Polyester Textiles as a Source of Microplastics ... - ACS Publications

May 24, 2017 - Empa, Swiss Federal Laboratories for Materials Science and Technology, Technology and Society Laboratory, Lerchenfeldstrasse 5,. 9014 S...
0 downloads 11 Views 4MB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Synthetic Textiles as a Source of Microplastics from Households: A Mechanistic Study to Understand Microfiber Release During Washing Edgar Hernandez, Bernd Nowack, and Denise M. Mitrano Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 24 May 2017 Downloaded from http://pubs.acs.org on May 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

Environmental Science & Technology

1

Polyester Textiles as a Source of Microplastics from Households: A

2

Mechanistic Study to Understand Microfiber Release During Washing

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Edgar Hernandez1, Bernd Nowack1 and Denise M. Mitrano*1,2

1Empa,

Swiss Federal Laboratories for Materials Science and Technology, Technology and Society

Laboratory, Lerchenfeldstrasse 5, 9014 St. Gallen, Switzerland

2Eawag,

Swiss Federal Institute for Aquatic Science and Technology, Process Engineering,

Überlandstrasse 133, 8600 Dübendorf, Switzerland

*Corresponding Author: Dr. Denise M. Mitrano Eawag – Swiss Federal Institute of Aquatic Science and Technology Process Engineering Überlandstrasse 133 8600 Dübendorf Switzerland [email protected]

25 26

ACS Paragon Plus Environment

Environmental Science & Technology

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

Abstract Microplastic fibers make up a large proportion of microplastics found in the environment, especially in urban areas. There is good reason to consider synthetic textiles a major source of microplastic fibers and it will not diminish since the use of synthetic fabrics, especially polyester, continues to increase. In this study we provide quantitative data regarding the size and mass of microplastic fibers released from synthetic (polyester) textiles during simulated home washing under controlled laboratory conditions. Consideration of fabric structure, washing conditions (use of detergents, temperature, wash duration, sequential washings) allowed us to study the propensity of fiber shedding in a mechanistic way. Thousands of individual fibers were measured (number, length) from each wash solution to provide a robust data set on which to draw conclusions. Among all the variables tested, the use of detergent appeared to affect the total mass of fibers released the most, yet the detergent composition (liquid or powder) or overdosing of detergent did not significantly influence microplastic release. Despite different release quantities due to the addition of a surfactant (approximately 0.025 and 0.1 mg fibers/g textile washed, without and with detergent, respectively), the overall microplastic fiber length profile remained similar regardless of wash condition or fabric structure, with the vast majority of fibers ranging between 100 µm and 800 µm in length irrespective of wash cycle number. This indicates that the fiber staple length and/or debris encapsulated inside the fabric from the yarn spinning could be directly responsible for releasing stray fibers. This study serves as a first look towards understanding the physical properties of the textile itself to better understand the mechanisms of fiber shedding in the context of microplastic fiber release into laundry wash water.

50

ACS Paragon Plus Environment

Page 2 of 27

Page 3 of 27

Environmental Science & Technology

51

Introduction

52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

Since the mass production of plastics began in the 1940’s, manufacturing techniques have been optimized, resulting in a plethora of lightweight, durable, persistent and corrosion resistant plastic varieties.1 These attributes have led to the extensive use of plastics in near inexhaustible applications.2 However, some of the same attributes that make plastics an attractive material, such as their durability, can also be problematic and cause environmental concern. While macroplastic objects have been the focus of environmental concern for some time, it is only since the turn of the century that tiny plastic fragments, fibers and granules, collectively termed microplastics (polymer fragments < 5 mm in size), have also been increasingly considered a pollutant. Microplastics in freshwater ecosystems have been detected3-5 and increased microplastic inputs have been correlated with population density, land use, and the level of sewage treatment6,7. Therefore, the origins of many microplastics are industrial and domestic water treatment flows8-11. Wastewater treatment plants (WWTP) and other industrial facilities are likely important hubs and transport pathways of microplastics into the environment; microplastics in high concentrations are consistently reported downstream of the discharge points.12-15 For domestic sewers and WWTP there are two main microplastic inputs: primary microplastic beads made of polyethylene, polypropylene, and polystyrene particles found in cleaning and cosmetics/toothpaste/personal care products that are used as scrubbers and secondary microplastic fibers/filaments that are breakdown products from the washing of polyester or other synthetic textiles. Secondary microplastics arising as fibers from washing clothes are mainly made of polyester, acrylic and polyamide and can be found in WWTP effluent.16 Fibers similar to those observed in household sewage effluent have been found to be dominant at sewage disposal sites and exhibit long residence times. Indeed, it has been reported that wastewater treatment facilities are unable to capture all microplastics and this contributes to their presence in freshwater bodies: there are higher microplastic densities downstream of WWTP than at reference points upstream of the plant.17,18 Additionally, synthetic fibers appear to be removed in the WWTP to a lesser extent than their natural counterparts.19 Significant legislative efforts have been underway to eliminate primary microplastic scrubbers from consumer goods such as cosmetics and body washes.20,21 Therefore, textile fibers will likely be one of the main microplastic sources to consider in domestic drainages in the future. While consumers can choose to buy clothing made from natural materials, synthetic fibers are entrenched in the clothing industry, either as pure polymers or as natural/polymer blends, and the global production of synthetic fibers (especially polyester) has surpassed the demand for natural alternatives.22 While the general occurrence of microplastic fibers from textiles is widely noted, there is little quantitative data to determine the microplastic fiber length or mass released when washing fabrics in the home. A point that is specifically missing are systematic studies under controlled conditions to

ACS Paragon Plus Environment

Environmental Science & Technology

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111

better understand the mechanisms behind fiber shedding through use, especially considering variables such as different fiber production (extruded filaments or wound staples), fabric structure (woven, knitted or non-woven), washing conditions (temperature, detergent/surfactant, length of washing, multiple washes) and specific size distributions and masses of fibers shed during this process. A few studies have provided some initial figures as to the mass of microplastic released from textiles and considered a few of these factors individually.16,23-26 In the study with the most variables, Napper et al. laundered three different textiles under a number of different conditions (wash temperature, detergents, multiple washes) and fibers were collected directly from the washing machine outlet and weighed after drying on a clean filter to determine total microplastic mass released.23 A few fibers were individually imaged to suggest average fiber dimensions released for each textile variant. Likewise, Hartline et al. washed a series of jackets (new or aged) and relied on serial filtration to determine the mass and size range of fibers released from the textiles.24 However, these reports had a few common factors which obscured the mechanisms of release including, 1) the use of commercially available textiles and using a household washing machine instead of textiles with known properties, standardized washing procedures and laboratory controls, 2) measuring only a small sample set to suggest fiber size (distribution) released, 3) collecting large volumes of effluent and using simplified measurements leading to rough calculations of microplastic fiber mass release and 4) limited or no discussion on the mechanistic properties of microplastic fiber release. Here we aim to improve on these studies by 1) using standardized textiles with known fiber and knit characteristics, 2) performing our tests under more controlled conditions using standardized protocols (which include extensive QA/QC for any new experimental and analytical tasks), 3) developing measurement methods which capture a more detailed and accurate fiber release profile in terms of both fiber length and total mass, and 4) perform a matrix of experiments which can suggest mechanisms for fiber shedding specifically related to textile construction and washing conditions when they are laundered in domestic washing conditions.

