high-density particles such as inorganic clay to sink to the bottom. Thus, microplastics could be recovered from the supernatant. The saturated NaCl solution with a density of 1.2 kg/L is the most commonly used to achieve this separating process due to its low cost and no toxicity to humans (Browne et al., 2011; Claessens et al., 2011). Filtered seawater and distilled water are also used as sepa￾rating solutions (Hanvey et al., 2017). However, for those denser microplastics that contain polyvinyl chloride or polyoxymethylene with densities of 1.16e1.58 kg/L and 1.41e1.61 kg/L, they cannot be recovered by this approach. Other salt solutions were also reportedly used. They are sodium polytungstate (SPT) solution with a density of 1.4e1.5 kg/L (Corcoran et al., 2009; Corcoran, 2015), calcium chloride (CaCl2) with a density of 1.3 kg/L (Stolte et al., 2015) and sodium iodide (NaI) solutions with a density of 1.8 kg/L (Nuelle et al., 2014). The selection is based on the separation efficiency and the cost of materials. As a non-toxic separating salt, the SPT was applied for liquid separation for years (Munsterman and Kerstholt, 1996). However, it is rather expensive when it is compared with other salts. A study by Stolte (2015) demonstrated that CaCl2 was not suitable as it caused interference in the measurement. NaI was used for the microplastics separation in several studies. Although the density of NaI solution is as high as 1.8 kg/L, it is expensive and must be handled with care (Nuelle et al., 2014). Zinc chloride (ZnCl2) solution with a dentistry of 1.6 kg/L was also used for the microplastics separation (Imhof et al., 2012, 2016); most types of microplastics can be recovered. However, it is haz￾ardous compared to other substances reported. Thus, the subse￾quent recycling and reuse of ZnCl2 with care are necessary in order to minimize environmental pollution. J. Li et al. / Water Research 137 (2018) 362e374 367

368J. Li et al. / Water Research 137 (2018)362374The extraction/separation methods are based on the densityAs the damage to the microplastics is negligible, the H2O2 basedseparationthroughthecombinationoffluidizationwithfloatationmethod is still themostpopularforthesampledigestion.(lvleva etal,2017).80%-100%ofmicroplastics (particles)could beOther chemical approaches such as ultrasonication can beobtained by saturated NaCl solution following the plain stirringcombined withthe deionized water or sodiumdodecyl sulfateextraction (Fries et al., 2013). However, the drawback is that smaller(SDS) solutions for the digestion (Cooper and Corcoran,2010:plastic particles of size <1 mm can only be separated to a smallEnders et al.,2015).However, some concerns had been raised asdegree of 40% (mass).tiny microplastics could be generated from the brittle plasticSeveral strategies were proposed and adapted. It was reportedsamples (Loderand Gerdts,2015).that the combination of fluidization in a Nacl solution and floata-Another emerging approach to remove the organic matters isthe enzymatic degradation. During the degradation process,tion in Nal solution gave a very good recovery rate of 99% (Nuelleet al., 2014).The newly developed instrument, Munich Plasticmicroplastics samples are incubated with a mixture of technicalSediment Separator, in combination with the use of ZnClz solution,enzymes such as lipase, amylase, proteinase, chitinase, and cellu-can yield a particle number recovery of up to 96-100% for largerlose (Coleet al,2014;Loder etal.,2015).Such organicmatters asmicroplastics and 96% for smaller microplastics, which was mainlyproteins, lipids and carbohydrates can be specifically removed. Coleusedforthemicroplasticsseparationforthesedimentsamplesetal.(2014)reportedtheapplicationoftheproteolyticenzyme(lmhofetal..2012)(proteinase-K)in the treatment of the marine samples, More thanAnother alternative is through the pressurized fluid extraction97% (by weight) of the materials present in plankton-rich seawater(PFE). By optimizing the PFE conditions, plastics can physically besamples were removed within a few hours, whereas the micro-separated from waste and soil samples (Fuller and Gautam,2016).plastic debris present was unaffected. This method should furtherAlthough theplastic particlesas small as 30 μmcan efficiently bebe optimized as the molecular biological enzyme is costly.extracted, the challenges of using the method are that: (1) sizeThe basic enzymatic purification protocol (BEPP) was used todistribution cannot well be determined; (2) morphology ofpurify samples for subsequent spectroscopic analyses (Loder et al.microplastics after extraction is changed.2017).Combined with the detergent sodium dodecyl sulfate (SDS)In addition to the density separation, alternative separationtechnical-grade enzymes (e.g-, protease, cellulase and chitinase) intechniques are reportedly used. For example, elutriation canphosphate buffer saline solutions and H2O2, with the density sep-effectivelyisolatemicroplasticsfrommarinesediments(Claessensaration by ZnClz, the purification efficiency can reach 98.3% foret al.,2013).Thelightmicroplasticscanberetainedthroughtheplankton samples.A universal enzymatic purification protocol (UEPP) differentupward water flow and aeration in a self-designed elutriation andfrom the aforementioned BEPP was further developed (Loder et al.,floatation apparatus, and then be collected on a 35-μm sieve. Anextraction efficiencyof 94-98%can beobtained.Its limitation is2017).IntheUEPP,anothertwomoreenzymes(lipaseandamylase)that it islessapplicableforwastewater,whichcontainsahighwereadded and digestion wasprocessed in tris(hydroxymethyl)concentration of organic matter. This is due to the negligible den-aminomethane (Tris) and sodium acetate buffer. The modifiedsitydifferencebetweenthemicroplasticsandthenaturallyoccur-protocol can be implemented on samples from different environ-ring particles within these matrices, making the elutriation methodmental matrices. Although these two protocols can reduce opera-difficult to achieve.tion timeof microscopy and have more reliable spectroscopicTo improve the efficiency of the extraction for samples takenresults, it requires long digestion time up to 15 days. Therefore, thefrom sediment and sewage,density separation to separate thenew enzymatic purification approach should still be optimized andlighter microplastics is always the first step,followed by the puri-the flexibility in the large-scale routine field sample purificationfication. This becomes especially important for bulk samples.still needs to be further investigation.However, thesequence inthe operations isless importantfor water5. Microplastics identification and quantificationsamples.The purification process can be divided into two main differentAfter the microplastics-containing water is treated, visual sort-categoriesof chemical degradation and enzymaticdegradation,inorder that the interfering matters such as organic tissues anding is conducted. This is commonly used by the majority of theinorganicdustscanberemoved.Inthefirstapproach,themicro-researchgroups,toselectthesuspectedmicroplasticsforfurtheranalysis and identification (Derraik,2002;Hidalgo-Ruz