Contributions of Fourier transform infrared spectroscopy in microplastic pollution research: A review

Abstract

Fourier transform infrared (FTIR) spectroscopy has been extensively used in microplastic (MP) pollution research since 2004. The aim of this review is to discuss and highlight the recent advances in FTIR (spectroscopy and chemical imaging) techniques that are used to characterize various polymer types of MPs and to trace their fate and transport in different environmental matrices. More than 400 research papers dealing with FTIR techniques in MP pollution research, which are published between January 2010 and December 2019, have been identified from the Scopus and Web of Science databases. The MPs present in sediment, water (marine and freshwater), biota, air/dust, waste water treatment plants and salt are further classified according to (1) characterization and identification, (2) weathering and aging, (3) ecotoxicology, and (4) analytical methods. The results revealed that the ATR-FTIR technique is mostly used to identify and characterize the MPs found in water and sediment. The µFTIR (FTIR imaging) is extensively used to study the ingestion of MPs in biota (both marine and freshwater). In this article, we have summarized the current knowledge of application of FTIR spectroscopy to MP research and provided insights to future challenges for understanding the risk of MPs.

Graphical Abstract

Keywords:

• Characterization
• FTIR
• microplastics

1. Introduction

Microplastic (MP) contamination in water (both marine and fresh water) (Gago et al., 2016; Koelmans et al., 2019), sediment (Cauwenberghe et al., 2015; Xu, Liu, et al., 2019), air (Zhang et al., 2020), salt (Peixoto et al., 2019), biota (Rezania et al., 2018; O'Connor et al., 2020; Markic et al., 2020), wastewater treatment plant (Sun, Dai, et al., 2019) and honey (Muhlschlegel et al., 2017) has aroused as a major global environmental and economic issue. The term “microplastics” was first coined by Thompson (Thompson et al., 2004) in 2004 to describe the small size (less than 5 mm in size) plastic particles in the oceans. Since then, significant amount of results has been published on this topic globally. Plastic debris in the environment is classified based on their chemical composition, solid state, solubility, size, shape, color and origin. Depending on size, plastics have been categorized as nanoplastics (1 to <1000 nm), MPs (1 to <1000 µm), mesoplastics (1 to <10 mm) and macroplastics (1 cm and larger) (Hartmann et al., 2019). MP consists of primary MPs, which are manufactured in microscopic size for specific purpose (including microbeads) and secondary MPs, which are formed from large plastic debris degraded and fragmented by long-term physical, chemical and biological effects in the environment (Cole et al., 2011; Veerasingam et al., 2017; Besseling et al., 2019).

Weathering process in MPs (both primary and secondary) is first initiated at the surface of particles, where the surface layer is discolored, oxidized, embrittled and crazed (Bond et al., 2018). Then, the interior weathering is proceeded by a diffusion-controlled process. MP undergoes degradation due to physical (wind wave action and mixing in-between the heavy and/or abrasive beach sediments) (Efimova et al., 2018; Chubarenko et al., 2020), chemical (UV light from solar radiation) (Andrady, 2011, 2017; Song et al., 2017) and biological (microorganisms) (Kooi et al., 2017) factors. Degradation processes of MPs are categorized as follows: photo degradation (action of light or photons, usually UV light from the sun), thermal degradation (high temperature causes molecular deterioration in laboratory condition), thermo-oxidative degradation (slow oxidative, molecular deterioration at moderate temperatures), hydrolysis (reaction with water) and biodegradation (decomposition of organic materials by microorganisms). Due to weathering process, MPs change its physical properties such as mechanical, optical or electrical characteristics in crazing, cracking, erosion, discoloration and phase separation. Degradation of MPs depends on many factors including polymer type, age, and environmental conditions such as sunlight, temperature, rain, humidity, irradiation, pH, pollutants, thermal cycles, and oxygen content (Halle et al., 2016; Smith et al., 2018). Weathering of MPs in water is much slower than that in air or on beaches, as water suppresses light-induced oxidative degradation. This can be attributed to low temperatures, low oxygen concentrations and reduced transmittance of UV irradiation in water as well as increased biofilm formation (Andrady, 2011, 2017; Bond et al., 2018). Due to higher rate of photo-oxidation process, plastics in the hot sand beaches degraded faster than those floating in the ocean water (Veerasingam, Mugilarasan, et al., 2016; Sathish et al., 2019).