112 113

Materials and Methods

114

Textiles and detergents

115 116 117 118 119

Two black colored synthetic textiles were obtained from the textile archive at Empa, representing fabrics that are in commercial production for undershirts meant as a first layer of clothing.27 Interlock fabric was knitted from 100% polyester (PES) yarns wheras the plain single knit jersey fabric was knitted with these same polyester yarns but with a 2% spandex plating. The fiber diameter and the density of both knit types was on average 20 μm and 950 kg/m3 , respectively.

ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27

Environmental Science & Technology

120 121

Yarns were constructed by spun fiber staples. Images of the difference in fabric knit can be seen in

122 123 124 125 126 127

Two washing solutions were obtained, the same as used in Mitrano et. al.28,29 Both were “grocery

128

Pretreatment of textiles

129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144

Textile samples were cut to the same size, tailored so no raw edges were exposed, weighed and

Figure S1. The pretreatment of the textiles before washing is described in the Supporting Information.

store brand” detergents from a Swiss store, are commercially available and intended for use in private homes. One liquid detergent and one powder detergent, both of the “All Purpose” variety were chosen. Detergent compositions are given in Table S1. The detergents are distinguished by the absence or presence of an oxidizer, the presence or absence of particulate matter and the differing pH (9.2 and 10.1 for liquid and powder detergents, respectively).

underwent a pre-wash (rinse) step before experimental treatments. Fabrics were cut to a size of 30 cm x 10 cm and edges were folded over twice and sewn with a white thread to not interfere with fiber counting if thread was lost during the washing steps. A finished tailored single swatch of fabric measured approxmately 25 cm x 7 cm with a surface area of 175 cm2. A single swatch of fabric weighed approximately 7 g for both interlock and jersey variants. Before fabrics were washed under their given experimental conditions for the first time, they underwent a short (5 min) rinse cycle in the laboratory washing machine with only DI water which ensured that 1) all loosely bound polymeric material from fabrication, fabric softening and cutting were removed and 2) all stray fibers from other materials were removed. This pre-wash step was essential for measuring fibers that were only released from the textile since without it a significant amount of material was collected from the rinse water that was not released from the textile itself, identified by differently colored fibers and debris (see Figure S2). Additionally, scanning electron microscopy (SEM) was performed on the textile yarns and the fibers released from the textile after washing (Figure S4, with accompanying analyisis details). A schematic of fabric processing and experimental wash procedures/conditions is shown in Figure 1.

ACS Paragon Plus Environment

Environmental Science & Technology

145 146

Figure 1: Schematic of experimental design and wash water analysis.

147 148

Standardized washing procedure

149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164

The washing procedure was carried out as described in our previously published work and based on the ISO Standard 105-C06:2010, with some modifications for these specific solutions.30-33 A WashtexP Roaches laboratory washing machine was operated at 40 ± 2 rpm with stainless steel vessels. Temperature was controlled by a thermostat at 40°C ± 2°C. A 5 min rinse cycle in DI H2O in the laboratory washing machine, where rinse water was disposed of, preceded the experimental washing which lasted for 45 min. For each washing 7 ± 0.37 grams of fabric was placed inside the stainlesssteel vessel with 200 ml DI water and, depending on the experimental conditions, liquid detergent or powder detergent (4 g detergent/L). In order to simulate the stress produced during a normal washing cycle, 10 stainless steel balls (Ø 6 mm) were also introduced in the samples to provide additional mechanical wear during the wash cycle. Each washing experiment was conducted in triplicate and results presented here are indicative of the average. Fabric swatches were completely air dried (fabric swatches hung on a line inside a protective box overnight) before the next wash cycle and stored individually in clear plastic bags after drying. Five wash cycles for each standard wash condition (DI H20, liquid detergent, powder detergent) were performed in triplicate, to determine how fiber release may change in terms of fiber length of quantity with increasing wash cycles.

ACS Paragon Plus Environment

Page 6 of 27

Page 7 of 27

Environmental Science & Technology

165 166

Non-standardized washing procedures – mechanistic understanding of microplastic

167

fiber release

168 169 170 171 172 173

While the standardized methods outlined above allowed us to estimate likely releases of microplastic

174

Extension of wash cycle duration: The extended wash durations implemented were 1, 2, 4 and 8 hours.

175 176 177

The same fabric size and other wash conditions remained the same as in the standardized washing

178 179 180 181 182 183 184 185

Addition of surfactant: In the grocery store liquid detergent, a combination of four different ionic and

186

Variation of wash temperature: Jersey fabric was washed in DI H2O and liquid detergents under the

187

standard protocol, with the exception of setting the temperature to 25, 40, 60 and 80 degrees Celsius.

fibers during typical household washing conditions, a parallel line of research was conducted to better understand the mechanisms that dictated the size and quantities of fiber released during each wash cycle. Therefore, we examined several aspects of the wash cycle that may influence release and exaggerated the wash conditions beyond that of typical household use to tease out factors that dominated microplastic fiber formation.

procedure. In this portion of the study, only jersey fabric was used for two wash water variants, DI H2O and liquid detergent.

non-ionic surfactants is present with a total of approximately 18% by weight. The predominant type is a LAS surfactant, and so for the mechanistic studies a linear alklylbenzenesufonic acid (97%) was purchased from Alfa Aesar for dilution in DI H2O. The total weight of all surfactant varieties in the liquid detergent is approximately 0.75 g/L and so an equivalent amount of LAS detergent was used for direct comparison. Two and three times the amount of surfactant was also tested (i.e. overdose of the recommended detergent concentration), with 1.5 g/L and 2.25 g/L added to DI H2O for the standard washing procedure.

188 189

Wash water collection, filtration and filter imaging

190 191 192 193 194 195

After washing, the fabric was removed from the vial and gently pressed to squeeze excess liquid back into the wash container. The water that remained inside the vial was then transferred to a filtration system using a 10 ml pipette, stirring the wash water continuously to keep fibers suspended homogeneously in solution. The filtration system consisted of a vacuum pump pulling the wash water through a Whatman 0.45 μm pore sized filter (Ø 4.5 cm). The filters were then left to dry for 24 hours inside an aluminum protective case that reduced possible airborne residue and contaminants. The

ACS Paragon Plus Environment

Environmental Science & Technology

196 197 198 199 200 201 202 203

white color of the filter paper contrasted to the black color of the fibers making identification of the

204 205 206 207 208

Depending on the wash condition, the volume of wash water filtrated was either 50, 100 or 200 ml.

fibers easier. A Keyence Digital Microscope System with a VHX Digital Microscope Multi Scan Lens was used to image the filters. The image captured with the microscope was a high-quality resolution image of 96 dpi, which was a large composite image digitally stitched together from multiple snapshots. Typically, 30 individual images were needed to image the whole filter and the individual images were automatically aligned by the microscope computer software to give one single image of the filter as a final result. An appropriate scale bar was placed on the composite image for later processing with the open source software ImageJ.