etal,2012:plastics samples aretreated withdifferentchemicals,mainly10%-30%hydrogenperoxide(H202)solution(Zettleretal.,2013:NuelleRochman,2015:Qiuetal.,2016),However,visualsortingmaynotetal.,2014;Zhaoetal.,2017)or peroxidemixed withmineralacidsprovideaccurateinformationonmicroplasticsabundanceduetosuch as sulfuric acid (lmhof et al.,2013).presence of such particles as clay and algae. If the aftermentionedCole etal. (2014)studied the efficiency of differentacid,alkalinetreatment is not conducted, it is very difficult to visually differen-and enzymatic digestion methods for digestion of biological ma-tiate the microplastics from other extracted organic and inorganicterials, including organisms/microorganisms and attached biofilmsparticles of similar size and shape, especially for the samples thatfrom marine surface trawls.The damage of microplastics may beare not pre-treated as aforementioned. Additional approaches suchless. The non-oxidizing acids such as hydrochloric acid at lowasthespectroscopicapproachesarerequired.Theycanserveasaconcentrations at room temperature are insufficient, yielding largemorereliable technical tool for the identificationof the plasticparticles, instead of relying on visual observation that wouldamount oforganicresidues after the digestion, whereas the strongoxidizingacids suchassulfuricacidandnitricacid would destroyorintroduce largevariations intheresultsThe current techniques for quantitative and qualitative researchdamage the microplastics made by polymers with less tolerance atlowpH.ofmicroplasticsincludepyrolysisgaschromatographycoupledtoA study on digestion of the microplastics-containing samplesmass spectrometry (Fabbri etal.,2000;Fabbri,2001;Fries et al.taken from marine environment by 1-M sodium hydroxide showed2013:Nuelle et al.,2014),Raman spectroscopy (Cole et al.,2013;a higher effectiveness of 90% (Cole et al., 2014). The digestion effi-Collard et al.,2015;Imhof etal.,2016:Qiu etal,2016:Wiesheuciency would increase with the increase in the molarity and theet al.,2016: Zhao et al.,2017).Fourier-transform infrared (FT-IR)temperature.Theplanktonweredigestedby10-MNaOHat60C;spectroscopy(Song etal.2014;Besseling etal.,2015:Qiu etal.however,suchaharshconditionwoulddamagethemicroplastics.2016),liquid chromatography (Hintersteiner et al,2015:Elert

The extraction/separation methods are based on the density separation through the combination of fluidization with floatation (Ivleva et al., 2017). 80%e100% of microplastics (particles) could be obtained by saturated NaCl solution following the plain stirring extraction (Fries et al., 2013). However, the drawback is that smaller plastic particles of size <1 mm can only be separated to a small degree of 40% (mass). Several strategies were proposed and adapted. It was reported that the combination of fluidization in a NaCl solution and floata￾tion in NaI solution gave a very good recovery rate of 99% (Nuelle et al., 2014). The newly developed instrument, Munich Plastic Sediment Separator, in combination with the use of ZnCl2 solution, can yield a particle number recovery of up to 96e100% for larger microplastics and 96% for smaller microplastics, which was mainly used for the microplastics separation for the sediment samples (Imhof et al., 2012). Another alternative is through the pressurized fluid extraction (PFE). By optimizing the PFE conditions, plastics can physically be separated from waste and soil samples (Fuller and Gautam, 2016). Although the plastic particles as small as 30 mm can efficiently be extracted, the challenges of using the method are that: (1) size distribution cannot well be determined; (2) morphology of microplastics after extraction is changed. In addition to the density separation, alternative separation techniques are reportedly used. For example, elutriation can effectively isolate microplastics from marine sediments (Claessens et al., 2013). The light microplastics can be retained through the upward water flow and aeration in a self-designed elutriation and floatation apparatus, and then be collected on a 35-mm sieve. An extraction efficiency of 94e98% can be obtained. Its limitation is that it is less applicable for wastewater, which contains a high concentration of organic matter. This is due to the negligible den￾sity difference between the microplastics and the naturally occur￾ring particles within these matrices, making the elutriation method difficult to achieve. To improve the efficiency of the extraction for samples taken from sediment and sewage, density separation to separate the lighter microplastics is always the first step, followed by the puri- fication. This becomes especially important for bulk samples. However, the sequence in the operations is less important for water samples. The purification process can be divided into two main different categories of chemical degradation and enzymatic degradation, in order that the interfering matters such as organic tissues and inorganic dusts can be removed. In the first approach, the micro￾plastics samples are treated with different chemicals, mainly 10%e 30% hydrogen peroxide (H2O2) solution (Zettler et al., 2013; Nuelle et al., 2014; Zhao et al., 2017) or peroxide mixed with mineral acids such as sulfuric acid (Imhof et al., 2013). Cole et al. (2014) studied the efficiency of different acid, alkaline and enzymatic digestion methods for digestion of biological ma￾terials, including organisms/microorganisms and attached biofilms from marine surface trawls. The damage of microplastics may be less. The non-oxidizing acids such as hydrochloric acid at low concentrations at room temperature are insufficient, yielding large amount of organic residues after the digestion, whereas the strong oxidizing acids such as sulfuric acid and nitric acid would destroy or damage the microplastics made by polymers with less tolerance at low pH. A study on digestion of the microplastics-containing samples taken from marine environment by 1-M sodium hydroxide showed a higher effectiveness of 90% (Cole et al., 2014). The digestion effi- ciency would increase with the increase in the molarity and the temperature. The plankton were digested by 10-M NaOH at 60 C; however, such a harsh condition would damage the microplastics. As the damage to the microplastics is negligible, the H2O2 based method is still the most popular for the sample digestion. Other chemical approaches such as ultrasonication can be combined with the deionized water or sodium dodecyl sulfate (SDS) solutions for the digestion (Cooper and Corcoran, 2010; Enders et al., 2015). However, some concerns had been raised as tiny microplastics could be generated from the brittle plastic samples (Loder and Gerdts, 2015 € ). Another emerging approach to remove the organic matters is the enzymatic degradation. During the degradation process, microplastics samples are incubated with a mixture of technical enzymes such as lipase, amylase, proteinase, chitinase, and cellu￾lose (Cole et al., 2014; Loder et al., 2015 € ). Such organic matters as proteins, lipids and carbohydrates can be specifically removed. Cole et al. (2014) reported the application of the proteolytic