Analytical pyrolysis for the characterization of microplastics and their degradation products in the environment

(1)

UNIVERSITÀ DI PISA

Dipartimento di Chimica e Chimica Industriale

Corso di Laurea Magistrale in Chimica

Curriculum Analitico

CLASSE: LM-54

Tesi di Laurea

Analytical pyrolysis for the characterization of microplastics

and their degradation products in the environment

Relatore Controrelatore

Prof.ssa Francesca Modugno Dr. Tommaso Lomonaco

Candidata

Greta Biale

(2)

(3)

“It all began the day I found That from my window I could only see A piece of sky” Yentl

(4)

(5)

1. Introduction

The worldwide production of plastic materials since the second half of the 20th_{century has}

made plastic pollution an environmental threat and a health risk [1]. Thus, the assessment and the quantification of plastic debris in the ecosystem, and in particular in the marine environment is the focus of an intense multidisciplinary research: chemistry, marine biology, medicine, among others.

Currently, about 5-13 million tons of plastic are estimated to enter the ocean every year in the form of item or debris. Plastic debris can be classified by size: mega-debris (> 100 mm), macro-debris (> 20 mm), meso-debris (20-5 mm) and micro-debris (< 5 mm) [2].

Microplastics (MPs) belong to micro-debris and are commonly defined as plastic particles in the 1 μm – 5 mm dimension range, even if a shared and worldwide accepted definition has not been recognized yet. An issue that has emerged along with the study of this topic, is the lack of common MPs classification criteria: during the last decade different research groups have suggested different classifications of MPs on the basis of their shape (film, foam, pellets, sheets…) color, or dimension [3]. On the basis of their origin MPs can be classified in primary and secondary MPs. The first ones are produced in the micro dimension range, while the second ones are the result of degradation and fragmentation of bigger plastic debris.

The polymers that are reported to contribute more to plastic debris pollution are: polypropylene (PP), polyethylene (PE), polyvinylchloride (PVC), polyurethane (PUR), polyethylene terephthalate (PET), and polystyrene (PS) [4].

The risks correlated to plastic debris can be enhanced by the reduction of their dimensions and by degradation, that make MPs easier to be ingested by marine organisms causing death and inflammatory responses; moreover, with degradation plastic can release additives and small molecular weight oxidized fraction harmful for the environment and the biota [5, 6]. However, the actual risks to human health and the environment are at present not well-established [7].

For this reason, not only the assessment of the presence of MPs is important but also the identification of the polymer type in environmental samples, because different polymers can have various and not completely known impacts on the biosphere [8]. Thus, at the present state of the art the research is focused on the development and application of analytical techniques suitable to identify MPs in environmental samples (water, soil, sand, marine organisms), to recognize the polymer type, and to quantify them.

The widespread interest in the topic is demonstrated by the several recent European calls published in 2019 and in 2020, as in example: the JPI Oceans calls “Microplastics in the Marine Environment”, and “Aquatic Pollutants”( http://jpi-oceans.eu/calls); the HORIZON 2020 calls “CE-SC5-30-2020 Plastics in the environment: understanding the sources,

(10)

2

transport, distribution and impacts of plastics pollution” and “SC1-BHC-36-2020 Micro- and nano-plastics in our environment: Understanding exposures and impacts on human health”. It is to be underlined that so far, the most common analytical techniques used for the analysis of MPs are spectroscopic techniques like Fourier transform infrared spectroscopy (FTIR), micro FTIR (μ-FTIR), or Raman spectroscopy [8-10]. The research activities in the field show how in the last years analytical pyrolysis is becoming a key candidate in the research and also routine analysis of MPs.

MPs analysis comprises in most cases many analytical steps, that need to be carefully evaluated. The published researches report the use of different analytical methodologies, highlighting the need of internationally accepted standard methods and procedures for the characterization of MPs; this comprises also standard methods regarding sample-handling in the laboratories in order to avoid cross-contamination [11].

A part of the work carried out in my thesis has represented a basis for the activities of the following projects: the University of Pisa project “PRA_2018_26 Advanced analytical pyrolysis to study polymers in renewable energy, environment, cultural heritage (2018-2020)”; the Fondazione Carilucca – Research 2019 project “Micro and nanoplastics: quantification, evaluation of their impact on marine and lacustrine ecosystems, and remediation strategies (2019-2021)”; the JPI-Oceans HOTMIC project “Horizontal and vertical oceanic distribution, transport and impact of microplastics (2020-2023)”. I participated to the virtual kickoff meeting, in which all the research project funded in the program JPI-Oceans microplastic call were presented (20th_{May 2020}

https://www.jpi- oceans.eu/news-events/news/six-new-jpi-oceans-microplastics-research-projects-launched-and-good-start). Assisting to the event gave me an overview of the hot topics in MP research today. For what concerns the analytical aspects that are more relevant for my thesis, one of the main issues is the development, validation and harmonization of analytical methods to quantify nano-and microplastic particles and to investigate their fate and degradation phenomena in the environment, and their interaction with the ecosystems including the intake by organisms. The target is to establish a risk-based assessment of microplastics pollution in the environment.

The aim of this thesis project is to contribute to evaluate the potentialities of analytical techniques based on analytical pyrolysis and mass spectrometry for the analysis of MPs. For this purpose, I initially carried out an extensive survey of the state of the art on analysis of MPs by means of analytical pyrolysis. This survey allowed me to gain knowledge of the critical issues that need to be faced approaching MPs research: sample representativeness, sample preparation, pre-treatments, interference of the matrix, specificity of pyrolysis signals, difficulties in the calibration.

The results of this literature study are described in Chapter 1 and have been published as part of a review article, written in collaboration with my supervisors and with Prof. Daniele Fabbri of University of Bologna [12].

(11)

3

This state-of-the-art basis allowed the design of my experimental work, that can be schematized into two work-packages:

• analysis of reference polymers samples subjected to artificial aging in order to characterize them and their degradation products

• development of a quantitative method based on microwave-assisted extraction and double-shot Py-GC-MS for the analysis of phthalates and polystyrene in environmental samples.

The first work-package (described in Chapter 4) was carried out in strict collaboration with the polymer research group M_4.0_LAB directed by Prof. Valter Castelvetro.

Reference polymers - polypropylene (PP), polystyrene (PS), polyethylene terephthalate (PET), low-density polyethylene (LDPE) and high-density polyethylene (HDPE) - subjected to artificial aging using a SolarBox for 4 weeks during a previous master thesis [13], were analyzed by means of evolved gas analysis-mass spectrometry (EGA-MS) and the results were compared with those obtained for the unaged polymers to study the thermal degradation behavior specific for each polymer after artificial aging. The unaged and the 4 weeks aged polymer samples were also analyzed by means of pyrolysis-gas chromatography-mass spectrometry (PY-GC-MS) to detect and characterize possible degradation products. The same samples were then subjected to solvent extraction in refluxing dichloromethane (DCM) -methanol for PS - in order to extract only the degraded, low molecular weight fractions. The extracts were then analyzed by size exclusion chromatography (SEC), and Py-GC-MS using hexamethyldisilazane (HMDS) as derivatizing agent to characterize the polar and low-volatile compounds formed as a consequence of photo-oxidation. The corresponding insoluble residues were analyzed by means of Py-GC-MS to verify the effectiveness of the extraction procedure.