This variation depended on the quantity of fibers that were seen on the filter during the filtration step to ensure an optimal quantity of fibers were deposited for analysis. Having too many fibers would create extensive overlapping of individual fibers and thus 1) make fiber identification more difficult manually and 2) cause fiber mass underestimation using image analysis.

209 210

Filter Image Processing – Fiber length distribution and mass calculations

211 212 213 214 215 216 217 218 219 220 221

Three metrics were obtained from the analysis, 1) fiber number, 2) fiber length distribution, and 3)

222 223 224 225 226

For all filters, a binary image was created of the filter where pixels in the image that contained a fiber

227 228

In order to facilitate the analysis of additional samples, a correlation between filter coverage and

fiber mass. In all cases, images were uploaded in the .tiff format into the ImageJ software with their native pixel resolution of 96 dpi. To obtain the fiber number and fiber length distribution, all individual fibers were counted manually over the entire filter for all standard wash condition filters. Using a digital tablet and stylus, lines were drawn on top of the fibers (Figure S5). The scale was set using the imbedded scale bar on the image and applied to lines drawn on the fibers. Lengths of individual fibers were exported to excel for further data processing. When all fibers were counted on the filter, the known fiber diameter and density could be used to calculate the mass of fibers on the filter. The fiber length distribution was converted to a volume distribution (using the known fiber diameter) and subsequently to a mass distribution when multiplied by the fiber density of 0.95 g/cm3. By integrating this mass distribution, the mass of fibers was calculated.

were shown as black and filter pixels were white. After uploading the filter image into ImageJ, it was converted to 8 bit and the threshold function was adjusted to create the binary image. By setting the filter diameter and overlaying a circular template the size of the exposed filter area, a percentage of black to white pixels on the filter was calculated.

fiber weight was developed using fiber weight data derived manually from counting each individual

ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27

229 230 231 232 233 234 235

Environmental Science & Technology

fiber on the filter and their respective filter coverage data obtained from the binary image analysis. This correlation was then used for the analysis of additional wash procedures without the need to measure fibers individually, since by using the percentage coverage of the filter alone, the correlation would yield an acceptable approximated mass of fibers on the filter. The construction of this correlation was done by taking several analyzed filters and plotting the mass calculated from the length distribution against the filter coverage from the binary image analysis. The procedure to derive the correlation is described in detail in the Supporting Information.

236 237

Quality Assurance/Quality Control Experiments

238 239 240 241 242 243 244 245 246 247 248

Reproducible fiber quantification (in terms of both fiber size and fiber number) was of utmost importance for the success of the method and so QA/QC experiments were performed in order to test the reliability of the wash water filtering, filter imaging and data processing sequences. For a select sample set (one sample each of DI water, liquid detergent and powder detergent for both jersey and interlock fabrics washed under the standard conditions), triplicate aliquots of the wash water were filtered through separate filters, imaged and analyzed for both fiber size distribution and released fiber mass. While experimental triplicates may have a certain amount of variability due to differing amounts of fiber release from the textile swatch, because these QA/QC samples originated from the same wash vial, both the fiber size distribution and fiber mass should be the same. Any variation in the QA/QC sample set, therefore, depicts uncertainty arising from the sample processing and/or image processing only.

249 250

Statistics

251 252 253 254 255 256 257 258 259 260 261

In order to differentiate between the amount of microplastic fibers released between washing procedures and knit types, pair-wise comparisons were made by ANOVA tests performed between the experimental groups. Three groups were tested for statistical differences including 1) knit variant, 2) washing condition and 3) the non-standardized washing procedures. For the knit variants, two analyses were preformed; firstly, between the different knits under the same washing conditions and secondly between washes of a single wash group (i.e. 1st to 5th wash of a given knit under a given experimental condition). For the second group, to determine differences in the washing conditions, the knit type was fixed and we compared the released microplastic fiber concentrations in DI water, liquid and powder detergents. In these cases, the focus was mainly on determining the holistic differences across the entire group and not in assessing the individual wash cycle number. The final analysis was of the non-standardized washing procedures, which consisted of the wash cycle

ACS Paragon Plus Environment

Environmental Science & Technology

262 263

duration and wash water temperature effects, analyzed as single groups. In any given comparison, the groups were considered statistically different when p values were less than 0.05.

264 265

Results and Discussion

266

Quality Assurance and Quality Control Studies

267 268 269 270 271 272 273 274 275 276 277

We explored two facets of quality assurance and quality control in these studies; the transfer of fibers

278 279 280 281

The minimum measurable length of a fiber with this analytical setup was determined to be 40 µm.

282 283 284 285 286 287

Method blanks of DI H2O, liquid detergent and powder detergent were analyzed for presence of

suspended in the wash water onto the filter and fiber analysis on the filter, both in terms of size distribution and total mass. By taking subsets from one wash solution and processing them individually, we determined that the deposition of fibers onto the filter was reproducible and secondly, we verified that the analytical techniques we used were consistent in providing similar released fiber size and mass metrics (Figure S6 and Figure S7, respectively). Therefore, we determined that 1) using our experimental setup we could subsample a portion of the wash water and it would be representative of the total fiber release and 2) the entire analytical work-flow (filter imaging, individual fiber counting for fiber size distribution and “simplified” binary filter analysis for measuring released fiber mass) was appropriate to precisely characterize the fiber release profile under a number of different washing conditions.

This corresponds to the minimum number of pixels that could definitively be considered a fiber when manually counting or when using the filter coverage/binary analysis method, which was between 2 and 5 pixels depending on the saturation of the particular image.

fibers but too few (< 3 per filter) were detected to image the entire filter in our system. While the filter started to build up a small cake when filtering samples washed in the powder detergent (especially as the solution cooled), there were no colored fibers or particles of any type noted in any of the method blanks. Since we only measured black fibers on the filters, any white powder or residue that was left on the filters did not interfere with the analysis.

288 289

Fiber Area Coverage Correlation Curve

290 291 292

Counting the number and length of all the individual fibers on the wash water filter achieved the most accurate metrics for fiber length distribution and calculated mass possible. However, as the number of fibers on each filter often numbered into the thousands, this method was very time

ACS Paragon Plus Environment

Page 10 of 27

Page 11 of 27

Environmental Science & Technology

293 294 295 296 297 298 299 300 301 302 303 304

consuming. While we investigated some options to automatically measure fibers through image

305 306

When using a correlation function between mass and percentage filter coverage, some accuracy is

307

simplified method remains in an acceptable range of ±10% of the total fiber mass calculated. Fiber

308 309 310 311 312 313 314

length did not affect the overall accuracy of this simplification, but the extent to which fibers

analysis, since the fibers overlapped we were unable to convolute the various strands automatically. In order to reduce filter analysis time for many of the experimental variables, a correlation was made between the released fiber mass (as determined from measuring all individual fibers on the filter manually) and the percent coverage of fibers on the filter (as measured by binary image processing). Filter analysis of fibers released from both jersey and interlock fabrics washed in DI H2O, liquid detergent and powder detergent (i.e. all standard washing variants) were plotted together on a single graph, which shows that a close correlation could be made between filter coverage and fiber mass (Figure 2). Therefore, in subsequent filter analysis of wash water from multiple washes or mechanistic studies we were able to estimate the mass of fibers released based on fiber filter coverage alone. For additional information regarding the correlation function, we refer to the Supplemental Information.

inevitably lost, but nevertheless the final result when comparing the precise method and the

overlapped had the potential to underestimate mass release since the area where the fibers cross would only be accounted for once in the calculation. However, this underestimation is limited, with an average underestimation of approximately 10% (see Table S2). This was determined by measuring the total number of pixels which were associated with individual fibers measured manually and comparing that to the more automated percent coverage of the filter method for a randomly selected sample set.