enzyme (proteinase-K) in the treatment of the marine samples. More than 97% (by weight) of the materials present in plankton-rich seawater samples were removed within a few hours, whereas the micro￾plastic debris present was unaffected. This method should further be optimized as the molecular biological enzyme is costly. The basic enzymatic purification protocol (BEPP) was used to purify samples for subsequent spectroscopic analyses (Loder et al., € 2017). Combined with the detergent sodium dodecyl sulfate (SDS), technical-grade enzymes (e.g., protease, cellulase and chitinase) in phosphate buffer saline solutions and H2O2, with the density sep￾aration by ZnCl2, the purification efficiency can reach 98.3% for plankton samples. A universal enzymatic purification protocol (UEPP) different from the aforementioned BEPP was further developed (Loder et al., € 2017). In the UEPP, another two more enzymes (lipase and amylase) were added and digestion was processed in tris(hydroxymethyl)- aminomethane (Tris) and sodium acetate buffer. The modified protocol can be implemented on samples from different environ￾mental matrices. Although these two protocols can reduce opera￾tion time of microscopy and have more reliable spectroscopic results, it requires long digestion time up to 15 days. Therefore, the new enzymatic purification approach should still be optimized and the flexibility in the large-scale routine field sample purification still needs to be further investigation. 5. Microplastics identification and quantification After the microplastics-containing water is treated, visual sort￾ing is conducted. This is commonly used by the majority of the research groups, to select the suspected microplastics for further analysis and identification (Derraik, 2002; Hidalgo-Ruz et al., 2012; Rochman, 2015; Qiu et al., 2016). However, visual sorting may not provide accurate information on microplastics abundance due to presence of such particles as clay and algae. If the aftermentioned treatment is not conducted, it is very difficult to visually differen￾tiate the microplastics from other extracted organic and inorganic particles of similar size and shape, especially for the samples that are not pre-treated as aforementioned. Additional approaches such as the spectroscopic approaches are required. They can serve as a more reliable technical tool for the identification of the plastic particles, instead of relying on visual observation that would introduce large variations in the results. The current techniques for quantitative and qualitative research of microplastics include pyrolysis gas chromatography coupled to mass spectrometry (Fabbri et al., 2000; Fabbri, 2001; Fries et al., 2013; Nuelle et al., 2014), Raman spectroscopy (Cole et al., 2013; Collard et al., 2015; Imhof et al., 2016; Qiu et al., 2016; Wiesheu et al., 2016; Zhao et al., 2017), Fourier-transform infrared (FT-IR) spectroscopy (Song et al., 2014; Besseling et al., 2015; Qiu et al., 2016), liquid chromatography (Hintersteiner et al., 2015; Elert 368 J. Li et al. / Water Research 137 (2018) 362e374

369J. Li et al. / Water Research 137 (2018)362374Table 3Summary of commonly used analytical techniques for analysis of microplastics. MethodMethodologyParticle Size by differentAdvantagesLimitationsanalytical methodsVisual MethodMicroscopicPretreated samples areThe particles with size down to Samples with a relative highThe nature of the samplesCountingidentified by microscopy. Themicro-meter(um)rangecanbeamountof largemicroplasticscannot be determined and it iscan be identified by thisnecessary to couple withparticles are counted directly.identified by themethods quickly. providing anidentification methods listedstereomicroscopeoverall picture of microplasticsbelow.abundanceinshorttimewithlow cost.SpectroscopicFTIRSamples are subjected toLarger particlesizeof>500 μmFTIR based techniques arenon-Samples must be IR-active; theMethodinfrared radiation with defined can be analyzed by ATR-FTIR,destructive methods.They aresamples below 20 μm might notrange and the excitableSmaller particledown to 20 μm well established,fast andquite yield not enough absorbancevibrations depend on thecan be analyzed by microscopy reliable. The newly emerginginterpretable spectra. Non-coupled FIR.composition and molecularautomatic FTIR imaging such as transparent particles arestructure of a substance, PlasticFPA makes it possible for thedifficult to be analyzed by thispolymers have highly specific IRfast acquisition of severalmethod. The highly specificspectra with distinct bandthousand spectra within an area instruments are expensive andthrough one singlerequire experienced personalspatterns.measurements and the analysis for operation and datatime is greatly shortened.processing. The detection isaffected by the environmentalmatrix (e.g., biofilm formationon polymer).which createsdifficulty in the datainterpretation. The sample mustbe pretreated to eliminate IR-active water.RamanThere are great interferences ofThe interaction of the irradiated Microscopy coupled RamanMicroscopy coupled RS allowsSpectroscopylaser light with the moleculesSpectroscopy (RS) method isthe analysis for small particlesfluorescence from biological,and atoms of the sample results suitable for particles size > 1between 1 to 20 μm with highorganic and inorganicin differences in the frequencyμm. It is the only methodspatial resolution and relatively impurities, hampering theof the back-scattered light when available for particles falling in low sensitivity towards water. It identification of microplastics.compared to the irradiatingis possible to analyse non-Sample requires purificationthe range of 1 to 20 μm.laser. The Raman shift can betransparent and dark particlesbefore analysis: appropriatedetected, leading to substance-usingRS method:the fastRaman acquisition parametersspecific Raman spectra.chemical mapping can be(e.g., wavelength, laser power.performed by RS method,and photo bleaching)areenablingfast and automaticimportant.The automaticdata collection /processingmapping of micro-RS is stillunder development. Theanalysis by RS is time-consumingScanningThe sample images areParticle of sizes downto micro The high-resolution image ofSamples need to be coated atElectronproduced by the interaction of scale can be analyzed.the samples can be produced by high vacuum; no detailedSpectroscopyan electron beam withthethis method.identification information isavailable.sample to measure thesecondary ions.Chromatographic Thermo-Samples are thermally treatedThe method is suitable forThe sample can be analyzedOnly one particle with certainMethodanalyticalunder ambient conditions andsamples with sizes>500 um,together with the organicweight can be assessed per run.methods such as the released gaseouswhich can be manually handled plastic additives in one runThe pyrolysis database is onlypyrolysis GC/MS compounds are trapped,andwithout the use of solvents and available for selected polymersby tweezers.subsequently transferred to athus avoid backgroundonly suchas PEandGC column,coupled to acontamination, The method ispolypropylene,quadrupole-MS. The spectra ofsensitive and reliable. Somecommon