The second work-package of the thesis project (described in Chapters 5 and 6 ) is focused on the development and the validation of an analytical method based on microwave-assisted extraction and double-shot Py-GC-MS for the quantification of phthalate plasticizers and PS. The conditions of the double shot Py-GC-MS method were selected on the basis of the EGA profiles of the phthalates and the polymers (PS, PP, and HDPE).

The validated method has been object of a publication [14] and has been applied to environmental sand samples in order to assess the presence and to quantify phthalates and PS pyrolysis products in the first shot and the second shot, respectively, of the double-shot Py-GC-MS analysis. Moreover, the analysis allowed the identification of PP and HDPE degradation products.

(12)

(13)

5

2. State of the art

Plastic particles in the range of 1 μm to 5 mm and those in the sub-micrometer range are commonly denoted as microplastics (MPs) and nanoplastics (NPs), respectively [15]; microplastics (MPs) are being recognized as nearly ubiquitous pollutants in water bodies, but their actual concentration in natural waters, sediments and biota is still largely unknown. A lack of reliable data emerges from the literature [2, 16]. Their wide distribution in the global environment as a result of human activities has been ascertained, and they represent a crucial source of pollution. Analytical chemistry plays a key role in developing methods for their identification and quantification, that are crucial for devising a global strategy for an effective evaluation and successful mitigation of this kind of pollution [9, 17].

Since 1950 the widespread and massive production (more than 8300 million tons) [9] of plastic materials has made plastic debris an increasing environmentally issue all over the world. MPs sources are diverse, and they are produced through multiple activities, across large areas, and can be found in quite different matrices, from soils to aquatic systems including rivers, shorelines, swamps and oceans, and organisms. The level of MP contamination is considered alarming and increasing in recent literatures, in fact, it is estimated that 5-13 million tons of plastic debris enter the oceans every year, and cumulative plastic input will probably double in the next decade [18] threatening both terrestrial and aquatic ecosystems [17]. The 12 polymers with the highest annual production are polypropylene (PP), low/very low density polyethylene (LDPE/LLDPE), high/medium density polyethylene (HDPE/MDPE), polyvinylchloride (PVC), polyurethane (PUR), polyethylene terephthalate (PET), polystyrene (PS), expanded PS (PS-E), styrene-acrylonitrile resin/acrylonitrile butadiene styrene (SAN/ABS), polyamide (PA), polycarbonate (PC) and poly (methyl methacrylate) (PMMA) [19].

So far, there are no accounts of biodegradable MPs, such as polylactic acid (PLA) or polycaprolactone (PCL) in the environment, although aging studies suggest that also biodegradable plastic form secondary MP through weathering [20].

2.1 Definition and classification of microplastics (MPs)

The term microplastics (MPs) usually refers to synthetic solid particles or polymer matrix in the 1 μm-5 mm size range [9, 21]. Plastic particles in the microscale dimension range were first observed in the marine environment in the early 1970s [22, 23], but only in 2004 the term “microplastics” became commonly used [21]. Moreover, the European Marine Strategy Framework Directive (MSFD) Technical Subgroup on Marine Litter suggested a further classification into small microplastics (< 1 mm), large microplastics (1–5 mm), mesoplastics

(14)

6

(5-25 mm) and macroplastics (> 25 mm) [24]. MPs exist in a variety of morphologies and they can be categorized on the basis of the shape: pellet, fragment, fiber, film and foam [3]. MPs are also often classified in primary and secondary microplastics [4, 17, 24]. Primary microplastics are specifically produced with micro- and nano-sizes, and they are released in the environment through industrial or domestic water discharge systems; primary MPs comprehend acrylic or polyester beads used in industrial abrasives for sandblasting, synthetic microfibers released in laundry wastewaters, plastic pre-production pellets or polyethylene microbeads used as exfoliating agents in cosmetic formulations [2]. On the other hand, secondary MPs derive from the mechanical (such as sea-wave action), photo-oxidative, and biological degradation of meso and macroplastic [1, 2] (Figure 2.1). Larger plastic debris are subjected to weathering degradation and fragmentation once in the environment on the basis of the type of polymers and exposure conditions [25]. Direct exposure to UV radiation from sunlight and high temperature are primary causes of fragmentation of plastics on land and floating in marine waters. In particular, polymers with a C-C backbone like polyethylene (PE), PP and PS are more susceptible to fragmentation [25, 26], and, due to their low-density, they are more likely to end up in the shoreline [27]. Even if, regarding polyolefins, chain scission and cross-linking are the main degradation pathways linked to thermal oxidation, in the presence of oxygen, chain scissions are the main reactions causing the reduction of the average molecular weight [28]. Compared to polyolefins, polyethylene terephthalate (PET) particles have a higher stability to photo-oxidative degradation and are more susceptible to hydrolytic degradation (even if it is very slow) [20, 26]; so they are expected to degrade leading to the formation of low molecular weight products including monomers and oligomers rather than secondary MPs and NPs [26, 29, 30]. Due to their higher density they are more likely found in water sediments with little or no exposure to photo-oxidizing conditions [30].

Figure 2.1 Secondary MPs resulting from surface fragmentation of a PP item on the beach exposed to photo-oxidative and thermal aging [25].

(15)

7

Photo-degradation occurs through three steps [26]: initiation, propagation and termination; the initiation step consists in the cleavage, by light or heat, of the chemical bonds in the main polymer chain to produce free radicals. Propagation, in the end, consists in chain scission reactions leading to the formation of olefins, aldehydes and ketones which should be more susceptible to photoinitiated degradation, because they contain unsaturated double bonds. Photo-oxidation can cause chemical changes making plastics brittle and thus more susceptible to fragmentation; fragmentation increases surface area and number of particles per unit of mass (Figure 2.1). On land, especially at the soil surface, temperature fluctuations will generally be greater than those in sea water [1, 31]. Also, as a consequence of photo-oxidation, degradation products like low-molecular weight polymer fragments with oxidized end groups (aldehydes and ketones, and dicarboxylic acids) are expected to be released into the environment like water, if floating litter are considered (Figure 2.2) [29].

Figure 2.2 Possible abiotic degradation pathways suggested for PE (R=H), PP (R=CH3) and PS (R=aromatic ring) after

initiation by photolytic cleavage of a C–H bond on the polymer backbone [26].

R P -H P R P R R R O O POO O2 R O HO R PO OH P POOH R O PO H2O R O OH P R R chain scission O R R O O hν

(16)

8

2.2 Microplastics as environmental threat

The issues related to MPs research are not limited to assess their presence, but there is a need also to investigate how these particles interact with both terrestrial and marine organisms, and their effects on them and the environment [7]. Moreover it has been studied that toxic additives present in plastics, such as plasticizers used in the manufacture of polymers, leach out from the plastic debris over time and can contaminate the environment [2, 6, 32].