ACS Paragon Plus Environment

Environmental Science & Technology

Page 12 of 27

0.35 Fitted Function

Interlock

Jersey

Mass (mg)

0.3 0.25 0.2 0.15 0.1 0.05 0 0

0.5

1

1.5

2

Percent Coverage (%) 315 316 317 318 319 320

Figure 2: Correlation between the mass of fibers released on one filter (where all fibers were measured individually over the entire filter) and the percent coverage of fibers on the filter (as measured by binary image analysis). The same correlation fit was obtained for the filters analyzed for both interlock (brown circles) and jersey (yellow circles) fabric weave variants under a number of different wash conditions including DI H2O, liquid detergent and powder detergent.

321 322

Microplastic Fiber Release from Standardized Washing Procedure – Simulation of

323

Home Washing

324

Microplastic Fiber Length Distribution Released

325 326 327 328 329 330 331 332 333 334 335

The size distribution of microplastic fibers shed from both textiles washed under all standard conditions (DI H2O, liquid detergent and powder detergent) and for multiple washes (wash one through wash five) was determined by manually measuring every fiber collected when filtering the wash water (normalized fiber length distribution histograms Figure 3, box plot summary of data Figure 4). Depending on the sample, this meant the number of fibers individually measured on a filter typically ranged between 500 and 2000, with some cases upwards of 3000 individual fibers (see Table S3). The size distributions between triplicate experiments were averaged when presenting these results, but there was overall good agreement of fiber length distributions within the experimental sets. While the first wash of textiles washed with detergents appeared to release the highest number of small sized fibers, there was otherwise relatively little change in fiber length distribution with subsequent washes across all washing conditions. Additionally, there was little

ACS Paragon Plus Environment

Page 13 of 27

Environmental Science & Technology

336 337 338

variability between the fiber release profiles for jersey and interlock textiles. Regardless of textile

339 340 341 342 343 344 345 346 347 348 349 350

One notable feature recognized in each of the filters measured was the presence of a few fibers that

knit or wash condition, the vast majority of fibers were under 1 mm in length with a slight shift of the distribution towards the smaller fiber sizes.

were significantly longer than the rest of the particle size distribution (i.e. > 1 mm), but these outliers accounted for only 0.7% of the entire fiber population measured. Depending on the total number of fibers released in a given sample, this equates to 2 – 5% of the mass of plastic that is in the > 1 mm size fraction. While these long fibers may be the least numerous, under different or less careful analysis conditions one may only observe these easier to spot fibers and thus misrepresent the characteristics of microplastic fiber release. For example, in Napper et. al the mean fiber length was based on data from only 10 individual fibers, where the authors found the average (mean) fiber length between 4.99 and 5.44 mm.23 While the type of textile and fibers therein will play a crucial role in average fiber length, the characterization of so few fibers, especially when they are seemingly so uncharacteristically long amongst the total dataset, may lead to overestimations of fiber release, especially when this metric is converted to mass.

351

ACS Paragon Plus Environment

Environmental Science & Technology

Page 14 of 27

352 12

N orm a lized Fibe r Le ngth D istrib u tio n (% )

10 Wash Wash Wash Wash Wash

8 6

1 2 3 4 5

4 2

8

Wash Wash Wash Wash Wash

6 4

2

12 00

1 50 0

1 40 0

15 00

4

1 50 0

6

14 00

8

1 2 3 4 5

1 40 0

Wash Wash Wash Wash Wash

13 00

N o rm a lized F ib er Le n gth D is trib u tio n (% )

10

1 30 0

12

Je rse y, P o w d e r D e te rge n t

11 00

F ib e r Len gth (μ m )

10 00

900

800

700

600

500

400

300

200

0

150 0

140 0

130 0

120 0

110 0

100 0

900

800

700

600

500

400

300

200

0

100

0

12

In te rlo ck, P o w d e r D e te rge nt

10

Wash Wash Wash Wash

8

1 2 3 4

6 4 2

Fib e r Le n gth (μm )

1 20 0

1 10 0

1 00 0

90 0

80 0

70 0

60 0

50 0

40 0

30 0

20 0

1500

1400

1300

1200

1100

1000

900

800

700

600

500

400

300

200

0

100

0

0

0

10 0

N o rm a liz ed Fib er Len gth D istrib u tio n (% )

1 2 3 4 5

2

Fib e r Le n gth (μ m )

354 355 356 357 358 359 360

1 30 0

In te rlo ck, Liq u id D e te rge n t

10

100

N o rm a lized Fib er Len gth D istrib u tio n (% )

12

Je rse y, Liq u id D e te rge n t

0

353

1 20 0

Fib er Len gth (μ m )

Fib er Len gth (μ m ) 12

1 10 0

0

1 50 0

1 40 0

1 30 0

1 20 0

1 10 0

900

1 00 0

800

700

600

500

400

300

200

0

100

0

2

900

2

1 2 3 4 5

4

1 00 0

Wash 5

800

4

6

700

Wash 4

600

Wash 3

500

6

Wash Wash Wash Wash Wash

8

400

Wash 2

300

8

In te rlo c k, D I W a te r

10

0

Wash 1

200

10

100

Je rse y, D I W a te r N o rm a liz ed Fib er Le n gth D is trib u tio n (% )

N o rm a lized F ib er Len gth D istrib u tio n (% )

12

Fib er Le n gth (μm )

Figure 3: Microplastic fiber size distribution released from washing 100% PES in various solutions including DI H2O, liquid detergent and powder detergent over sequential wash cycles. All fibers were counted on the entire filter resulting in the fiber size distribution histogram shown here, which has been scaled to the normalized number of particles measured across all datasets for easier comparison between wash conditions with varying fiber release quantities. Values presented here were averaged across three experimental replicates. The grey dashed line indicates the method detection limit of 40 µm. This data corresponds to Figure 4 presented as a box plot in the main text.