polymers spectra datapyrolysis products are thenlibrary is available.compared to a database ofcommon plastic types.LiquidSamples are dissolved bySufficient sample size of several The recoveries of selectedInability to determine theChromatography selected solvents.Differentmilligrams is required for thepolymers were highphysical characteristics, such asmolar mass distribution ischemical extraction to conductsizeinformation,andmeasured by size exclusionthis testing.restrictions on polymer typeschromatography andlimit its applications toquantification is based on HPLCenvironmental samples. Onlyanalysis.small amountofsamplescanbeassessed per run. Only specificpolymers such as PS and PETcan be analyzed by this methodOther MethodsTagging Method Hydrophobic dye adsorbs ontoMicroplastics with sample size This method is straightforward Other particles such as organicsurfaces of microplastics anddown tomicroscale range candebris might be stained by theand allows fast screening ofrenders them fluorescent when be visualized and counted.microplastics with low cost, The dye. This would lead to over-they iradiated with blue light.fluorescent particles can beestimation of microplasticscounted and may be identified abundance

Table 3 Summary of commonly used analytical techniques for analysis of microplastics. Method Methodology Particle Size by different analytical methods Advantages Limitations Visual Method Microscopic Counting Pretreated samples are identified by microscopy. The particles are counted directly. The particles with size down to micro-meter (mm) range can be identified by the stereomicroscope. Samples with a relative high amount of large microplastics can be identified by this methods quickly, providing an overall picture of microplastics abundance in short time with low cost. The nature of the samples cannot be determined and it is necessary to couple with identification methods listed below. Spectroscopic Method FTIR Samples are subjected to infrared radiation with defined range and the excitable vibrations depend on the composition and molecular structure of a substance. Plastic polymers have highly specific IR spectra with distinct band patterns. Larger particle size of > 500 mm can be analyzed by ATR-FTIR, Smaller particle down to 20 mm can be analyzed by microscopy coupled FTIR. FTIR based techniques are non￾destructive methods. They are well established, fast and quite reliable. The newly emerging automatic FTIR imaging such as FPA makes it possible for the fast acquisition of several thousand spectra within an area through one single measurements and the analysis time is greatly shortened. Samples must be IR eactive; the samples below 20 mm might not yield not enough absorbance interpretable spectra. Non￾transparent particles are difficult to be analyzed by this method. The highly specific instruments are expensive and require experienced personals for operation and data processing. The detection is affected by the environmental matrix (e.g., biofilm formation on polymer), which creates difficulty in the data interpretation. The sample must be pretreated to eliminate IR￾active water. Raman Spectroscopy The interaction of the irradiated laser light with the molecules and atoms of the sample results in differences in the frequency of the back-scattered light when compared to the irradiating laser. The Raman shift can be detected, leading to substance￾specific Raman spectra. Microscopy coupled Raman Spectroscopy (RS) method is suitable for particles size > 1 mm. It is the only method available for particles falling in the range of 1 to 20 mm. Microscopy coupled RS allows the analysis for small particles between 1 to 20 mm with high spatial resolution and relatively low sensitivity towards water. It is possible to analyse non￾transparent and dark particles using RS method; the fast chemical mapping can be performed by RS method, enabling fast and automatic data collection /processing. There are great interferences of fluorescence from biological, organic and inorganic impurities, hampering the identification of microplastics. Sample requires purification before analysis; appropriate Raman acquisition parameters (e.g., wavelength, laser power, and photo bleaching) are important. The automatic mapping of micro-RS is still under development. The analysis by RS is time￾consuming. Scanning Electron Spectroscopy The sample images are produced by the interaction of an electron beam with the sample to measure the secondary ions. Particle of sizes down to micro scale can be analyzed. The high-resolution image of the samples can be produced by this method. Samples need to be coated at high vacuum; no detailed identification information is available. Chromatographic Method Thermo￾analytical methods such as pyrolysis GC/MS Samples are thermally treated under ambient conditions and the released gaseous compounds are trapped, and subsequently transferred to a GC column, coupled to a quadrupole - MS. The spectra of pyrolysis products are then compared to a database of common plastic types. The method is suitable for samples with sizes > 500 mm, which can be manually handled by tweezers. The sample can be analyzed together with the organic plastic additives in one run without the use of solvents and thus avoid background contamination. The method is sensitive and reliable. Some common polymers spectra data library is available. Only one particle with certain weight can be assessed per run. The pyrolysis database is only available for selected polymers only such as PE and polypropylene. Liquid Chromatography Samples are dissolved by selected solvents. Different molar mass distribution is measured by size exclusion chromatography and quantification is based on HPLC analysis. Sufficient sample size of several milligrams is required for the chemical extraction to conduct this testing. The recoveries of selected polymers were high. Inability to determine the physical characteristics, such as size information, and restrictions on polymer types limit its applications to environmental samples. Only small amount of samples can be assessed per run. Only specific polymers such as PS and PET can be analyzed by this method. Other Methods Tagging Method Hydrophobic dye adsorbs onto surfaces of microplastics and renders them fluorescent when they irradiated with blue light. Microplastics with sample size down to microscale range can be visualized and counted. This method is straightforward and allows fast screening of microplastics with low cost. The fluorescent particles can be counted and may be identified. Other particles such as organic debris might be stained by the dye. This would lead to over￾estimation of microplastics abundance. J. Li et al. / Water Research 137 (2018) 362e374 369