Since the size of MPs is very small, it can be ingested by a wide range of organisms by mistake as food for prey (Botterell et al., 2019). MPs are not only contaminants by themselves, they are also associated with different chemical additives, which were added during their manufacturing process to optimize their physical attributes. These chemicals, including those incorporated during plastic production (additives), can leach into biological tissues, posing health risk to organisms and bio-accumulating in the food chain (Koelmans, 2015; Gall & Thompson, 2015; Galloway et al., 2017). Therefore, MPs have the potential to act as vectors for the transport of hydrophobic organic pollutants (Rochman, 2015). The vector effect of particle mediated transport of pollutants can be divided into three groups (Syberg et al., 2015): (1) an environmental-vector effect (MPs with adhered pollutants are transported through the environment), (2) an organismal-vector effect (the pollutant is transported into the organisms) and (3) a cellular-vector effect (the pollutant is transported with the particle into cells).

Many analytical methods have been used to identify polymeric composition of plastic debris in different aquatic environments (summarized in Table 1). These methods are time consuming, involving laborious sample preparation, and destructive in nature. Moreover, most of these methods are limited to volatile or ionizable compounds such as small oligomeric fragments or additives within the bulk material. FTIR and Raman techniques are versatile vibrational spectroscopic methods used to characterize different types of polymers (Kappler et al., 2016; Jung et al., 2018). In fact, FTIR spectroscopy can identify all the molecular and functional groups present in plastic polymers (Bhargava et al., 2003; Mecozzi et al., 2016).

Table 1. Analytical methods for the characterization of MPs.

Technique	Purpose	Nature of technique	Reference
Thermogravimetric analysis (TGA)	Determines thermal degradation pathway of microplastics	Destructive for sample	Yu et al. (2019)
Differential scanning calorimetry (DSC)	Determines endothermic phase transition, melting characteristics, reaction kinetics of microplastics	Destructive for sample	Majewsky et al. (2016)
Thermogravimetric analysis—solid-phase extraction process/thermal desorption gas chromatography mass spectrometry (TGA-SPE/TDS-GC-MS)	Enables unambiguous and convenient detection of characteristic decomposition products of microplastics	Destructive for sample	Dumichen et al. (2015)
X ray diffraction	Structural and functional group analysis, degree of crystallinity	Destructivity depends on sample preparation method used	Ariza-Tarazona et al. (2019)
Nuclear magnetic resonance (NMR)	Structural and functional group analysis, identification of exact structure, chemical moieties and conformational state	Laborious sample preparation, nondestructive for sample	Peez et al. (2019)
Scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS)	Characterization of surface structure and elemental composition in microplastics	Destructive for sample	Wang, Wagner, et al. (2017)
Gel permeation chromatography (GPC) with fluorescence detection (FLD)	Semi-quantitative selective determination of microplastics	Laborious sample preparation, destructive for sample	Biver et al. (2018)

FTIR spectroscopy deals with measurement of infrared (IR) radiation absorbed by the MP sample, allowing the study of molecular composition. An infrared spectrum represents a fingerprint of a sample (MP) with absorption peaks correspond to the frequencies of vibration between the bonds of the atoms making up the material. Because each different polymer material is a unique combination of atoms, no two compounds produce exactly the same infrared spectrum. Therefore, the chemical structure of a polymer molecule can be determined by FTIR (Chalmers, 2006). The IR region of the electromagnetic spectrum is divided into three regions: (1) the higher energy near infrared (NIR) region with wavenumbers of 14,000–4000 cm⁻¹ (0.78–2.5 µm wavelength) range, which is sensitive to overtone and combinations of vibrations, (2) the mid infrared (MIR) region with wavenumbers of 4000–400 cm⁻¹ (2.5–30 µm wavelength) range to study the fundamental vibrations and (3) the far infrared (FIR) region with wavenumbers of 400–10 cm⁻¹ (30–1000 µm wavelength) range to study rotations (Mukherjee & Gowen, 2015). Among these three IR spectral regions, MIR is the most common region in the field of MP characterization. FTIR spectroscopy can be used to study the solid, liquid and gaseous samples. Since MPs are generally solid samples, we will focus our discussion on these materials. FTIR spectra of different polymers are covered in this review to illustrate the key features involved in spectral interpretation and its application for the analysis of MPs.

The objectives of this article are (1) to critically review the work of identification of MPs using FTIR technique in various environmental matrices, (2) to highlight the suitable evaluation and data processing methods for MPs identification based on FTIR measurements, and (3) to look at the current limitations in the methods and analyses, and how do we improve and harmonize the practices for future studies of MPs using FTIR measurements.