2.2.1 Ecotoxicological impact

One of the most critical issues associated to MPs in the natural environment is the possibility to be ingested by vertebrates and invertebrates, causing problems at different levels, for instance on the mortality rate and the reproductive capability, causing damages to the ecosystem. Three factors can influence the possible ingestion of MPs by marine organisms: size, color and shape [17]. The possibility to be ingested by organisms increases when the particles are smaller, furthermore particles of small size can permeate body tissues resulting in an inflammatory response [2]. This aspect has been studied mostly on marine species [33] and they are poorly studied on terrestrial species. Ingestion by living organisms is potentially harmful to the entire ecosystem, including humankind, since MPs ingested by fish and mussels have recently been suspected to enter in the human food chain [34, 35]. Analytical procedures that can efficiently characterise MPs in biological tissues are thus urgently required [36]. It is also important to evaluate how MPs degradation rate enhances MPs toxicity; for instance, it has been reported that exposure to artificially aged nano-polystyrene can have some serious effects on the planktonic crustacean, Daphnia magna, whereas pristine nano-polystyrene cause no significant effects [37].

2.2.2 Chemical hazard

During the manufacture of plastic materials, additives are used to give specific properties, such as fillers and reinforcement to improve mechanical properties, antioxidants and stabilizer for resistance to heat and aging, UV stabilizers to provide resistance to light degradation, pigments, and colorants. Many of these additives like bisphenol A, phthalates esters (PAEs) , polybrominated diphenyl ethers (PBDEs) and metals are reported to be toxic or endocrine disruptors [2]. The leach out of these additives into the environment (water, sediments) occurs over time and it is sped up by UV irradiation and heath [1, 32]. Different parameters have been reported to alter plasticizers leaching like temperature, oxygen presence, and pH of the sediments or marine environment where MPs are [2]. One of the first studies about the monitoring of PAEs in soil samples dates back to 2008 [38] where a wide range of soil samples from agricultural and peri-urban sites in China were analysed. These studies proved that the concentration of phthalates increases with the proximity to solid waste sites.

Moreover, studies have reported that MPs interact with hydrophobic organic chemicals (HOCs) within the environment like organochlorine pesticides, polycyclic aromatic

(17)

9

hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs), enhancing their potential toxic effect [2].

The release of potentially harmful volatile organic compounds (VOCs) such as carbonyls, lactones, esters, acids, alcohols, ethers, and aromatics, during the chemical degradation of MPs induced by high temperature and UV irradiation has been recently reported [5, 25, 39] and deserves further investigation.

2.3 Analytical pyrolysis in the analysis of MPs

Instrumental analytical techniques based on analytical pyrolysis and thermal analysis has found a wide use in the investigation of polymers [40, 41]. The more common approaches are pyrolysis-gas chromatography-mass spectrometry (Py-GC-MS) [42-44] and thermal extraction-desorption-GC-MS [45] (TED-GC-MS, a combination of TGA, solid phase extraction SPE, and thermal desorption).

In the last years analytical pyrolysis and thermal techniques have become great candidates in the routine analysis of MPs [12], as integration methods to spectroscopic and microscopic techniques.

Figure 2.3 reports a summary of analytical procedures involving thermochemical methods for the analysis of MPs.

Figure 2.3 Schematic summary of analytical procedures for MP analysis. a_{Samples from biosphere (mussel, fish);}

lithosphere (sand, river sediments); hydrosphere (seawater, river water, lake water, sea salt). b_3Na₂_WO₄_9WO₃_{; ZnCl}₂_{; NaCl;} NaBr; NaI. c_{KOH 10%; H}₂_O₂_{30%; proteases.}d_{Stainless sieve > 0.25 mm, < 5 mm).}e_{Glass microfiber filter (GF/F), Ø 15 mm,} 0.2 µm. f_{Fragments picked up with tweezers, milled filter, etc.}g_{Anthracene-d10.} h_{tetramethylammonium hydroxide} (TMAH). i_{Or in alternative off-line pyrolysis [12].}

Thermochemical methods can provide both qualitative quantitative information and more accurate degradation information. In Py-GC-MS usually a sample is instantaneously pyrolyzed at about 400-600 °C, with or without reactive agents (like derivatizing agents) under a flow of carrier gas (N2 or He). The pyrolysis products are transferred into the GC

column, separated and analyzed by MS (Figure 2.4) to yield a pyrogram. Py-GC-MS can be performed as single-shot (flash pyrolysis) or as multi-shot pyrolysis; multi-shot-Py-GC-MS

ENVIRONMENTAL SAMPLESa

BIOSPHERE

LITOSPHERE

HYDROSPHERE

ATMOSPHERE Solvent extraction

Direct analysisi

Density separationb

Matrix degradationc Py-GC-MS

TED-GC-MS TG-MS TG-FTIR Py-APCI-MS Py-HRMS Py-Q-TOF …. TERMOCHEMICAL METHODS Size fractionationd Filtratione Single particle Mixed particlesf TEST SAMPLES homogenization drying Internal standardg Derivatisationh PRE-TREATMENTS

(18)

10

enables more analytical runs on a single sample: a thermal desorption for the volatile compounds such as additives and a pyrolysis for the polymer itself [46].

Figure 2.4 Instrumental asset of a furnace PY-GC-MS system [12].

Another promising technique used for MPs analysis is Evolved Gas Analysis-Mass spectrometry (EGA-MS) which is a complementary technique to Py-GC-MS. In EGA-MS, evolved gases formed during the programmed heating are directly transferred into a mass spectrometer by a deactivated open transfer line [41]. It can be suitable for degradation studies to characterize different fraction of MP samples on the basis of their thermal degradation temperature. The direct coupling of pyrolysis with MS was recently tested in a custom-made portable system [47].

Recently Py-GC coupled with atmospheric pressure chemical ionization-time-of-flight mass spectrometry (Py-GC-APCI-TOF-MS) was applied to the identification of polymers [48]. The APCI ionization/fragmentation process and the high mass resolution of TOF-MS were exploited to identify specific fragment ions in the pyrolyzates formed from model plastic mixtures, showing that the technique is a promising tool to characterize environmental MPs. Through direct analysis in real time-MS (DART-MS) thermally desorbed additives and pyrolysis products of polymers were detected in samples heated at 600 °C. The complexity of the mass spectra of real samples required a data analysis and statistical approach typical of petroleomics [49]. Py-GC-QTOF was used to identify the presence of PP, PS and PVC in river waters, and its potential for quantitation was also investigated [50].

The identification of polymer markers in complex pyrograms can be enhanced by high-resolution mass spectrometry, which has seldom been applied to the analysis of MPs. Also Py-GC interfaced with an Exactive® Orbitrap MS (Resolution 60,000 at m/z 200) was applied to detect methyl methacrylate (m/z 99.0441) and styrene (m/z 104.0621) for the quantitation of PMMA and PS in a spiked fishmeal alkaline digestate [51].