370J.Lietal./WaterResearch137(2018)362-374et al.,2017)and themost recently reported tagging method (Shimfor simultaneous visualization,mapping and collection of spectra.etal.,2016).Table3summarizesmostoftheanalyticaltechniques,Themicro-FTIR is particularlyuseful as themembranefilter canbeand provides their advantages and limitations.directly used for the visualization with little sample preparation.Amongthem,manualcountingunderastereomicroscopeistheThe micro-FTIR analyses of plastics can beperformed ineithermost straightforward and thus widely used. Noren (2007) proposedtransmission or reflectance mode. The transmission mode givesseveral standardized criteria on a strict and conservative exami-high-quality spectra, but requires substrates to be infrared-nation of microplastics with the key points as follows:transparent. The analyses in reflectance mode can be done forthick samples. Meanwhile, irregularly-shaped microplastics would(1) The particles or fibers that have structure of biological or-produce non-interpretable spectra dueto refractive error (Harrisonganisms should not be included as microplastics:et al.,2012).Thus only transparent microplastics with certain(2) The fibers can be counted as microplastics if they have three-thickness with regular shape can be analyzed: otherwise the signalwill be disturbed/distorted by reflection error caused by lightdimensional structure;(3) The particles can be counted as microplastics if they can bescattering.Thesedrawbackscanhardlybeavoidedhomogeneouslycolored;Themicro-FTiRrequirestremendoustimeandeffortstofindthe(4) The microplastics must be transparent or whitish in colorsuitablemicroplasticsparticlesfortheanalytical work.Thecontactand be studied under high magnification with the help ofof the crystal with the inorganic particles could impose greatdamage to the expensive instrumentation.fluorescence microscopyLevin and Bhargava (2005) used the approach of micro-FTIRThere are a few drawbacks to the aforementioned criteria. Themapping. with the sequential measurement of IR spectra atresults from the visual sorting are strongly affected by severalspatially separated and manually-selected points on the samplefactors including: (1) personal factors (e.g. carelessness), (2) mi-surface. However, only small areas of the filter paper could becroscopy quality, and (3) sample matrix. Furthermore, visualanalyzed and the process is very time-consuming.counting suffersthe drawback of size limitation dueto the reso-A FPA-based FTIR imaging with several detectors placed in a gridlution of the microscopy. Up to 70% error rates can be observed; thepattern was applied for microplastics analysis (Tagg et al,2015).number of error increases with a decrease in particle size (Hidalgo-This method allows for detailed and unbiased high-throughputRuzetal,2012)screeningoftotalmicroplasticsonthewholefilterpaper.ItcanA study was conducted on the microplastics abundance in theenablethesimultaneousrecordingofseveralthousand spectrainasediment samples collected from the North Sea. The results by thetargeted area within a single measurement run, and thus generatevisual counting approach seemed overestimated based on thechemical images for the whole filter paper. The screening andadditional study by the focal plane array (FPA)-based micro-FTIRanalysis for the whole sample filter paper may become possible thisspectroscopy.Itwasfoundthat,only1.4%oftheparticlesvisuallytechniqueobserved were the synthetic polymer by the micro-FTIR spectros-This method has several limitations as follows. The lateral res-olution of micro-FTIR spectroscopy is always limited to certaincopy. The other particles visually found were of other origins suchas organic matters and dusts (Loder etal,2015)diffraction range (e.g. 10 μmat 1000 cm-). The smaller particlesAmong the commonly reported techniques, Fourier transformwith irregular shape are less applicable. The samples down toinfrared spectroscopyand Ramanspectroscopyaremostcommonly20 μm cannot be analyzed (Loder and Gerdts,2015).As water isusedto identifymicroplastics(Besseling etal,2015Ivlevaetal.,stronglyIR active,which produces broad peaks over 3000cm-1,the2017;Zhao et al., 2017).In the testing, the microplastic samplessample preparation is required prior to measurement.are excited, by which the structure specific vibrations can beRaman spectroscopyas a surface analytical technique allows fordetected. The produced characteristic spectra with the fingerprintthe study of large and visually sorted particles. The micro-Ramanmicroscopy combined with Raman spectra imaging techniquerange allow the identification of nature of materials (namely plas-tics and non-plastics).Thepolymeridentification wasachieved bytheoretically allows forthespectraanalysis of wholefilters at acomparing the obtained spectra with the known reference spectra.spatial resolution below1 μm(lmhofet al,2016:Zhaoetal.,2017).For FTIR spectroscopy, the sample is irradiated with IR light withHowever,as Raman spectroscopy isbased on the methodology thata defined wavelengthrange and the IR radiation absorbancethefluorescentsamplesareexcitedbythelaser,thecontaminantsby biological residual and others would interfere with the spectra,collectedbytheequipmentisstructurespecificTheprerequisitefor theIR absorptionisthe change of thepermanent dipoleleadingtothedifficultyinproducingtheinterpretablespectramoment of a chemical bond, thus this technique is mainly used toThe thermoanalytical methods such as pyrolysis-GC/MS andanalyze the molecules with the polar functional groups such asTGA-MS have been used for the microplastic analysis. Samples arecarbonyl groups.firstly thermally degraded and the resultant products are subse-Raman spectroscopy is a scattering method. The laser with thequently sent to the mass spectrometer for analysis. The collectedsingle wavelength is applied to excite the molecule; the radiationdataarecomparedwithreferenceto obtainsuchsampleinforma-interactionwiththesample isdetected.Theprerequisite of thetion as identity and concentration.Raman spectroscopy is the change in the polarizability of a chem-However, the techniques have a limitation on size of plasticical bond and thus this technique is mainly for compounds withmatters.Itisdifficulttohandlesamples500μmcanbeanalyzedusingATR-FTIRraphy mass spectrometer (TDS-GC/MS). The environmental fieldFor smaller particles, the micro-FT-IR spectroscopy is a good toolsample with certain impurities such as the organics and

et al., 2017) and the most recently reported tagging method (Shim et al., 2016). Table 3 summarizes most of the analytical techniques, and provides their advantages and limitations. Among them, manual counting under a stereomicroscope is the most straightforward and thus widely used. Noren (2007) proposed several standardized criteria on a strict and conservative exami￾nation of microplastics with the key points as follows: (1) The particles or fibers that have structure of biological or￾ganisms should not be included as microplastics; (2) The fibers can be counted as microplastics if they have three￾dimensional structure; (3) The particles can be counted as microplastics if they can be homogeneously colored; (4) The microplastics must be transparent or whitish in color and be studied under high magnification with the help of fluorescence microscopy. There are a few drawbacks to the aforementioned criteria. The results from the visual sorting are strongly affected by several factors including: (1) personal factors (e.g. carelessness), (2) mi￾croscopy quality, and (3) sample