2. Data collection

A comprehensive literature search was conducted using a combination of keywords including “microplastic,” “microplastics,” “plastic,” “FTIR,” “sediment,” “water,” “waste water treatment plant,” “atmosphere,” “dust,” “biota,” “ingestion,” “salt” in Scopus and Web of Science database. The retrieved articles were screened, and only those studies carried out with FTIR technique were selected. This has resulted into 412 research articles, published between January 2010 and December 2019, and these were taken-up for detailed review. These research articles are categorized according to the topics of investigation: (1) characterization and identification, (2) weathering and aging, (3) ecotoxicology, and (4) analytical methods and future challenges. The statistics show that the number of publications pertaining to MP studies using FTIR technique has increased rapidly in the last 5 years (Figure 1).

Figure 1. Literature search results on microplastics using FTIR displayed as the number of publications from January 2010 to December 2019. The published papers were further classified into various environmental matrices (sediment, water, waste water treatment plants, biota, dust/air, and salt).

3. Characterization and identification

Characterization and identification of different polymer types of MPs found in a sample are important as each polymer has its own impacts on the biosphere. MP pollution has been studied in various environmental matrices (sediment, water, biota, salt and air/dust) using FTIR technique around the world (Table 2). However, only few studies have been carried out using FTIR technique so far to assess the MPs in air (dust) samples. Most frequently obtained shapes of MPs are fibers, fragments, films, foams, microbeads and pellets. Abundance of secondary MPs is higher (fragments, fibers, and foams) than the primary MPs (microbeads and pellets). MPs collected from various environmental matrices might be linked with different organic pollutants, and that may affect the identification of their polymer types. Therefore, purification of MPs is the key to instrumental analysis. Various digestion methods including oxidant digestion (reagent: 30% H₂O₂/35% H₂O₂/5% NaCIO), acid digestion (reagent: 65% HNO₃/37% HCl), alkaline digestion (reagent: 10 M NaOH/10% KOH) and enzymatic digestion (reagent: trypsin/proteinase K) are used for purification (Lusher et al., 2017). The review analysis shows that various polymer types have been covered in the studies—polyethylene (PE), polypropylene (PP), polystyrene (PS), polyamide (PA), and polyethylene terephthalate (PET). PE and PP are the dominant polymer types found in all environmental matrices. The MPs extracted from sediment, water, biota, waste water treatment plants, salt, and air (dust) are grouped into two categories after visual identification: (1) large MPs (5 mm–500 µm) and (2) small MPs (size <500 µm) (Eo et al., 2018; GESAMP, 2019; Hidalgo-Ruz et al., 2012; Lorenz et al., 2019; Olesen et al., 2019; Song, Hong, Jang, Han, Rani, et al., 2015).

Table 2. Comprehensive list of research papers dealing with microplastic analysis using FTIR, as of December 2019.