(19)

11

2.3.1 Sample pre-treatments

In order to isolate and produce a suitable portion of material to introduce into the Py-GC-MS system, an opportune pre-treatment is required. Plastic particles can be collected by visual inspection– using a microscope and tweezers - for particles with a minimum size suitable for handling, that can be approximatively estimated as > 0.1 mm in size or > 0.1 mg in weight [52], or even lower depending on the type of fragment and polymer [53]. Plastic particles can be directly picked up from the sample and transferred into the pyrolysis holder, as reported for the sand surface [53] or stomach content of marine fish [52]. Even when plastic particles are manually selected, a concentration/separation step from the samples is often necessary, except for few cases, in which the analysis of MPs in environmental matrices is based on the whole matrix after homogenisation. This approach was applied by Dümichen et al. to perform a preliminary screening using TED-GC-MS of different samples from a biogas plant. The samples were homogenized by a cutting mill, subsequently cooled in liquid nitrogen, and further homogenized with a centrifugal mill and an overhead shaker [45]. Dümichen et al. also used a similar cryo-milling and direct analysis approach for the characterisation of mussels and soil samples from the Spree River (Berlin) [54]. Analysing the whole sediment sample without separation is more suitable with TED-GC-MS than with Py-GC-MS due to the amount of sample that can be used for the analysis (ca. 20 mg), which is usually around 100 times greater than the sample weight normally introduced in Py-GC-MS. Off-line pyrolysis is another approach for the direct analysis, which presents the advantage to allow the analysis of high sample amounts, improving sensitivity and representativeness [55]. Off-line pyrolysis GC-MS was applied to study the accumulation of PS by mussels (M. galloprovincialis) exposed to different concentrations of the polymer in laboratory aquaria [56].

The sample treatments are strongly dependent on the features of the environmental matrix under investigation. The main types of matrices that have been analysed by thermoanalytical methods for MPs fragment analysis are water [45, 52, 53], sediments [57], soil [58], and marine organisms [44, 53], which can be sampled as whole individuals, or animal tissue, or as the content of fish stomachs.

Due to the selectivity achieved by the coupling with GC-MS, analytical pyrolysis techniques are suitable for the analysis of not only single-particles, but also mixtures of different micro- and nanoplastics obtained by extraction or filtration [44, 45, 59-62], and can also be potentially exploited for the analysis of solid microportions of homogenized ground mixtures of particles of different polymers. Cryo-milling is recommended in the homogenization of solid plastic samples in order to facilitate milling and to avoid heating and alterations in the polymers but needs to be carefully analyzed [54].

Filtration

MPs can be collected on filters that are subjected to Py-GC-MS or TED-GC-MS. Filtration can be directly applied to water samples where MPs are present as suspended particulate matter or floating fragments or from aqueous solutions derived from the last step of the

(20)

12

isolation procedure (e.g. after density separation). The pore size, the chemical nature of the filter to be used, and the amount of water to be filtered need to be evaluated very carefully. Glass fiber filters are the most common ones used in the analysis of MPs. The filters can be cut, milled or wrapped prior to pyrolysis [43, 63]. Styrene, probably derived from the sizing agent of the glass filter [64] or from the filter packaging made of PS was identified as a contaminant [44]. This contamination can be overcome heating the filter in a muffle at 300-400 °C or even 590 °C. Other aspects worth considering are the size of the filter in comparison to the maximum quantity of material that can be sampled for the pyrolytic run, as well as the need for particular care in the filtration procedure to prevent the uneven distribution of particles on the filter surface area. In a study a glass adaptor was set-up to reduce the deposition area of the particles on the filter surface [63]. Filtration was combined with centrifugation under constant operation conditions to process river water samples for TED-GC-MS [45].

Density separation

The most common approaches to separate MP particles from large amounts of sediments are based on density separation in salt solution (flotation), exploiting the low density of the majority of polymeric materials. The most common salts used for density separation [21, 65] are sodium chloride (NaCl, inexpensive, density of the saturated solution 1.2 g/cm3_[66]),

sodium bromide (NaBr, 1.46 g/cm3_{[43]) and sodium iodide (NaI, density of the saturated}

solution 1.8 g/cm3_{[65]). High-density (1.8 g/cm}3_{) salt solutions of zinc chloride [67] and}

sodium or zinc polytungstate [68] can be used as an alternative to enable higher density polymers, like PVC and PET, to be extracted. Density separation has been applied to sediments and other sample types including benthic fish for the elimination of sand [44], and MP particles in sea salts [42].

The amount of sediment to extract is a crucial issue in the density separation of MPs from sediments. Increasing the volume of extracted sediments has the advantage of allowing the analysis of very heterogeneous samples or samples with low concentration of MPs, in order to guarantee a representative sample examined; but the main drawback is its high cost. Nuelle et al. [65] balanced this choice by proposing a two-step density separation procedure (referred to as “two-step air induced overflow extraction method”) where the initial “fluidification” of the sediment was achieved in a low-cost NaCl saturated solution used to decrease the sample mass through the bubbling of air in the suspension in a device exploiting a pump to enrich the matrix in MPs. MP particles were then separated from the enriched sample through a flotation step with a saturated NaI solution. The supernatant portion was filtered using 0.45 m nitrocellulose filters. The authors tested the recovery of the separation procedure on several types of reference polymer particles, and the approach was used to investigate the heterogeneous distribution of MPs in the < 1mm sieved fraction starting with 1 kg portion of dry sediment from the German North Sea. After density separation, floating particles in the supernatant can be further separated into different size class fractions using steel sieves [63].

(21)

13

Innovative micro and nanoparticle extraction methods have recently been applied, based on the use of magnetic nanoparticles [69] and cloud-point extraction [70].

Purification from biogenic organic materials

A problem is that the flotation separation of MPs from sediments suffers from the presence of low-density biogenic natural organic matter (plant or animal parts) which floats on the surface of the salt solution, and can be difficult to visually distinguish it from plastic. The elimination of biogenic organic material before the analysis is carried out through exposure to alkalis (NaOH or KOH solutions), acids (HCl, HNO3, HClO4), or an oxidising H2O2 solution

(35%, 7 days) [65]. The acidic and alkaline digestion is not suitable for condensation polymers such as PET or polyamides, that are destructed [33]. On the other hand, some polymers are altered and/or degraded by oxidising solutions of H2O2 [65] or by alkaline

treatment [71]. In addition, coloured plastics can be discoloured by H2O2 treatment, making

very difficult the visual differentiation between e.g. coloured anthropogenic fibres and colourless natural fibers.

Even if the efficiency of the H2O2 solution was found to be higher than NaOH and HCl, with

the H2O2 treatment,only 50% of the biogenic organic matter was eliminated [41], so the

Py-GC-MS identification of the polymers needs to be based on highly specific pyrolysis markers. The H2O2 treatment was used to characterize MPs ingested by fish [44], and to characterize

water tow and sediments [52]. Extraction with organic solvents

Isolation from sediments and soil, including sand, has been tackled in several cases by extraction with organic solvents. One of the most common solvents used is dichloromethane (DCM) as it dissolves most polymers present in the environment as MPs. This solvent was used [72, 73] to extract polymers in sediments from the Adriatic Sea costal lagoon using Soxhlet-extraction in DCM of wet sediment samples (around 10 g), and subsequently precipitation in n-hexane, to investigate the presence of PS and PVC (following the EPA method for non-volatile organics) [74].

The extraction with DCM was also adopted [75] using a modified Kumagawa-type apparatus enabling the sample pre-treatment to be performed on 160 g of sediment. In 2019 recent evolution of solvent extraction was proposed [59] based on the use of a pressurized system and on a two-step procedure, a first step using methanol to reduce the matrix effects and a second one based on the use of tetrahydrofuran to extract the MPs. This pressurized system was used for the Py-GC-MS analysis of MPs in biosolids [76]. Trichlorobenzene has also been investigated for the extraction of PE, PP, and PS at 120 °C from soil samples [58].