361

ACS Paragon Plus Environment

Page 15 of 27

Environmental Science & Technology

362 In te rlo ck , D I W a ter Fib er Le n gth (μ m )

Fib er Len gth (μ m )

Jer s ey, D I W a ter 1 000

1 1

2

3

4

10 0 0

1

5

1

2

W a s h C ycle

4

5

F ib e r Len gth (μ m )

Fib e r Len gth (μ m )

1 3

10 00

1 1

2

363 364 365 366 367 368 369 370 371

100 0

1 3

4

5

In ter lo ck , P o w d e r D e ter ge n t Fib e r Le n gth (μ m )

Fib e r Le n gth (μ m )

Jer s ey, P o w d er D e tergen t

2

3

W a sh C ycle

W a s h C ycle

1

5

In terlo ck , Liq u id D etergen t

1 00 0

2

4

W a s h C ycle

Jers ey, Liq u id D e te rge n t

1

3

4

5

1 0 00

1 1

W a s h C ycle

2

3

4

5

W a s h C ycle

Figure 4: Microplastic fiber size distribution released from washing 100% PES in various solutions including DI H2O (blue), liquid detergent (green) and powder detergent (yellow) over sequential wash cycles. Two knits of the same yarns were investigated; jersey (light color variant, left column) and interlock (dark color variant, right column). 25% quantiles appear within the box plot with average particle size indicated by black line therein. Whiskers represent 95% of particle size distribution. Outliers, signified by the red dots, represent only 0.7% of the total fiber distribution. Values presented here were averaged across three experimental replicates. For corresponding full MP fiber size distribution histograms of this data, see Figure 3.

372 373

Microplastic Fiber Mass Released

374 375 376 377

Despite similar fiber length release profiles, the mass of microplastic fibers shed during laundering was dependent on the wash solution (Figure 5). The lowest masses recorded were those textiles washed in DI H2O only. In contrast to other published studies,23,25 in our tests there did not appear to be a reduction in the amount of fibers released over subsequent washes and instead there was a

ACS Paragon Plus Environment

Environmental Science & Technology

378 379 380 381 382

steady average release of approximately 0.025 mg microplastic fibers/g textile regardless of wash

383 384 385 386 387 388 389

Washing with either liquid or powder detergent variants released a larger mass of microplastic fibers

390 391 392 393 394 395 396 397 398 399 400

The variability between triplicate experiments of one laundering variant was not predictable. In

cycle number. When taken as whole considering the entire wash set (1 through 5 washes), there was not a significant difference between the mass of fibers released from the jersey and interlock knit variants, nor was there any statistically important relationship between individual samples of a given wash cycle number.

into the wash water than when laundering with DI H2O alone (two-way ANOVA analysis p < 0.05), but there were no statistically significant differences between the mass released when washing with different detergent formulations or between the fabric knit types. As with the DI H2O wash samples, there did not appear to be a clear trend in the mass of microplastic fibers released over subsequent washes. An average of approximately 0.1 mg microplastic fiber/g textile was shed during each wash, which is approximately four times as much as when washing with only DI H2O.

some instances, there was a very small standard deviation and others the triplicate experiments showed rather different masses of microplastic fibers released. The fluctuation in these releases did not appear to be influenced by the fabric knit, the wash solution, the wash order or a specific swatch of textile (i.e. one textile swatch was not continually releasing more or less fibers than the other two in the experimental set). However, this variance in the standard deviation between triplicate experiments is not simply due to sample or analytical processing error, since QA/QC experiments of select wash waters proved that there was a high precision when measuring a single sample (Section on quality assurance and quality control in the results). Therefore, the variance observed within an experimental triplicate set is simply the variation of the possible mass released. We concluded there is some implicit variance in the washing from an unidentified source.

ACS Paragon Plus Environment

Page 16 of 27

Page 17 of 27

Environmental Science & Technology

m g Fiber/g Textile

DI Water 0.25

Jersey Interlock

0.2 0.15 0.1 0.05 0 1

2

3

4

5

Wash Cycle

mg Fiber/ g Textile

Liquid Detergent 0.25 Jersey Interlock

0.2 0.15 0.1 0.05 0 1

2

3

4

5

Wash Cycle

m g Fiber/g Textile

Powder Detergent 0.25 Jersey Interlock

0.2 0.15 0.1 0.05 0 1

401

2

3

4

5

Wash Cycle

402 403 404 405

Figure 5: Mass of microplastic fibers released from washing 100% PES textiles (jersey and interlock weaves, light and dark variants, respectively) in various solutions including DI H2O (blue), liquid detergent (green) and powder detergent (yellow) over sequential wash cycles. Error bars indicate triplicate experiments.

406

Mechanistic Experiments to Understand Microplastic Fiber Releases

407 408 409 410 411

There was no statistical difference in the mass of microplastic fibers released when textile swatches were laundered at different temperatures including 25 oC, 40 oC, 60 oC and 80 oC (Figure 6A). This was the case when fabrics were washed in either DI H2O or liquid detergent. The extended washing time (Figure 6B) also did not have a significant impact on the fiber release profile. An ANOVA analysis confirmed that there were no statistical differences in mass of fibers released for either

ACS Paragon Plus Environment

Environmental Science & Technology

412 413 414 415 416 417 418

laundering at different temperatures or increasing the duration of the wash cycle. In contrast, the

419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434

One of our initial hypotheses was that mechanical stress from the laundering processes was

435 436 437 438 439 440 441 442 443 444 445 446 447

Nevertheless, we wanted to determine if part of our experimental regime was responsible for

addition of surfactant to the wash solution always mobilized more microplastic fibers from the textile surface than when washed in DI H2O alone. However, adding more surfactant did not necessarily equate with a higher quantity of fibers released (Figure 6C). When the equivalent concentration of surfactant is added to the wash water as the grocery store liquid detergent (0.75 g/L surfactant), a similar mass of fibers is released under the standard washing conditions. However, when the concentration of the surfactant alone is doubled or tripled, no additional fiber release is observed.

responsible for the breakage (or loosing of fibers from the yarn) and subsequent shedding of fibers into the wash solution. Therefore, if this was true, then the cumulative release of fibers over time should be “additive” in the sense that a two-hour wash cycle would “produce” and release twice as many microplastic fibers as the one hour wash cycle, that the four-hour wash cycle would produce and release twice as many fibers as the two-hour wash cycle, and so on. However, this is not what we observed and instead we found only a “fixed” amount of fibers released regardless of the additional mechanical stress. Another hypothesis was that the surfactant was responsible for mobilizing the broken fibers from the surface of the fabric and into the wash solution. In this case, we did find a positive correlation between the mass of microplastic fibers in the wash water and the presence of surfactant in solution. In all cases where surfactant was used (liquid detergent, powder detergent, or laboratory surfactant solution), there were significantly higher quantities of microplastic fibers in the wash water solution than when using DI H2O alone. Moreover, similar amounts of fibers release were observed between all of these surfactant sets, indicating that, in our case, the exact nature of the surfactant does not contribute to the quantity of microplastic fibers mobilized from the textile surface.

creating a “maximum” amount of fibers that could be released, since it appeared that while some aspects could decrease the number of fibers observed in the wash water, there was seemingly an upper limit which was not surpassed by adjusting laundering parameters according to the two hypotheses above. Since the microplastic fibers should not “saturate” the wash water in a way that a chemical textile additive may, we conducted a series of washes to further investigate the link between mechanical stress, added surfactant and potential drying effects to the fabrics. Here we washed three groups of fabric swatches all under standard conditions with liquid detergent; 1) two 45 min wash cycles with fabrics air dried at ambient temperature, wash water processed after each cycle, 2) two 45 min wash cycles with fabrics not dried between washes, wash water processed at the end of each cycle, and 3) one 90 minute wash cycle with wash water processed at the end (Figure S8). In all cases, approximately the same mass of microplastic fibers was collected on the respective wash filters. This indicates that 1) (air) drying has no effect on the quantity of fibers released for the

ACS Paragon Plus Environment

Page 18 of 27

Page 19 of 27

448 449 450 451

Environmental Science & Technology

subsequent washing cycle and 2) each time the surfactant is refreshed in the solution, another “batch” of fibers can be released with subsequent washes, but typically only a given number of fibers will be released into the solution in one wash cycle and there are not additional fibers released with additional mechanical stress/wash time.