Reference guide for works on FTIR identification of microplastics
Sediment	Abidli et al., 2018; Acosta-Coley, Cuadro, et al., 2019; Acosta-Coley, Duran-Izquierdo, et al., 2019; Acosta-Coley and Olivero-Verbel, 2015; Alvarez-Hernandez et al., 2019; Alves and Figueiredo, 2019; Antunes et al., 2013; Ashwini and Varghese, 2019; Atwood et al., 2019; Bancin et al., 2019; Battulga et al., 2019; Bayo et al., 2019; Blaskovic et al., 2018; Blettler and Ulla, 2019; Blumenroder et al., 2017; Brandon et al., 2019; Camacho et al., 2019; Carson et al., 2011; Ceccarini et al., 2018; Chen et al., 2018; Chen et al., 2019; Cheung et al., 2016; Chouchene et al., 2019; Claessens et al., 2011; Constant et al., 2019; Coppock et al., 2017; Corcoran et al., 2015; Crichton et al., 2017; Dai et al., 2018; Dikareva and Simon, 2019; Dowarah and Devipriya, 2019; Edo et al., 2019; Eo et al., 2018, 2019; Fan et al., 2019; Filgueira et al., 2019; Fok et al., 2017; Frias et al., 2010; Garaba and Dierssen, 2018; Garces-Ordonez et al., 2019; Graca et al., 2017; Gray et al., 2018; Haave et al., 2019; Hanachi et al., 2019; Harrison et al., 2012; Herrera et al., 2018; Horn et al., 2019; Hu et al., 2018; Imhof et al., 2017; Jiang et al., 2018; Karkanorachaki et al., 2018; Karthik et al., 2018; Kim et al., 2015; Klein et al., 2015; Koongolla et al., 2018; Korez et al., 2019; Krishnakumar et al., 2018; Kunz et al., 2016; Leads and Weinstein, 2019; Leslie et al., 2017; Li, Zhang, et al., 2018; Li, Wu, et al., 2019; Lin et al., 2018; Liu, Lu, et al., 2018; Liu, Song, et al., 2019; Lo et al., 2018; Lorenz et al., 2019; Lourenco et al., 2017; Lv, Zhou, et al., 2019; Mani, Primpke, et al., 2019; Martins and Sobral, 2011; Martin et al., 2017; Masia et al., 2019; Matsuguma et al., 2017; Miller et al., 2017; Mistri et al., 2017, 2018; Mohsen et al., 2019; Mu, Qu, et al., 2019; Munari et al., 2017; Nabizadeh et al., 2019; Naidoo et al., 2015; Naji, Esmaili, and Khan, 2017; Naji, Esmaili, Mason, et al., 2017; Naji et al., 2019; Nakao et al., 2020; Nor and Obbard, 2014; Pagter et al., 2018; Palombini et al., 2018; Pannetier et al., 2019; Peng et al., 2017, 2018; Phuong, Poirier, et al., 2018; Piehl et al., 2019; Pinon-Colin et al., 2018; Piperagkas et al., 2019; Qin et al., 2020; Qiu et al., 2015; Reed et al., 2018; Rodrigues, Abrantes, et al., 2018; Ruiz-Compean et al., 2017; Sagawa et al., 2018; Saliu et al., 2018; Sanchez-Vidal et al., 2018; Sarkar et al., 2019; Sathish et al., 2019; Scheurer and Bigalke, 2018; Scopetani et al., 2019; Scott et al., 2019; Shim et al., 2016; Simon-Sanchez et al., 2019; Song, Hong, Jang, Han, Rani, et al., 2015; Su et al., 2016, 2018; Su, Zhang, et al., 2019; Syakti et al., 2017; Tang et al., 2018; Tiwari et al., 2019; Toumi et al., 2019; Tsang et al., 2017; Turner, 2018; Turner and Holmes, 2011; Turner et al., 2019; Tziourrou et al., 2019; Vaughan et al., 2017; Veerasingam, Mugilarasan, et al., 2016; Veerasingam, Saha, et al., 2016; Velez et al., 2019; Vianello et al., 2013; Vidyasakar et al., 2018; Wang, Peng, et al., 2017, Wang, Wang, et al., 2019; Wang, Zhou, et al., 2018; Wang, Su, et al., 2018, 2019; Wu, Zhang, et al., 2019; Yabanli et al., 2019; Yao et al., 2019; Yu et al., 2016, 2019; Zbyszewski et al., 2014; Zhang, Wu, et al., 2019; Zhang, Zhou, et al., 2019; Zhang et al., 2018; Zhang, Zhao, et al., 2019; Zhao et al., 2018, 2019; Zheng et al., 2019; Zhou et al., 2018; Zhu, Wang, et al., 2018
Seawater, freshwater, and bottled water	Abayomi et al., 2017; Aliabad et al., 2019; Amelineau et al., 2016; Atwood et al., 2019; Baini et al., 2018; Barrows, Cathey, et al., 2018; Barrows, Christiansen, et al., 2018; Brandon et al., 2016; Cabernard et al., 2018; Cai et al., 2018; Caldwell et al., 2019; Castillo et al., 2016; Castro et al., 2016; Chae et al., 2015; Chen et al., 2018, 2019; Cheung and Fok, 2016; Cheung et al., 2018, 2019; Cordova et al., 2019; Courtene-Jones, Quinn, Gary, et al., 2017; Dai et al., 2018; Dikareva and Simon, 2019; Ding et al., 2019; Dris et al., 2018; Eo et al., 2019; Fan et al., 2019; Gallagher et al., 2016; Gasperi et al., 2014; Gewert et al., 2017; Goldstein et al., 2013; Gray et al., 2018; Green et al., 2018; Gündoğdu et al., 2017, 2018; Halle et al., 2016; Hendrickson et al., 2018; Hu et al., 2018; Isobe, 2016; Isobe et al., 2014, 2015, 2017, 2019; Iwasaki et al., 2017; Jahan et al., 2019; Kang et al., 2015; Kanhai et al., 2017, 2018; Kataoka et al., 2019; Kedzierski et al., 2019; Koongolla et al., 2018; Kosore et al., 2018; Kroon, Motti, Talbot, et al., 2018; Kuklinski et al., 2019; Lacerda et al., 2019; Lahens et al., 2018; Lebreton et al., 2018; Lefebvre et al., 2019; Leslie et al., 2017; Li, Green, et al., 2018; Li, Wolanski, et al., 2018; Lin et al., 2018; Liu, Olesen, et al., 2019; Lorenz et al., 2019; Luo et al., 2019; Lusher et al., 2015; Lv, Zhou, et al., 2019; Mai et al., 2018, 2019; Mani et al., 2015; Mani, Blarer, et al., 2019; Martin et al., 2017; Mason et al., 2016; Mauro et al., 2017; Mintenig et al., 2019; Mohd Khalik et al., 2018; Mu, Zhang, et al., 2019; Naidoo et al., 2015; Obbard et al., 2014; Olivatto et al., 2019; Pedrotti et al., 2016; Peeken et al., 2018; Pivokonsky et al., 2018; Qu et al., 2018; Reisser et al., 2013; Rivers et