In the analysis of sediment samples from Lake Bracciano [25], sieved and homogenized samples were extracted in refluxing DCM and the resulting insoluble DCM residues were then extracted in refluxing xylene. The soluble DCM fraction, containing PS, low molecular weight and oxidized polyolefin fragments, was analyzed by different analytical techniques including Py-GC-MS. The soluble xylene fraction was evaporated to a few mL volume and treated with an excess of warm methanolic KOH in order to precipitate the polyolefins

(22)

14

possibly present in the sample, and to separate most of the remaining biogenic contaminants; then, the dry solid was analyzed by means of infrared spectroscopy and Py-GC-MS to gain information about high molecular weight low-density PE (LDPE), high-density PE (HDPE), and PP. The insoluble xylene residue, containing condensation heteropolymers such as PET, was then subjected to acid hydrolysis to depolymerize the polyamide MPs. The solid residue from the acid hydrolysis step, was then subjected to strong alkaline hydrolysis, purification and preconcentration to quantify PET. In figure 2.5 is represented the experimental protocol for the fractionation, characterization and quantification of different polymer types present as MPs and NPs previously described.

Figure 2.5 Scheme of the experimental protocol for the fractionation, characterization and quantification of different polymer types present as MPs and NPs in sediment samples used in [25]. (TPA=terephthalic acid from PET depolymerization; PEox and PPox refer to the oxidized, low MW, DCM-soluble highly degraded polyolefin fraction).

Analysis of MPs in biological tissues

The sample pre-treatments described in the literature for the analysis of MPs in biological tissues are highly heterogeneous and far from being standardized. The most common approaches to pre-treat these samples are based on alkaline [77] or acidic [78] digestion or chemical oxidation [79] of the biological materials.

Different protocols were compared for the digestion of biological samples for the analysis of MP contamination in fish and shellfish [33] in order to assess the capacity of Py-GC-MS to identify polymers after that. Both Raman micro-spectroscopy and Py-GC-MS led to the correct identification of the type of polymer - in a set of 15 different plastic materials - after the application of alkaline digestion procedures (KOH 10% for 24h at 60°C), except in the case of cellulose acetate, which was not correctly identified after treatment with strong alkalis. The same protocol based on the use of alkaline digestion was also applied to characterize MPs particles collected during sailing [60], and to characterize the MPs in mussels and cockles [53].

(23)

15

Enzyme treatment are milder and have no significant effect on the structure of microplastics, thus continuous enzymatic digestion and purification methods based on proteolytic enzymes [80] also find increasing application [81] for the digestion of biological matrices. However, the drawbacks of enzymatic treatments are the higher cost and a relatively lower digestion efficiency.

2.3.2 Qualitative analysis

When qualitative analyses are performed by thermochemical approaches, the polymer is identified or classified on the basis of the molecular profile of the pyrolytic products produced in the thermal decomposition. Figure 2.6 reports the molecular structures of the main polymers investigated in MPs research and of their main pyrolysis products, with the corresponding m/z. One of the first applications of analytical pyrolysis in the field of environmental contamination by synthetic polymers was the identification of rubbers in roadway dust in 1966 [82]. Tire wear particles are a substantial source of MPs in the environment and their presence in air dust was investigated by several research groups by means of analytical pyrolysis [83]; they are difficult to measure with spectroscopic techniques. The main pyrolysis products indicative of the principal elastomers in cars (styrene-butadiene rubber, SBR) and truck tire tread (natural rubber, NR; butadiene rubber, BR) were reported to be the monomers, styrene (SBR), isoprene (NR), butadiene (BR, SBR), and the dimers, vinylcyclohexene (BR, SBR) and dipentene (NR). More recently, TED-GC-MS analysis of reference elastomers for the selection of specific markers to target the characterisation of street runoff samples was studied [64]. The presence of non-visible traces of PS in soil samples, highlighted by Py-GC-MS, was described in 1986 [84]: the identification, among the pyrolysis products, of styrene, methylstyrenes and dimethylstyrenes were indicative of traces of PS, resulting that plastic particles had already been reported as an emerging contaminant in the environment [85].

In 1998 Fabbri et al. focused on determining the presence of PS and PVC by Py-GC-MS analysis of sediment samples [73] and of extracts in organic solvent [72]. Sediment samples were collected from a coastal lagoon on the Adriatic Sea (Italy) impacted by the industrial production of synthetic polymers. Py-GC-MS revealed a specific case of pollution by resin pellets. Plastic pellets are solid particles ranging in size between 1- 5 mm which raised environmental concerns in the early 1990s [86] and are today referred to as primary MPs. PS was identified on the basis of intense signals for styrene and α-methylstyrene, together with the presence of characteristic PS pyrolysis products (styrene-dimer, biphenyl, diphenylpropane). PVC presence was indicated by an intense benzene peak and confirmed by the presence of chlorobenzene. A critical issue emerged in the quantitation and also in the qualitative assessment due to the fact that the most abundant pyrolysis products of PS and PVC (styrene and benzene respectively) are poorly specific, while the most specific markers, styrene-dimer and chlorobenzene, have a low pyrolysis yield and thus their use leads to poor detection limits (see Section 2.3.3). The same research group reported the identification of a larger set of synthetic polymers in the dichloromethane extracts of lagoon

(24)

16

sediments [87] through the identification of characteristic pyrolysis products: polybutadiene (PB), poly(acrylonitrile-co-styrene-co-butadiene) (ABS), styrene-butadiene random (SBR) and block (SBS) copolymers, and polyvinyl acetate (PVA), in addition to PVC, and PS. The specific pyrolysis products exploited as markers were chlorobenzene for PVC, acetic acid for PVA, benzene-butanenitrile for ABS, and cyclohexenylbenzene for styrene–butadiene rubbers. Fries et al. used Py-GC-MS to analyse individual marine MP particles from sediment samples from the island of Norderney (Germany) [66]. In particular, MPs were analysed after density separation with NaCl and NaI [65], and selection through optical microscopy.

Their work exploited the potential of fractionated Py and GC-MS to rapidly identify not only the polymer type, but also the associated organic plastic additives (OPAs) [66, 88] in one single analytical run. Thermal desorption at 350 °C was used for the analysis of OPAs followed by pyrolysis at 700 °C of the same particle. The particles were identified as PE, PP, PS, PA, chlorinated PE (CPE) and chlorosulfonated PE; MPs containing CPE and chlorosulfonated PE were identified for the first time in a marine environment. The identified OPAs were phthalates, benzaldehyde and 2,4-di-tert-butylphenol. The analyses were complemented by scanning electron microscopy (SEM) determination of the inorganic plastic additives (IPAs). This analytical approach was successfully exploited to investigate the spatial distribution of small MPs and their correlation with macroscopic/visible plastic debris in beach sediments (Germany) [89], demonstrating that the presence of macroscopic plastic debris was not significantly correlated with the occurrence of small plastic particles. Interestingly, fibers were not investigated due to their presence in procedural blanks (see Section 2.3.3). Besides, enabling the simultaneous identification of polymer types and OPAs at a molecular level, these first studies highlighted the advantages of desorption-GC-MS of additives, avoiding the use of solvents, compared to the analysis of additives based on solvent extraction.