452

453 454 455 456 457

Figure 6: A) Effects of laundering water temperature variation on the mass of microplastic fibers shed from the textile when washed in DI H2O (blue) and liquid detergent (green), B) Extended wash cycle to induce additional mechanical stress and C) addition of surfactant in various concentrations, representing the concentration in a recommended detergent dose (0.75 g/L) and overdoses (orange

ACS Paragon Plus Environment

Environmental Science & Technology

458 459

bars) compared to the standard addition of liquid laundry detergent (green bar). Error bars represent triplicate experiments.

460 461 462

Important Variables for Shedding of Microplastic Fibers from Textiles

463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478

The extent to which the textile will shed fibers (fragments or staples) depends on a number of

479 480 481 482 483 484 485 486 487 488 489 490 491

There is good evidence that washing processes generally contribute more to fabric damage than does

variables including fabric type (woven, knit or non-woven), the texture (open or dense), the yarn type and the nature and number of the different fiber types involved. During the yarn spinning process, there are many harsh mechanical influences that can further break down and cut the fibers. There short fibers (fragments) will be embedded in the yarn and they can later be washed out. These fragments are likely the majority of the microplastic fibers emitted from our samples, with a few true fiber staples also being released, which typically have lengths in the millimeter size range. The ways in which water and variables such as (water) temperature and mechanical action affect fabrics during the washing process, especially during the first few washes, often depend strongly on the previous history for the fibers, yarns and fabrics as well as on fabric geometry and the physicochemical properties of the fibers themselves. Consequently, factors such as 1) yarn structure and the nature of stresses built into yarns during spinning, 2) fabric structure and the nature of stresses built into fabrics during knitting or weaving, 3) physical and chemical effects of scouring, bleaching, dyeing and chemical finishing process, and 4) drying methods should all be considered when studying the performance of individual fabrics.34 Depending on the fiber and fabric, certain aspects may be of more important for the given finished fabric.

use or wear (up to 90% of damage can be caused by the wash process35), though “damage” to the fibers is a broad category which may include many metrics including tensile strength, shrinking, dimensional stability, etc. and not necessarily fiber shedding, as we are mostly interested in in this particular case relating to microplastic fiber release. The combined effects of water and wash temperature will influence the fiber glass transition temperature, fiber swelling and water diffusion rate between the fiber strands. While water hardness or laundering products may slightly affect the rate of water diffusion into the fiber, total fabric wetting and fiber swelling generally occurs within 30 seconds of being immersed, immediately putting strain across the fabric network. Polyester fibers are highly crystalline, mechanically tough, hydrophobic, and thus do not swell significantly in water.34 This inherent toughness and stability of polyester fibers makes individual fiber breakage less likely, though when exposed to temperatures above the glass transition temperatures (typically 100 oC for polyester), the fibers could stretch and change mechanical properties.34,36 In our samples,

ACS Paragon Plus Environment

Page 20 of 27

Page 21 of 27

Environmental Science & Technology

492 493

which have fiber staples spun into yarn, shedding of the smaller staples can occur without

494 495 496 497 498 499 500 501 502 503 504 505 506 507 508

For both natural and synthetic fibers, the use of washing detergents significantly modifies the

509 510 511 512 513 514 515 516 517 518

Understanding the shedability of textiles is in itself not a new topic, albeit with very different aim

519 520 521 522 523 524 525 526

In contrast to fiber shedding, pilling is another result of abrasion associated mechanical stress on the

breakdown of the base material itself if they are pulled or loosened from the yarn.

physical effects of mechanical action during laundering in two main ways; 1) agitation of the surfactant solution generates foam and this cushions the fabrics against the beating and rubbing action during washing, thereby reducing fabric damage and 2) surfactants can deposit on the fiber surface which reduces frictional forces: essentially lubricating the fibers and mitigating fiber damage. In our samples, we hypothesize that some surfactant is necessary to mobilize the fiber fragments or staples off of the fabric itself, hence more microplastic fibers are found in the wash water when laundered in detergent vs. H2O. Conversely, the addition of increasing amounts of surfactant into the wash solution does not release increasingly higher amounts of microplastic fibers, since the fibers may be sufficiently protected from additional damage. Additionally, when washed at or below recommended temperatures, the relatively hydrophobic synthetic fibers are not noticeably affected by alkaline detergent solutions, thus this is why we did not observe a significant difference between detergent variants on microplastic fiber release. It should be noted, though, that other synthetic fibers, such as polyamide, are more susceptible to oxidative attack on hydrocarbon elastomers which can cause fiber failure.34

than to describe the amount of microplastic fibers that could be released into the environment. Fiber shedding is a critical problem in biomedical textile materials since stray or emitted fibers can lead to infection or otherwise impair wound healing. In these tests, the number of particles released from the testing fabric is counted and classified using a particle counter within the size range from 0.3 to 25 µm.37-39 The cause of fiber shedding was found to be predominantly fiber slippage, coating point rupture and/or fiber breakage taking into account fabric and fiber metrics such as stitch size, pile density, number of ground yarns, back coating and coating and surface chemistry of the individual fibers.40 Other testing methods, such as abrading and tape methods, are used to estimate the fiber shedding propensity of apparel textiles.41-45

fabric which could release polymeric material into the wash water. We did not (visually) note any pills either on the wash water filter nor on the fabric surface of our test swatches, yet Napper et al. did discuss pilling in relation to fiber release from laundering a selection of textiles under various washing conditions.23 In that work it was unclear if a differentiation was made between discrete fibers and pills, but the authors used cotton/polyester blends in some instances and in those systems pills may occur more readily than in synthetic only textiles.23. Chiweshe et al. noted that polyester interlock and woven fabrics has very little pilling and that detergent type (with and without

ACS Paragon Plus Environment

Environmental Science & Technology

527 528 529 530 531 532 533 534 535

enzymes) did not significantly influence pilling; an expected result given that cellulose enzymes are

536 537 538 539 540 541 542 543 544 545 546 547

The mechanical properties of knitted fabrics upon extended aging (up to 40 washes) of both