al., 2019; Rodrigues, Abrantes, et al., 2018; Rodrigues, Goncalves, et al., 2018; Rose and Webber, 2019; Rudduck et al., 2017; Sagawa et al., 2018; Saliu et al., 2018, 2019; Savoca et al., 2019; Scheurer and Bigalke, 2018; Scott et al., 2019; Sighicelli et al., 2018; So et al., 2018; Song et al., 2014; Song, Hong, Jang, Han, Rani, et al., 2015; Song, Hong, Jang, Han, and Shim, 2015; Song et al., 2018; Su et al., 2016, 2018; Suaria et al., 2016; Sun, Liang, et al., 2018; Syakti et al., 2017, 2018; Tan et al., 2019; Tang et al., 2018; ter Halle et al., 2017; Tsang et al., 2017; Tuncer et al., 2018; Uurasjarvi et al., 2020; Vianello et al., 2018; Virsek et al., 2017; Wang, Zhou, et al., 2018; Wang, Ndungu, et al., 2017; Whitaker et al., 2019; Xu et al., 2018; Yu et al., 2019; Zeri et al., 2018; Zhang, Zhang, et al., 2019; Zhang et al., 2015, 2017; Zheng et al., 2019; Zhu, Zhang, et al., 2019; Zhu, Bai, et al., 2018
Biota	Abidli et al., 2019; Akindele et al., 2019; Alomar and Deudero, 2017; Alvarez et al., 2018; Andrade, Winemiller, et al., 2019; Avery-Gomm et al., 2016; Avio et al., 2015, 2017; Axworthy and Padilla-Gamino, 2019; Baalkhuyur et al., 2018; Bernardini et al., 2018; Bessa et al., 2018, 2019; Besseling et al., 2015; Biginagwa et al., 2016; Birnstiel et al., 2019; Blettler et al., 2019; Bour et al., 2018; Brate et al., 2016; Brate, Blazquez, et al., 2018; Brate, Hurley, et al., 2018; Calderon et al., 2019; Caron et al., 2018; Carreras-Colom et al., 2018; Catarino et al., 2018; Cau et al., 2019; Chan et al., 2019; Cho et al., 2019; Cole et al., 2014; Courtene-Jones, Quinn, Gary, et al., 2017; Courtene-Jones, Quinn, Murphy, et al., 2017; Courtene-Jones et al., 2019; Deng et al., 2018; Didier et al., 2017; Digka et al., 2018; Ding et al., 2018, 2019; Donohue et al., 2019; Doyle et al., 2019; Fang et al., 2018; Felice et al., 2018; Feng et al., 2019; Figueiredo and Vianna, 2018; Foekema et al., 2013; Fossi et al., 2017; Garcia-Garin et al., 2019; Gassel and Rochman, 2019; Gomiero et al., 2019; Hall et al., 2015; Halstead et al., 2018; Hasselerharm et al., 2018; Hermsen et al., 2017; Hipfner et al., 2018; Hodson et al., 2017; Hossain et al., 2019; Hurley et al., 2017; Iannilli et al., 2018, 2019; Jabeen et al., 2017; Jahan et al., 2019; Jang et al., 2018; Jemec et al., 2016; Karthik et al., 2018; Khan et al., 2015, 2017; Kokalj et al., 2018; Kolandhasamy et al., 2018; Kroon, Motti, Jensen, et al., 2018; Kuhn, Oyen, et al., 2018; Kuhn, Schaafsma, et al., 2018; Kumar et al., 2018; Lagarde et al., 2016; Lefebvre et al., 2019; Leslie et al., 2017; Li, Ma, et al., 2018; Li et al., 2015, 2016; Li, Green, et al., 2018; Li, Su, et al., 2019; Li, Wu, et al., 2019; Lourenco et al., 2017; Lusher et al., 2013, 2015; Lv, Zhou, et al., 2019; Markic et al., 2018; McGoran et al., 2017, 2018; Mohsen et al., 2019; Morgana et al., 2018; Murphy et al., 2017; Naidoo et al., 2017; Naji et al., 2018; Nelms et al., 2018; Nelms, Barnett, et al., 2019; Nelms, Parry, et al., 2019; Neves et al., 2015; Ory et al., 2018; Patterson et al., 2019; Pegado et al., 2018; Pellini et al., 2018; Pham et al., 2017; Phillips and Bonner, 2015; Phuong, Poirier, et al., 2018; Piarulli et al., 2019; Pozo et al., 2019; Qu et al., 2018; Reichert et al., 2019; Renzi, Grazioli, and Blaskovic, 2019; Renzi, Specchiulli, et al., 2019; Ribeiro et al., 2017; Roch and Brinker, 2017; Rodriguez-Seijo et al., 2018; Rummel et al., 2016; Savoca et al., 2019; Scott et al., 2019; Slootmaekers et al., 2019; Song et al., 2019; Steer et al., 2017; Su et al., 2016, 2018; Su, Deng, et al., 2019; Su, Nan, et al., 2019; Sun et al., 2017; Sun, Liang, et al., 2018; Sun, Chen, et al., 2018; Sun, Liu, et al., 2018; Sun, Li, et al., 2019; Tanaka and Takada, 2016; Teng et al., 2019; Tian et al., 2017; Ugolini et al., 2013; Valente et al., 2019; Verlaan et al., 2019; von Friesen et al., 2019; Wagner et al., 2019; Wang, Wang, et al., 2019; Wang, Mao, et al., 2019; Watermann et al., 2017; Webb et al., 2019; Weber et al., 2018; Welden et al., 2016, 2018; Wen et al., 2018; Wesch et al., 2016; Wu, Tao, et al., 2019; Yu et al., 2018; Zhang, Man, et al., 2019; Zhang, Wang, et al., 2019; Zhu, Li, et al., 2019; Zhu, Chen, et al., 2018; Zhu, Yu, et al., 2019; Zhu, Zhang, et al., 2019; Zhu, Wang, et al., 2019
Waste water treatment plants	Blair et al., 2019; Carr et al., 2016; Conley et al., 2019; Dyachenko et al., 2017; Gies et al., 2018; Kalcikova et al., 2017; Lares et al., 2018, 2019; Lee and Kim, 2018; Leslie et al., 2017; Li, Chen, et al., 2018; Li, Mei, et al., 2019; Lv, Dong, et al., 2019; Magni et al., 2019; Mahon et al., 2017; Mintenig et al., 2017; Murphy et al., 2016; Tagg et al., 2015; Talvitie, Mikola, Koistinen, et al., 2017; Talvitie, Mikola, Setala, et al., 2017; Xu, Jian, et al., 2019; Yang et al., 2019; Ziajahromi et al., 2017
Air/dust	Cai et al., 2019; Dris et al., 2016, 2017; Kaya et al., 2018; Liu, Li, et al., 2019; Liu, Wang, Fang, et al., 2019; Liu, Wang, Wei, et al., 2019; Vianello et al., 2019; Wang, Li, et al., 2019; Wesch et al., 2017
Salt	Iniguez et al., 2017; Karami et al., 2017; Kim et al., 2018; Lee et al., 2019; Renzi and Blaskovic, 2018; Renzi, Grazioli, et al., 2019; Seth and Shriwastav, 2018; Yang et al., 2015