Hendrickson et al. [90] used Py-GC-MS and Fourier transform infrared spectroscopy (FTIR) to investigate the distribution of MP pollution in surface waters of the Western Lake Superior (Canada). Samples were subjected to an oxidation step with H2O2 and then to a density

separation step with NaCl solution to isolate lighter plastic particles. Single plastic particles were selected for the analysis after microscopy morphological examination. The particles analysed by Py-GC-MS were identified as PVC, PE, PP, PET, CPE and PS. Incongruence between the two analytical techniques was found in some cases in the identification of CPE, PE and PVC , which the authors attributed to the heterogeneous chlorine content derived from the chlorination of PE and PVC and/or to the presence of copolymers.

An algorithm for the Py-GC/MS data processing and the automated identification of eleven types of synthetic polymers in MP samples containing plastic mixtures has been recently proposed [62].

(25)

17

Figure 2.6 Molecular structures of polymers investigated in MPs analysis and of their main pyrolysis products, with corresponding m/z (TMAH: pyrolysis products produced in the presence of tetramethylammonium hydroxide) [12].

(26)

18

Peters et al. [52] characterised single MP particles ingested by 1381 marine fish from six different species by washing their stomach contents with distilled deionized water through four filters, adapting a procedure previously developed for the analysis of the stomach content of sunfish in 2016 [91]. The composition of the MP particles analysed resulted to be: 44.1% PVC and PET, 2.3% epoxy resin, 2.3% silicone, 9.3% nylon and, interestingly, 42% of the examined particles were classified as unknown, due to a low pyrolytic abundance or to the lack of any clear polymer match. A large fraction of the unknown particles showed different morphologies but produced similar pyrograms (which included the presence of diethylphthalate among the pyrolysis products) suggesting a shared compositional origin. The authors hypothesized that this group of particles with a pyrolytic profile that was difficult to interpret could have been contaminated by - or originated from - petroleum waste or coal tar. This can be linked to the potential of the polymers to absorb environmental pollutants.

Ter Halle et al. [60] recently performed analysis of the nanoplastic (< 1 m) colloidal fraction of marine water selected by ultrafiltration through a 1.2 μm poly(ether sulfone) membrane and characterized by dynamic light scattering before Py-GC-MS analysis. The ultrafiltered colloidal fraction obtained was freeze-dried prior to Py-GC-MS analysis in different conditions: pyrolysis at 700 °C, thermodesorption at 300 °C, and thermochemolysis at 400 °C with tetramethylammonium hydroxide (TMAH). No pretreatment was applied to eliminate the organic matrix contribution. The authors used principal component analysis (PCA) of the hydrocarbons and aromatic pyrolysis profiles of the samples to obtain the composition of the mixture of polymers present in the nanocolloidal fraction. The study allowed the comparison of the pyrolysis response of hydrocarbons and aromatic compounds of different size fractions of plastic debris collected in the North Atlantic subtropical gyre: meso- (5-200 mm range), large micro- (1-5 mm range), small micro- (< 1 mm) and nanoplastics (1-999 nm). The pyrolytic signals of PE were observed to change with decreasing debris size, which could be related to aging and weathering effects.

Mintenig et al. used Py-GC-MS to identify PS in size fraction from the Asymmetrical Flow Field Flow Fractionation (A4F) of water samples spiked with nano and microplastic of the polymer. Samples were pyrolyzed at 560 °C [92].

2.3.3 Quantitative analysis

When analytical pyrolysis is used to perform quantitative analysis, the intensity of the instrumental signals associated to specific pyrolysis products of a polymer is used to determine the amount (mass) of that polymer in the sample by means of calibration protocols. Table 2.1 reports the identification and examples of concentration of total MPs obtained by means of pyrolysis in environmental samples. In 1998 Fabbri et al. studied the use of Py-GC-MS to quantify PS in sediment samples [73]. The experiments involved the quantification of PS in sediment samples without preliminary separation treatment and revealed the presence of PS at mg/g levels in the superficial layers of sediments in the

(27)

19

Ravenna Lagoon (Italy). Quantification was based on calibration curves built on peak areas of styrene obtained through SIM of the styrene molecular ion m/z 104.

Even if the use of only the styrene peak could lead to misleading results, this limitation can be partially overcome in those cases where the precise and univocal origin of the styrene peak can be assessed. In some cases this is possible on the basis of the styrene/toluene peak area ratio: e.g. the styrene/toluene ratio is reported to be in the range of 0.1-0.4 for the pyrolysis products from the most abundant natural organic materials in soil and sediments [93], while the ratio is generally higher than 1 when the styrene derives from synthetic polymers [41]. The presence or absence of butadiene marker peaks (butadiene, butadiene dimer, butadiene trimers and styrene-butadiene) – and their relative abundance in comparison with the intensity of the styrene peak – can be used to evaluate the contribution of styrene-butadiene rubbers to the plastic components. The yield of styrene depends on the molecular weight of PS. A solution to this problem can be the selection of styrene-dimer or trimer as identification and quantification markers for PS.

Since 2017 Fischer and Scholz-Böttcher [42-44] have been exploring the potential of Py-GC-MS for the quantification of synthetic polymers present as MP mixtures in different environmental samples at trace levels. In their first research, they used Curie point Py-GC-MS combined with thermochemolysis using TMAH (25% in water) and selected characteristic pyrolysis fragments for the quantification. They quantified MPs in fish samples after a pretreatment consisting in enzymatic and chemical digestion in order to remove (or at least reduce) the biological matrix and to preconcentrate potential MPs. The aim was to obtain the simultaneous, selective, and sensitive polymer-specific identification and mass quantification of MPs in environmental samples after the clean-up step [44]. In 2019 Fischer and Scholz-Böttcher [43] compared two different pyrolysis techniques, Curie point pyrolysis (CP-Py) along with thermochemolysis with TMAH and micro-furnace pyrolysis (MF-Py), for the simultaneous identification and quantification of MPs in environmental samples including sea salt, tidal flat sediments and North Sea surface water samples. All environmental samples were pre-treated to reduce the organic components that are in the matrix and which could lead to interference. These authors quantified all the polymers previously analysed [44] along with methyl dimethyl diisocyanate-polyurethane (MDI-PUR) using a microfurnace pyrolyzer (MF-Py). In [42] Py-GC-MS along with thermochemolysis (TMAH), was applied to study MP contamination in different commercially available marine salt samples from the Atlantic Ocean and Mediterranean Sea. In particular, “fleur de sel” was analysed, which is an unprocessed natural product whose crystals are directly harvested from the sea surface and was proposed by the authors as an indicator substrate for the monitoring of the MP load of coastal waters. The qualitative MP composition of sea salts has shown very distinctive regional and supra-regional patterns, which are useful for assessing MP contamination levels on both temporal and spatial scales.

(28)

20

Table 2.1 Identification and concentration of total MPs obtained by pyrolysis in environmental samples.