548 549 550 551 552 553 554 555 556 557

Nevertheless, depending on textile type, there can be a large number of microplastic fibers released,

capable of hydrolyzing only cellulose fibers.46 However, the use of detergents with cellulose enzymes (which are contained in both detergents used in this study), was shown to significantly reduce the negative effects of fabric softeners which created pills on the fabric, hypothetically by hindering softener buildup on the polyester fibers. While pilling is mostly associated with in use wear (especially knitwear and blends opposed to purely synthetic textiles), this can still happen through laundering, though typically only after approximately 10 wash cycles.34 However, there is little or no tendency for pills to wear off after further washing because the strength of polyester fibers anchors the pill to the fabric structure.

cellulosic (cotton and viscose) and synthetic (polyester) textiles was investigated by Agarwal et. al in an extensive matrix of fiber type, yarn construction and knitting structure.47 The authors found that the influence of wash aging on different mechanical parameters (tensile extension, sheer rigidity, bending rigidity compression energy and roughness) was only prominent for viscose fabrics; polyester fabrics remained unchanged even after 40 wash cycles. The metric of roughness may have the most direct influence on the propensity for fibers to be released from the textile, and while extended aging caused viscose fibers to breakdown, lint and pill, polyester fibers remained intact. In relation to our experimental work, we also note that there is not an increase in fiber release over multiple wash cycles. i.e. the fabric releases a similar amount of material in the first wash as in subsequent washes. This would indicate that the repetitive stress of washing does not mechanically degrade the fibers significantly to continually release more fibers over time.

in a variety of sizes, through the laundering process. There is good cause to consider synthetic textiles a major source of microplastic fibers that will not diminish due to changes in consumer habits since the use of polyester fabric continues to increase.48 Indeed, when making a priority list that indicates which microplastic sources can or should be targeted for emission reduction measures, shedding of microplastic fibers from clothing ranks high.49 Several groups have simply suggested that a filter could be placed on the washing machine outflow to reduce the flux of microplastic fibers into the WWTP (and beyond). While this is certainly a good starting point to help diminish the large fibers released from textiles, the filter designs underway may miss the smallest (and most numerous) fraction of microplastic fibers released.

558 559

Acknowledgements

ACS Paragon Plus Environment

Page 22 of 27

Page 23 of 27

560 561 562 563

Environmental Science & Technology

We would like to thank Dr. Maike Quandt and Leonie El Issawi-Frischknecht, both from Empa, for helping to obtain the fabrics used in this study, practical assistance in preparing the fabric swatches and for valuable discussions on the topic of textiles. Additionally, Brian Sinnet (Eawag) captured the SEM images.

564 565 566

Supplemental Information

567 568 569 570 571 572 573 574

Images of polyester interlock and jersey fabrics exemplifying differences in knit, images of wash water filter with and without pre-wash, discussion of fabric pre-treatment, SEM of textile fibers, examples of lines drawn on fiber images using ImageJ, discussion on filter coverage correlation curve, table of chemical composition of washing detergents used, fiber length distributions from a subset of wash waters for the QA/QC experiments, filter coverage analysis from QA/QC experiments, table of percent of underestimation using binary versus manual fiber counting, table of number of individual fibers measured on each filter, effects of sequential washing and drying on microplastic fiber release amounts.

575

ACS Paragon Plus Environment

Environmental Science & Technology

576

References

577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620

1. Cole, M.; Lindeque, P.; Halsband, C.; Galloway, T. S., Microplastics as contaminants in the marine environment: a review. Marine pollution bulletin 2011, 62, (12), 2588-2597. 2. Hammer, J.; Kraak, M. H.; Parsons, J. R., Plastics in the marine environment: the dark side of a modern gift. In Reviews of environmental contamination and toxicology, Springer: 2012; pp 1-44. 3. Lechner, A.; Keckeis, H.; Lumesberger-Loisl, F.; Zens, B.; Krusch, R.; Tritthart, M.; Glas, M.; Schludermann, E., The Danube so colourful: a potpourri of plastic litter outnumbers fish larvae in Europe's second largest river. Environmental pollution 2014, 188, 177-181. 4. Castañeda, R. A.; Avlijas, S.; Simard, M. A.; Ricciardi, A.; Smith, R., Microplastic pollution in St. Lawrence River sediments. Canadian Journal of Fisheries and Aquatic Sciences 2014, 71, (12), 1767-1771. 5. Eerkes-Medrano, D.; Thompson, R. C.; Aldridge, D. C., Microplastics in freshwater systems: A review of the emerging threats, identification of knowledge gaps and prioritisation of research needs. Water research 2015, 75, 63-82. 6. Desforges, J.-P. W.; Galbraith, M.; Dangerfield, N.; Ross, P. S., Widespread distribution of microplastics in subsurface seawater in the NE Pacific Ocean. Marine pollution bulletin 2014, 79, (1), 94-99. 7. Ballent, A.; Purser, A.; de Jesus Mendes, P.; Pando, S.; Thomsen, L., Physical transport properties of marine microplastic pollution. Biogeosciences Discussions 2012, (9), 18755-18798. 8. Fok, L.; Cheung, P., Hong Kong at the Pearl River Estuary: A hotspot of microplastic pollution. Marine pollution bulletin 2015, 99, (1), 112-118. 9. Lechner, A.; Ramler, D., The discharge of certain amounts of industrial microplastic from a production plant into the River Danube is permitted by the Austrian legislation. Environmental Pollution 2015, 200, 159-160. 10. Fendall, L. S.; Sewell, M. A., Contributing to marine pollution by washing your face: Microplastics in facial cleansers. Marine Pollution Bulletin 2009, 58, (8), 12251228. 11. Derraik, J. G., The pollution of the marine environment by plastic debris: a review. Marine pollution bulletin 2002, 44, (9), 842-852. 12. Carr, S. A.; Liu, J.; Tesoro, A. G., Transport and Fate of Microplastic Particles in Wastewater Treatment Plants. Water Research 2016, 91, 174-182. 13. Murphy, F.; Ewins, C.; Carbonnier, F.; Quinn, B., Wastewater Treatment Works (WwTW) as a Source of Microplastics in the Aquatic Environment. Environmental science & technology 2016, 50, 5800-5808. 14. Mahon, A. M.; O’Connell, B.; Healy, M. G.; O’Connor, I.; Officer, R.; Nash, R.; Morrison, L., Microplastics in Sewage Sludge: Effects of Treatment. Environmental Science & Technology 2016, 51, (2), 810-818. 15. Talvitie, J.; Heinonen, M.; Pääkkönen, J.-P.; Vahtera, E.; Mikola, A.; Setälä, O.; Vahala, R., Do wastewater treatment plants act as a potential point source of microplastics? Preliminary study in the coastal Gulf of Finland, Baltic Sea. Water Science and Technology 2015, 72, (9), 1495-1504.