FTIR is a well recognized, rapid and quite reliable method to identify polymer types of different MPs by comparing the resulting FTIR spectra with known plastic polymers in the spectral library (Table 3). Usually, larger MPs are identified by ATR-FTIR, whereas smaller MPs (particle down to 20 µm) are characterized by μFTIR (Okoffo et al., 2019). Single particles with size larger than 200–300 µm are commonly presorted using visual examination before FTIR analysis (Primpke, Christiansen, et al., 2020). Among the research articles considered for this review, nearly 60% of studies used ATR-FTIR technique to characterize the large single particle MPs extracted from different environmental matrices (Table 4) since it is a cost-efficient method and no sample preparation or complicated mathematical correction is required. Though reflection is fully nondestructive (Primpke, Christiansen, et al., 2020), ATR may destroy MPs due to pressure applied (Wu, Tao, et al., 2019). In some studies, the larger single MPs were ground with potassium bromide (KBr) for transmission measurements either by diffuse reflection or via ATR measurements (Renner et al., 2016). However, in these techniques, the working time per particle is high (∼3 min per particle). Aforementioned techniques require sample transport to a laboratory. In a few studies, handheld FTIR technique was used for the direct measurement of MPs in the field (Abidli et al., 2018; Tang et al., 2019). However, the handheld FTIR system is more expensive than the benchtop FTIR instruments. Handheld FTIR systems are commonly limited to characterizing the MP particles with size of 300–500 µm (Primpke, Christiansen, et al., 2020).