Origin Concentration MP type Notes Rif

Wadden Sea sediment 48-166 µg Kg-1

dw PE, PVC > PP, PS, PET,

PMMA

From replicates of

a single sample (n=4) [43] Boknaf Fjord (NO)

sediments 41-495 µg Kg

-1

dw PE, PVC, PET > PP, PA66,

PS, PMMA Range from different samples [63] Fleur de Sel (Atlantic/Mediterranean) 138-1993 µg Kg -1 PE, PP, PS, PET, PVC,

PC, PMMA, PA6, PUR

Range from different samples [42] Fleur de Sel (Mediterranean) 43 ± 12 µg per 200 g of sample PE, PP, PS, PET, PVC, PC, PMMA, PA6, PUR

From replicates of

a single sample (n=5) [42] Sea salt

(Atlantic/Mediterranean) 14-60 µg Kg

-1 PE, PP, PS, PET, PVC,

PC, PMMA, PA6, PUR

Range from

different samples [42]

Atlantic Ocean water 0.12 µg L-1 PE, PP, PS, PET, PVC,

PC, PMMA, PA6, PUR

Calculated from Fleur de Sel samples (mean value n=7)

[42]

Mediterranean

Sea water 0.20 µg L

-1 PE, PP, PS, PET, PVC,

PC, PMMA, PA6, PUR

Calculated from Fleur de Sel samples (mean value n=5)

[42]

North Sea water 0.3 µg L-1 _{PE, PP, PET, PVC, PMMA} _{One sample} _[43]

Biosolids 2.8-6.6 mg g-1 _{PE, PVC, PP, PS, PMMA}

Range from 25 samples from single wastewater treatment plant

[76]

Soil 1–86 μg g-1 _{PE, PP, PS} _{Validated on 2 samples [58]}

Soil 0.039-0.85 mg g-1 _{PE, PP, PS} _{RSD% below 10} _[59]

Ter Halle et al. [60] described a semi-quantitative approach to evaluate the relative amounts of different polymers in mixtures of MPs extracted from seawater, based on calculating the relative proportion of aromatic and aliphatic hydrocarbons in the colloidal fraction of seawater. The peak area of the selected m/z was integrated for each aromatic and aliphatic hydrocarbon and corrected by a mass spectra factor (MSF) calculated as the reciprocal of the proportion of the m/z (used for the integration) relating to the entire library mass spectra. The proportion of PVC, PS and PET in the aromatic fingerprint of nanoparticles were determined using PCA.

Selection of pyrolysis products for the identification and quantitation of MPs

Table 2.2 shows a list of pyrolytic markers that have been used for the identification and quantification of MPs. The choice of markers is crucial in quantification since the same pyrolysis products can derive from multiple origins; it follows that in some cases the markers

(29)

21

used for quantification are not the most abundant pyrolysis products as those derived from the qualitative identification.

Another issue is represented by polyolefins where a wide range of peaks can be used for the quantification. For example, the indicator ions for PE quantification proposed by Fischer et al. in [44] are m/z “83+85”, while the same authors used m/z 82 - corresponding to α,ω-alkenes - in subsequent papers [42, 43]. In addition, PE yields several homologue pyrolysis products (see Figure 2.6), thus different marker compounds have been used, including summed alkadienes [54], 1-pentadecene (more sensitive) and 1,14-pentadecadiene (more selective) [59], or a single pyrolysis product (e.g. tetradec-1-ene [63]).

Regarding PS quantification, an advantage for method sensitivity is that styrene is the pyrolysis product that gives the highest yield. However, as mentioned before styrene is not often selected as a quantitation marker because it is not a univocal pyrolysis product; in fact, it can derive from the thermal degradation of various sources including humic substances [94], vegetables and fruit tannins, degraded lignins from wood and paper, plastic waste and diesel exhaust [59, 83]. In addition, it is present among the pyrolysis products of several other synthetic polymers such as acrylonitrile-butadiene-styrene (ABS) and styrene-butadiene rubbers (SBR and SBS).

Dierkes et al. [59] monitored styrene (m/z 78) for the quantification of PS, prefixing that the results did not represent the PS concentration but a sum parameter of styrene-containing polymers. These authors tested the selectivity of indicator compounds analyzing organic matrixes that were not contaminated by plastics. They found that wood and engine oil interfered in the quantification of PS and PE, respectively. Also, fish filets can interfere in the quantification of PE because their high content of fatty acids releases alkenes and n-alkanes during thermolysis. For these reasons, 2,4-diphenyl-1-butene (styrene dimer) and 2,4,6-triphenyl-1-hexene (styrene trimer) are recommended as quantification marker peaks of PS.

Relatively polar pyrolysis products, such as those produced by PET, give broad peaks in Py-GC-MS with a non-polar GC stationary phase [45]. Different studies explored the advantages of using thermally-assisted hydrolysis and methylation (THM) with TMAH for condensation polymers and addition polymers with oxygenated side chains, in order to improve the GC performance and to increase the range of MPs detectable in a single pyrolysis run. The alkaline environment produced by TMAH promotes the chain-scission and methylation of the thermal degradation products of condensation polymers and polyesters such as polyvinyl, acryls, polycarbonated and polyamide resin. On the other hand, the pyrolysis mechanisms of polyolefins are not altered by the presence of TMAH [44]. The most abundant and/or polymer-specific compounds derived from THM-TMAH chosen as markers for polymer specific qualitative and quantitative analyses, are reported in Table 2.2 (indicated with an asterisk). Partial methylation may also complicate quantitation [44, 57]; for instance, both ε-caprolactam and its methylated derivative have been used to quantify PA6 by CP-Py [44].

(30)

22

Table 2.2 Examples of pyrolytic markers and peaks proposed for the identification and quantification of MPs. *Only in presence of TMAH. f_n-C

16-26 alkadienes used for the quantification of PE.

Polymer Identification Quantification

Pyrolysis products m/z Pyrolysis products m/z ref.

PE

alkanes (e.g. C20) 85 alkanes (e.g. C20) 83 + 85 [44]

α-alkenes (e.g. C20) 83 1-tetradecene (C14) 83 [63]

α,ω-alkenes (e.g. C20) 82 α,ω-alkenes (e.g. C20)f 82 [42, 43] 1,14-pentadecadiene 81 [59] 1-pentadecene 97 PP 2,4-dimethyl-1-heptene 70, 126 2,4-dimethyl-1-heptene 70 [42-44] 2,4-dimethyl-1-heptene 126 [59] 2,4,6,8-tetramethylundecene (isotactic) 69, 111

2,4,6,8-tetramethylundecenes (three isomers) 69, 210 [44, 63] 2,4,6,8-tetramethylundecene (heterotactic) 69, 111

2,4,6,8-tetramethylundecene (syndiotactic) 69, 111

PS

styrene 104 styrene 104 [44, 59]

2,4-diphenyl-1-butene (styrene dimer) 91 2,4-diphenyl-1-butene (styrene dimer) 208 [63] 2,4,6-triphenyl-1-hexene (styrene trimer) 91 2,4,6-triphenyl-1-hexene (styrene trimer) 91 [42-44]

PVC benzene 78 benzene 78 [42-44]

chlorobenzene 112 1-methylnaphthalene 142 [63]

PA6 ε-caprolactam 113 ε-caprolactam 113 [42-44]

N-methyl caprolactam* 113, 127 N-methyl caprolactam* 127 [42-44]

PA-66 hexene 113, 84 hexene 84 [63]

PMMA methyl methacrylate 69, 100 methyl methacrylate 100 [42-44, 63]

(31)

23

PET dimethyl terephthalate* 163, 194 dimethyl terephthalate* 163 [42-44, 63]