ACS Paragon Plus Environment

Page 24 of 27

Page 25 of 27

621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666

Environmental Science & Technology

16. Browne, M. A.; Crump, P.; Niven, S. J.; Teuten, E.; Tonkin, A.; Galloway, T.; Thompson, R., Accumulation of microplastic on shorelines woldwide: sources and sinks. Environmental science & technology 2011, 45, (21), 9175-9179. 17. Eriksen, M.; Mason, S.; Wilson, S.; Box, C.; Zellers, A.; Edwards, W.; Farley, H.; Amato, S., Microplastic pollution in the surface waters of the Laurentian Great Lakes. Marine pollution bulletin 2013, 77, (1), 177-182. 18. Hoellein, T.; Rojas, M.; Pink, A.; Gasior, J.; Kelly, J., Anthropogenic litter in urban freshwater ecosystems: distribution and microbial interactions. PLoS ONE 2014, 9, (6). 19. Magnusson, K.; Wahlberg, C., Mikroskopiska skräppartiklar i vatten från avloppsreningsverk. Rapport NR B 2014, 2208, 33. 20. Rochman, C. M.; Kross, S. M.; Armstrong, J. B.; Bogan, M. T.; Darling, E. S.; Green, S. J.; Smyth, A. R.; Veríssimo, D., Scientific evidence supports a ban on microbeads. Environmental science & technology 2015, 49, 10759 - 10761. 21. http://www.beatthemicrobead.org/ 22. Shui, S.; Plastina, A., World Apparel Fiber Consumption Survey. In Food and Agriculture Organization of the United States and International Cotton Advisory Committee: 2013. 23. Napper, I. E.; Thompson, R. C., Release of synthetic microplastic plastic fibres from domestic washing machines: Effects of fabric type and washing conditions. Marine Pollution Bulletin 2016, 112, (1), 39-45. 24. Hartline, N. L.; Bruce, N. J.; Karba, S. N.; Ruff, E. O.; Sonar, S. U.; Holden, P. A., Microfiber Masses Recovered from Conventional Machine Washing of New or Aged Garments. Environmental Science & Technology 2016, 50, (21), 11532-11538. 25. Pirc, U.; Vidmar, M.; Mozer, A.; Kržan, A., Emissions of microplastic fibers from microfiber fleece during domestic washing. Environmental Science and Pollution Research 2016, 23, (21), 22206-22211. 26. Astrom, L. Shedding of synthetic microfibers from textiles. University of Gotherburg, 2016. 27. Frackiewicz-Kaczmarek, J.; Psikuta, A.; Bueno, M.-A.; Rossi, R. M., Air gap thickness and contact area in undershirts with various moisture contents: influence of garment fit, fabric structure and fiber composition. Textile Research Journal 2015, 80, (20), 2196 - 2207. 28. Mitrano, D. M.; Arroyo Rojas Dasilva, Y.; Nowack, B., Effect of Variations of Washing Solution Chemistry on Nanomaterial Physicochemical Changes in the Laundry Cycle. Environmental Science & Technology 2015, 49, (16), 9665-9673. 29. Mitrano, D. M.; Lombi, E.; Arroyo Rojas Dasilva, Y.; Nowack, B., Unraveling the Complexity in the Aging of Nanoenhanced Textiles: A Comprehensive Sequential Study on the Effects of Sunlight and Washing on Silver Nanoparticles. Environmental Science & Technology 2016, 50, (11), 5790-5799. 30. Geranio, L.; Heuberger, M.; Nowack, B., The Behavior of Silver Nanotextiles during Washing. Environ Sci Technol 2009, 43, (21), 8113-8118. 31. Lorenz, C.; Windler, L.; von Goetz, N.; Lehmann, R.; Schuppler, M.; Hungerbühler, K.; Heuberger, M.; Nowack, B., Characterization of silver release from commercially available functional (nano) textiles. Chemosphere 2012, 89, (7), 817824.

ACS Paragon Plus Environment

Environmental Science & Technology

667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709

32. Windler, L.; Lorenz, C.; von Goetz, N.; Hungerbuhler, K.; Amberg, M.; Heuberger, M.; Nowack, B., Release of Titanium Dioxide from Textiles during Washing. Environmental Science & Technology 2012, 46, (15), 8181-8188. 33. Mitrano, D. M.; Rimmele, E.; Wichser, A.; Erni, R.; Height, M.; Nowack, B., Presence of Nanoparticles in Wash Water from Conventional Silver and Nano-silver Textiles. ACS nano 2014, 8, (7), 7208-7219. 34. Bishop, D., Physical and chemical effects of domestic laundering processes. In Chemistry of the Textiles Industry, Springer: 1995; pp 125-172. 35. Mohamed, S. S., Comparison of Phosphate and Carbonate Built Detergents for Laundering Polyester/Cotton. Textile Chemist & Colorist 1982, 14, (3). 36. Weinberger, C. B. Synthetic fiber manufacturing; 1996. 37. Mehta, S.; Hersh, S.; Tucker, P.; Bullerwell, A., Evaluating textiles and apparel for controlling contamination in cleanrooms. In Performance of Protective Clothing: Fourth Volume, ASTM International: 1992. 38. 9073-10:2003, I., Test methods for nonwovens - Part 10: Lint and other particles generation in the dry state. In 2014. 39. 1644-2:2000, E., Test Methods for Nonwoven Compresses for Medical Use Part 2: Finished Compresses. In. 40. Fu, Y.; Xie, Q.; Lao, J.; Wang, L., In Vitro Evaluation and Mechanism Analysis of the Fiber Shedding Property of Textile Pile Debridement Materials. Materials 2016, 9, (4), 302. 41. De Wael, K.; Lepot, L.; Lunstroot, K.; Gason, F., Evaluation of the shedding potential of textile materials. Science & Justice 2010, 50, (4), 192-194. 42. Coxon, A.; Grieve, M.; Dunlop, J., A method of assessing the fibre shedding potential of fabrics. Journal of the Forensic Science Society 1992, 32, (2), 151-158. 43. McDermott, S. D., Trace Evidence: Transfer, Persistence, and Value. John Wiley & Sons, Ltd: 2013. 44. Grieve, M.; Dunlop, J.; Haddock, P., Transfer experiments with acrylic fibres. Forensic Science International 1989, 40, (3), 267-277. 45. Bresee, R. R.; Annis, P. A., Fiber Transfer and the Influence of Fabric Softener. Journal of Forensic Science 1991, 36, (6), 1699-1713. 46. Chiweshe, A.; Crews, P. C., Influence of household fabric softeners and laundry enzymes on pilling and breaking strength. Textile Chemist and Colorist and American Dyestuff Reporter 2000, 32, (9), 41-47. 47. Agarwal, G.; Koehl, L.; Perwuelz, A.; Lee, K. S., Interaction of textile parameters, wash-ageing and fabric softener with mechanical properties of knitted fabrics and correlation with textile-hand. I. Interaction of textile parameters with laundry process. Fibers and Polymers 2011, 12, (5), 670. 48. Carmichael, A., Man-made fibers continue to grow. Textile World 2015, 165, 20-22. 49. Verschoor, A.; de Poorter, L.; Roex, E.; Bellert, B., Quick scan and prioritization of microplastic sources and emissions. RIVM letter report 2014-0156 2014.

710

ACS Paragon Plus Environment

Page 26 of 27

Page 27 of 27

Environmental Science & Technology

ACS Paragon Plus Environment