matrix. Furthermore, visual counting suffers the drawback of size limitation due to the reso￾lution of the microscopy. Up to 70% error rates can be observed; the number of error increases with a decrease in particle size (Hidalgo￾Ruz et al., 2012). A study was conducted on the microplastics abundance in the sediment samples collected from the North Sea. The results by the visual counting approach seemed overestimated based on the additional study by the focal plane array (FPA)-based micro-FTIR spectroscopy. It was found that, only 1.4% of the particles visually observed were the synthetic polymer by the micro-FTIR spectros￾copy. The other particles visually found were of other origins such as organic matters and dusts (Loder et al., 2015 € ). Among the commonly reported techniques, Fourier transform infrared spectroscopy and Raman spectroscopy are most commonly used to identify microplastics (Besseling et al., 2015; Ivleva et al., 2017; Zhao et al., 2017). In the testing, the microplastic samples are excited, by which the structure specific vibrations can be detected. The produced characteristic spectra with the fingerprint range allow the identification of nature of materials (namely plas￾tics and non-plastics). The polymer identification was achieved by comparing the obtained spectra with the known reference spectra. For FTIR spectroscopy, the sample is irradiated with IR light with a defined wavelength range and the IR radiation absorbance collected by the equipment is structure specific. The prerequisite for the IR absorption is the change of the permanent dipole moment of a chemical bond, thus this technique is mainly used to analyze the molecules with the polar functional groups such as carbonyl groups. Raman spectroscopy is a scattering method. The laser with the single wavelength is applied to excite the molecule; the radiation interaction with the sample is detected. The prerequisite of the Raman spectroscopy is the change in the polarizability of a chem￾ical bond and thus this technique is mainly for compounds with aromatic bonds, CeH and C¼C double bonds. FTIR and Raman spectroscopy are complementary vibrational techniques to each other. Both would provide complementary in￾formation on microplastic samples. FTIR spectroscopy is a non-destructive technique with well￾established polymer database. There are three different operating modes available for FTIR, namely, transmission, reflection and attenuated total-reflectance (ATR) mode. Larger plastic samples >500 mm can be analyzed using ATR-FTIR. For smaller particles, the micro-FT-IR spectroscopy is a good tool for simultaneous visualization, mapping and collection of spectra. The micro-FTIR is particularly useful as the membrane filter can be directly used for the visualization with little sample preparation. The micro-FTIR analyses of plastics can be performed in either transmission or reflectance mode. The transmission mode gives high-quality spectra, but requires substrates to be infrared￾transparent. The analyses in reflectance mode can be done for thick samples. Meanwhile, irregularly-shaped microplastics would produce non-interpretable spectra due to refractive error (Harrison et al., 2012). Thus only transparent microplastics with certain thickness with regular shape can be analyzed; otherwise the signal will be disturbed/distorted by reflection error caused by light scattering. These drawbacks can hardly be avoided. The micro-FTIR requires tremendous time and efforts to find the suitable microplastics particles for the analytical work. The contact of the crystal with the inorganic particles could impose great damage to the expensive instrumentation. Levin and Bhargava (2005) used the approach of micro-FTIR mapping, with the sequential measurement of IR spectra at spatially separated and manually-selected points on the sample surface. However, only small areas of the filter paper could be analyzed and the process is very time-consuming. A FPA-based FTIR imaging with several detectors placed in a grid pattern was applied for microplastics analysis (Tagg et al., 2015). This method allows for detailed and unbiased high-throughput screening of total microplastics on the whole filter paper. It can enable the simultaneous recording of several thousand spectra in a targeted area within a single measurement run, and thus generate chemical images for the whole filter paper. The screening and analysis for the whole sample filter paper may become possible this technique. This method has several limitations as follows. The lateral res￾olution of micro-FTIR spectroscopy is always limited to certain diffraction range (e.g. 10 mm at 1000 cm
1 ). The smaller particles with irregular shape are less applicable. The samples down to 20 mm cannot be analyzed (Loder and Gerdts, 2015 € ). As water is strongly IR active, which produces broad peaks over 3000 cm
1 , the sample preparation is required prior to measurement. Raman spectroscopy as a surface analytical technique allows for the study of large and visually sorted particles. The micro-Raman microscopy combined with Raman spectra imaging technique theoretically allows for the spectra analysis of whole filters at a spatial resolution below 1 mm (Imhof et al., 2016; Zhao et al., 2017). However, as Raman spectroscopy is based on the methodology that the fluorescent samples are excited by the laser, the contaminants by biological residual and others would interfere with the spectra, leading to the difficulty in producing the interpretable spectra. The thermoanalytical methods such as pyrolysis-GC/MS and TGA-MS have been used for the microplastic analysis. Samples are firstly thermally degraded and the resultant products are subse￾quently sent to the mass spectrometer for analysis. The collected data are compared with reference to obtain such sample informa￾tion as identity and concentration. However, the techniques have a limitation on size of plastic matters. It is difficult to handle samples <500 mm as the tiny small samples cannot be put into the tube. In addition, the method is less applicable for mixtures with high concentration of impurities (Ivleva et al., 2017). Another thermoanalytical method was used by Duemichen et al. (2014). It allows the direct microplastics assessment from the field environmental samples with ease in sample preparation. The thermogravimetric analysis is combined with the solid-phase extraction (TGA-SPE) and thermal desorption gas chromatog￾raphy mass spectrometer (TDS-GC/MS). The environmental field sample with certain impurities such as the organics and 370 J. Li et al. / Water Research 137 (2018) 362e374