PC p-methoxy-tert-butylbenzene* 149, 164 p-methoxy-tert-butylbenzene* 149 [63] 2,2-bis(4'-methoxyphenyl)propane* 241, 256 2,2-bis(4'-methoxyphenyl)propane* 241 [42-44] MDI-PUR 4,4'-methylenbis(N-methylaniline)* 226 4,4'-methylenbis(N,N-dimethylaniline)* 254 [42, 43] N,N-dimethyl-4-(4-methylamino)benzylaniline* 240 4,4'-methylenbis(N,N-dimethylaniline)* 253, 254 SBR, BR vinylcyclohexene 54, 79, 93, 108 vinylcyclohexene 54 [64, 83] butadiene 39, 54 SBR styrene 51, 78, 104 methylstyrene 78, 103, 118 cyclopenthylbenzene 115, 129, 144 cyclohexenylbenzene 104, 115, 129, 158 phenyl-[4.4.0]bicyclodecene 91, 104, 156, 212 BR butadiene trimers and homologous 91, 148, 162, 176 NR

dipentene 68, 93, 121, 136

dipentene 68

isoprene trimers 119, 162, 189, 204

(32)

24 Calibration

Calibration protocols have mainly been set up by weighing single polymer particles, for instance in the 0.5-50 µg range, and, with few exceptions, Py-GC-MS has demonstrated a satisfactory linearity (R2_{0.860.99) [44, 63]. The calibration method is limited by the}

smallest particle that can be weighed by a sensitive balance and transferred into the sample holder of the pyrolysis apparatus. Under these circumstances, the limit of detection (LOD) by Py-GC-MS is below 0.1 µg and dependent on the polymer type and density. Using calibration solutions for soluble polymers can significantly reduce the LOD (to 3 ng) and the limit of quantitation (LOQ), as estimated for PS depending on the quantitation marker (e.g. LOQ 16 ng using styrene and 282 ng using a styrene trimer) [44]. The use of calibration solutions for low polymer concentrations and weighed solid MP for higher concentrations is feasible within the same calibration curve [43].

External calibration curves were obtained by adding Al2O3 as an inert dilution matrix for the

eight main common plastics PE, PP, PS, PET, PVC, PMMA, PC and polyamide 6 (PA6) [44]. It was demonstrated that internal calibration worked better than external calibration for PET. This was confirmed by David et al. [95] in the TG-MS analysis of PET in soil samples using the m/z 105 ion with LOD and LOQ of 0.07 and 1.7 wt % PET, respectively.

A major calibration issue is that the analytical measurement of pyrolysis products can be affected by many different factors such as experimental conditions, polymer properties and matrix composition. As far as the analytical parameters are concerned, pyrolysis temperature and split ratios affect the peak area, especially for PE [53]. In the case of PS, thermal extraction conditions in TED-GC-MS can modify the proportion of the abundance of the styrene monomer, dimer and trimer [45].

Matrix effects should be carefully evaluated. Pyrolysis of PS in the presence of sediment components (quartz, calcite, various clays) considerably influences the yield of pyrolysis products affecting the slope of the calibration curves [73]. Besides inorganic constituents, natural organic matter is also critical, specifically for the production of pyrolysis products that are identical to those of the target polymer, and more generally for the background contamination of the pyrolysis system which could reduce the number of samples that can be analysed sequentially [44]. Matrix effects can be mitigated by performing the calibration using the sample matrix or matrix-matching materials [73]. Dümichen et al. [54] prepared the calibration of PE in soil (0.15-5 % range) for the TED-GC-MS.

Finally, the characteristics inherent in the polymer itself should be taken into consideration. For instance, styrene yields have been shown to be dependent on the molecular weight of the polymer [96]. However according to some authors, at analytical concentrations, the molecular weight does not appear to influence the styrene peak area determined by Py-GC-MS [97]. The effect of aging did not appear to significantly change the pyrolytic behavior of the MPs [45].

(33)

25

Internal pyrolysis process standards (ISTDpy) have been proposed [42, 43, 98] to improve

quantification by avoiding variabilities within the calibration curves due to variable organic loads. The use of the proposed internal standards aims to mimic the potential interactions of polymer specific indicator compounds with pyrolytic products of residual organic sample matrices during the pyrolysis process. The internal standards proposed by Fisher et al. include: an aliphatic compound (androstane), a planar aromatic compound (9-dodecyl-1,2,3,4,5,6,7,8-octahydro anthracene (DOHA) or 9-tetradecyl-(9-dodecyl-1,2,3,4,5,6,7,8-octahydro anthracene (TOHA) and anthracene-d10_{), and a polar compound (cholanic acid) whose acid}

group undergoes methylation during thermochemolysis. Individual calibration curves were created using the ratio of the area of the preselected indicator ions (Table 2.2) to the area of the matching ISTDpy which was selected individually for each polymer, on the basis of the

coefficient of determination, r2_{, and the standard deviation, s} x0.

A completely different internal standardisation was used by Dierkes et al. [59], where a polymeric internal standard, deuterated PS-d5, was used as the quantification internal standard. The method was based on the combination of pressurized liquid extraction and Py-GC-MS, and it led to low quantification limits for environmental samples. For the calibration curves they diluted different amounts of polymers in calcined sea sand.

Gomiero et al. [63] used Py-GC-MS in combination with TMAH thermochemolysis to study sample sediments from an urban fjord in Norway after an enzymatic and peroxide treatment. In order to identify a possible trend in polymer space and size distribution, PCA was subsequently applied to the data obtained. PCA has proven effective in the interpretation of Py-GC-MS data, differentiating between sites close to urban coastal areas and more open water sites.

The significant environmental impact of the pollution caused by tire and road wear particles (TRWP) was also explored by a quantitative Py-GC-MS approach [83]. Unice et al. proposed a calibration method including deuterated polymers (PS, polyisoprene and polyisobutadiene) as internal standards in order to quantify tire tread in environmental samples. In order to assess the content of tire wear-off, Eisentraut et al. [64] quantified the amount of SBR, which is found in most of car tire samples, in reference samples, and real samples from street runoff. Specific quantification marker compounds for SBR, BR and NR are listed in Table 2.2. The focus point in the quantification of MPs in environmental and fish/biota samples is the optimization of sample pre-treatments, in order to avoid losses of analytes that could hamper or bias the obtained results.

Quality control and quality assurance

Microplastic contamination from the sampling and laboratory environment is a major issue in the analysis of MPs in environmental samples, which can interfere with both qualitative and quantitative determination. Thus, MPs analysis requires quality control and quality assurance protocols on different levels: during sampling, pre-treatments, and the analysis step. Carrying out the entire treatment, digestion (in the analysis of animal or plant tissue), separation and drying steps in the same crucible or vial is advisable when possible, and in

Analytical pyrolysis for the characterization of microplastics and their degradation products in the environment

UNIVERSITÀ DI PISA

Dipartimento di Chimica e Chimica Industriale

Corso di Laurea Magistrale in Chimica

Curriculum Analitico

CLASSE: LM-54

Tesi di Laurea

Analytical pyrolysis for the characterization of microplastics

and their degradation products in the environment

Relatore Controrelatore

Prof.ssa Francesca Modugno Dr. Tommaso Lomonaco

Candidata

Greta Biale

Contents

1. Introduction

2. State of the art