371J. Li et al. / Water Research 137 (2018) 362374microplastics particles (up to 20 mg) was placed in a TGA crucible,identification of microplastics should be first developed and veri-followedbyaheatingprocessunderinertconditionssuchasNzupfiedsothatdatafromdifferentresearchescanbemorecomparableto8ooCTheresultantproductsthataresorbedontheextractionandreliable.KeyissuessuchasenvironmentalriskassessmentcansubstratecanbetestedbyGC/MSforthedeterminationof thethen be conducted scientifically. Subsequently, rules and regula-composition. Only the results of polyethylene (PE) were published.tions can be formulated and enforced.In addition to gas chromatography, liquid chromatography wasapplied in quantification ofmicroplastics(Hintersteiner et al.,2015:7. ConclusionsElert et al, 2017).The method takes advantage of the differentsolubility of plastics. Appropriate solvents are used to dissolveMicroplastics have become one of emerging contaminants indifferent polymers. For instance, tetrahydrofurane and hexa-the aquatic environment. The presence has been reported in manyfluoroisopropanol could dissolve PS and polyethylene terephthalateplaces around the world, and has caused great public concerns.(PET), respectively (Elert et al, 2017).However, there is still a lack of sufficient knowledge about micro-After the preparation of polymer extracts, the samples can beplastics in freshwaters such as their health effects and fastanalyzed by High Performance Liquid Chromatography (HPLC)monitoring.coupledwithsizeexclusionsystems.ThoughunabletodetermineThis review article summarized the current status of micro-the sizes of plastic particles, this method reportedly has high re-plastics contamination in freshwaters, including rivers,lakes, watercoveries and can quantify the microplastics. However, this methodtreatment plants and drinking water. The potential environmentalhas yet to be applied to real environmental water samples. Moreconsequences including entanglement, ingestion, vector of water-studies are required to verify this method for the microplasticsbornpollutantsandpossibletoxicitywereaddressed.samplesinfreshwaters.The sampling methods of volume-reduced method and bulkIn addition, a simple staining method was reported developedsample method weredescribed to provideresearcherswithsuit-(Shim etal.,2016).CommerciallyavailableNile Red (NR)wasableone(s)touse,Detailedillustrationonsamplepurificationandapplied to stain highly hydrophobic microplastics. The NR moleculeseparation was given,FTiR spectroscopy, Raman spectroscopy.is only fluorescent in the presence of a hydrophobic environment.pyrolysis-GC/MS, liquid chromatography and tagging method (forThe molecule specifically binds to the microplastics. As a result, thecounting) were described and their advantages and disadvantagesNRstainedmicroplasticparticlesarevisibleunderafluorescentwere discussed.microscope,which can easilybe countedThrough the literature research, we found that it was difficult toThe NR staining could be useful to determine the microplasticsuse a single method for quantitation and quantification of micro-abundance. The main drawback is that the NR also can stain theplastics from freshwater samples.The combination of differentnatural organic matters. Hence, pre-purification is required. Thetechniques can greatly improve our understanding of this newstaining method cannot be used alone unless the interferingenvironmental problem and morereliable datafor environmentalorganic matters are fully removed.risk assessment and preparation/enforcement of rules and regula-tions inthefuture.6.Challenges and recommendationsAcknowledgementsMicroplastics(polymers in nature)collected from freshwaterenvironment have undergone long-time exposed to variationA PhD scholarship from the National University of Singapore todegradation processes, including UV-induced photo-degradation,JL is appreciated. This research work was partially supported by thethermal degradation, mechanical action, and biological interactionsNational Research Foundation Prime Minister's Office, Singapore(e.g biodegradation). The surface morphology and properties suchunderitsCampusforResearchExcellenceandTechnologicalEn-as hydrophobicity would be greatly altered from original polymerterprise programme,as well as the Public Utilities Board, Singapore.compositions(Hidalgo-Ruz etal.,2012).Organisms and microorganisms can colonize microplastics,Referencesforming successional biofilms.Such processes in combination withpolymeradditiveswouldfurthercomplicatespectroscopicanalysis,Alencastro,D., 2012.Pollution due to plastics and microplastics in Lake Geneva andin theMediterraneanSea.Arch Sci.65.157-164becausespectralchangesresultfrombiofilm/additivesandrefer-Allsopp,M,Walters,A.,Santillo,D.Johnsto,P.206.PlasticDebrisintheWorld'sencespectraarelesssufficientfordegradationproducts.ItbecomesOceans.Greenpeace.International,Amsterdam,Netherlandsmore challenging for smaller particles, where the higher surface toAnderson,PJ,Warrack,S.,Langen,V.,Challis,J.K.,Hanson,M.L,Rennie,M.D.,2017.volume ratio makes the signals from the surface material moreMicroplastic contamination in Lake Winnipeg. Canada. Environ. Pollut 225.223-231significant. Thus it is rather difficult to differentiate the micro-Andrady,AL_ 2011.Microplastics in themarine environmentMar.Pollut. Bull.62plastics from natural polymers such as chitosan, cellulose, and81.1596-1605chitin.AngilllodionzoarmnoMaelloSanngoCau,A,Mastascusa,V.,Cau,A_Sacco,F,Canese,S.,2015.Distribution andIt is difficult to quantify and qualify microplastics from theIssessment of marine debris in the deep Tyrrhenian Sea (NW Mediterraneancomplex environmental samples using a single analytical methodSea.Italy)Mar.Pollut.Bull.92(1-2).149-159Thus, combination of multiple methods is preferred, which stronglyAntunes.J.CFriasJ..LMicaeloA.CSobralP,23.ResinpelletsfrombeachesfthePortuuesecoastand adsorbed persistentorganicpollutants.Estuar.Coastdependsonthesizesofmicroplastics.ShelfS130When the sizes of microplasticsfall in the range of<1mm andArthur.CBakerJBamford2009.In:ProceedingsoftheIntenationalResearchthe minimal cut-off size is tens of microns, the combination of theWorkshop on the Occurrence, Effects,and Fate of Microplastic Marine Debristember9-11,2008microscopicanalysis with the spectroscopicanalysisis highlyrec-Fauziah,S.,2017.Distribution and importanceofmicroplasticsAuta,H,Emenike,C.ommended. If the minimal cut-off size is in a range of a few mi-in themarine environment: a review of the sources, fate,effects, and potentialcrons, Raman spectroscopy is preferred as it is the only techniquesolutions.Environ.Int102,165-176.for such sizes. If the samples after digestion/pretreatment have lessBakir,A,'onnor,A,Rowland,J.Hendriks,AJ.Thompson,R.C,216.Relativimportance of microplastics as a pathway for the transfer of hydrophobicimpurity, thermal methods and liquidchromatographyareorganicchemicals tomarine life.Environ.Pollut219.56-65.recommendedBakir,A,Rowland,SJ.Thompson,R.C,2014.Enhanced desorptionof persistent Standardizedand robust methodsfor quantificationandorganic pollutants frommicroplastics under simulated physiological conditions

microplastics particles (up to 20 mg) was placed in a TGA crucible, followed by a heating process under inert conditions such as N2 up to 800 C. The resultant products that are sorbed on the extraction substrate can be tested by GC/MS for the determination of the composition. Only the results of polyethylene (PE) were published. In addition to gas chromatography, liquid chromatography was applied in quantification of microplastics (Hintersteiner et al., 2015; Elert et al., 2017). The method takes advantage of the different solubility of plastics. Appropriate solvents are used to dissolve different polymers. For instance, tetrahydrofurane and hexa- fluoroisopropanol could dissolve PS and polyethylene terephthalate (PET), respectively (Elert et al., 2017). After the preparation of polymer extracts, the samples can be analyzed by High Performance Liquid Chromatography (HPLC) coupled with size exclusion systems. Though unable to determine the sizes of plastic particles, this method reportedly has high re￾coveries and can quantify the microplastics. However, this method has yet to be applied to real environmental water samples. More studies are required to verify this method for the microplastics samples in fresh waters. In addition, a simple staining method was reported developed (Shim et al., 2016). Commercially available Nile Red (NR) was applied to stain highly hydrophobic microplastics. The NR molecule is only fluorescent in the presence of a hydrophobic environment. The molecule specifically binds to the microplastics. As a result, the NR stained microplastic particles are visible under a fluorescent microscope, which can easily be counted. The NR staining could be useful to determine the microplastics abundance. The main drawback is that the NR also can stain the natural organic matters. Hence, pre-purification is required. The staining method cannot be used alone unless the interfering organic matters are fully removed. 6. Challenges and recommendations Microplastics (polymers in nature) collected from freshwater environment have undergone long-time exposed to variation degradation processes, including UV-induced photo-degradation, thermal degradation, mechanical action, and biological interactions (e.g biodegradation). The surface morphology and properties such as hydrophobicity would be greatly altered from original polymer compositions (Hidalgo-Ruz et al., 2012). Organisms and microorganisms can colonize microplastics, forming successional biofilms. Such processes in combination with polymer additives would further complicate spectroscopic analysis, because spectral changes result from biofilm/additives and refer￾ence spectra are less sufficient for degradation products. It becomes more challenging for smaller particles, where the higher surface to volume ratio makes the signals from the surface material more significant. Thus it is rather difficult to differentiate the micro￾plastics from natural polymers such as chitosan, cellulose, and chitin. It is difficult to quantify and qualify microplastics from the complex environmental samples using a single analytical method. Thus, combination of multiple methods is preferred, which strongly depends on the sizes of microplastics. When the sizes of microplastics fall in the range of <1 mm and the minimal cut-off size is tens of microns, the combination of the microscopic analysis with the spectroscopic analysis is highly rec￾ommended. If the minimal cut-off size is in a range of a few mi￾crons, Raman spectroscopy is preferred as it is the only technique for such sizes. If the samples after digestion/pretreatment have less impurity, thermal methods and liquid chromatography are recommended. Standardized and robust methods for quantification and identification of microplastics should be first developed and veri- fied so that data from different researches can be more comparable and reliable. Key issues such as environmental risk assessment can then be conducted scientifically. Subsequently, rules and regula￾tions can be formulated and enforced. 7. Conclusions Microplastics have become one of emerging contaminants in the aquatic environment. The presence has been reported in many places around the world, and has caused great public concerns. However, there is still a lack of sufficient knowledge about micro￾plastics in freshwaters such as their health effects and fast monitoring. This review article summarized the current status of micro￾plastics contamination in freshwaters, including rivers, lakes, water treatment plants and drinking water. The potential environmental consequences including entanglement, ingestion, vector of water￾born pollutants and possible toxicity were addressed. The sampling methods of volume-reduced method and bulk sample method were described to provide researchers with suit￾able one(s) to use. Detailed illustration on sample purification and separation was given. FTIR spectroscopy, Raman spectroscopy, pyrolysis-GC/MS, liquid chromatography and tagging method (for counting) were described and their advantages and disadvantages were discussed. Through the literature research, we found that it was difficult to use a single method for quantitation and quantification of micro￾plastics from freshwater samples. The combination of different techniques can greatly improve our understanding of this new environmental problem and more reliable data for environmental risk assessment and preparation/enforcement of rules and regula￾tions in the future. Acknowledgements A PhD scholarship from the National University of Singapore to JL is appreciated. This research work was partially supported by the National Research Foundation Prime Minister's Office, Singapore under its Campus for Research Excellence and Technological En￾terprise programme, as well as the Public Utilities Board, Singapore. References Alencastro, D., 2012. Pollution due to plastics and microplastics in Lake Geneva and in the Mediterranean Sea. Arch. Sci. 65, 157e164. Allsopp, M., Walters, A., Santillo, D., Johnsto, P., 2006. Plastic Debris in the World's Oceans. Greenpeace. International, Amsterdam, Netherlands. Anderson, P.J., Warrack, S., Langen, V., Challis, J.K., Hanson, M.L., Rennie, M.D., 2017. Microplastic contamination in Lake Winnipeg, Canada. Environ. Pollut. 225, 223e231. Andrady, A.L., 2011. Microplastics in the marine environment. Mar. Pollut. Bull. 62 (8), 1596e1605. Angiolillo, M., di Lorenzo, B., Farcomeni, A., Bo, M., Bavestrello, G., Santangelo, G., Cau, A., Mastascusa, V., Cau, A., Sacco, F., Canese, S., 2015. Distribution and assessment of marine debris in the deep Tyrrhenian Sea (NW Mediterranean Sea, Italy). Mar. Pollut. Bull. 92 (1e2), 149e159. Antunes, J.C., Frias, J.G.L., Micaelo, A.C., Sobral, P., 2013. Resin pellets from beaches of the Portuguese coast and adsorbed persistent organic pollutants. Estuar. Coast. Shelf. S 130. Arthur, C., Baker, J., Bamford, H., 2009. In: Proceedings of the International Research Workshop on the Occurrence, Effects, and Fate of Microplastic Marine Debris, September 9-11, 2008. Auta, H., Emenike, C., Fauziah, S., 2017. Distribution and importance of microplastics in the marine environment: a review of the sources, fate, effects, and potential solutions. Environ. Int. 102, 165e176. Bakir, A., O'Connor, I.A., Rowland, S.J., Hendriks, A.J., Thompson, R.C., 2016. Relative importance of microplastics as a pathway for the transfer of hydrophobic organic chemicals to marine life. Environ. Pollut. 219, 56e65. Bakir, A., Rowland, S.J., Thompson, R.C., 2014. Enhanced desorption of persistent organic pollutants from microplastics under simulated physiological conditions. J. Li et al. / Water Research 137 (2018) 362e374 371