Table 3. FTIR characteristic peak assignments for various types of MPs.

S. No.	Polymer	Characteristic peaks (cm^−1)	Assignment	Reference
1	High density polyethylene (HDPE)	2915 2845 1472 1462 730 717	C–H stretching C–H stretching CH2 bending CH2 bending CH2 rocking CH2 rocking	Nishikida and Coates, 2003; Noda et al., 2007; Asensio et al., 2009; Mecozzi et al., 2016; Jung et al., 2018
2	Low density polyethylene (LDPE)	2915 2845 1467 1462 1377 730 717	C–H stretching C–H stretching CH2 bending CH2 bending CH2 bending CH2 rocking CH2 rocking	Nishikida and Coates, 2003; Noda et al., 2007; Asensio et al., 2009; Mecozzi et al., 2016; Jung et al., 2018
3	Polyethylene terephthalate (PET)	1713 1241 1094 720	C = O stretching C–O stretching C–O stretching Aromatic CH out-of- plane bending	Verleye et al., 2001; Noda et al., 2007; Asensio et al., 2009; Mecozzi et al., 2016; Jung et al., 2018
4	Polypropylene (PP)	2950 2915 2838 1455 1377 1166 997 972 840 808	C–H stretching C–H stretching C–H stretching CH2 bending CH3 bending CH bending, CH3 rocking, C–C stretching CH3 rocking, CH3 bending, CH bending CH3 rocking, C–C stretching CH2 rocking, C–CH3 stretching CH2 rocking, C–C stretching, C–CH stretching	Verleye et al., 2001; Noda et al., 2007; Asensio et al., 2009; Mecozzi et al., 2016; Jung et al., 2018
5	Polystyrene (PS)	3024 2847 1601 1492 1451 1027 694 537	Aromatic C–H stretching C–H stretching Aromatic ring stretching Aromatic ring stretching CH2 bending Aromatic CH bending Aromatic CH out-of- plane bending Aromatic ring out-of- plane bending	Verleye et al., 2001; Noda et al., 2007; Asensio et al., 2009; Mecozzi et al., 2016; Jung et al., 2018
6	Polyvinyl chloride (PVC)	1427 1331 1255 1099 966 616	CH2 bending CH bending CH bending C–C stretching CH2 rocking C–Cl stretching	Beltran and Marcilla, 1997; Verleye et al., 2001; Noda et al., 2007; Jung et al., 2018
7	Polyurethane (PU)	2865 1731 1531 1451 1223	C–H stretching C = O stretching C–N stretching CH2 bending C(=O)O stretching	Verleye et al., 2001; Noda et al., 2007; Asefnejad et al., 2011; Jung et al., 2018
8	Nylon (all polyamides)	3298 2932 2858 1634 1538 1464 1372 1274 1199 687	N–H stretching CH stretching CH stretching C = O stretching NH bending, C–N stretching CH2 bending CH2 bending NH bending, C–N stretching CH2 bending NH bending, C = O bending	Rotter and Ishida, 1992; Verleye et al., 2001; Noda et al., 2007; Mecozzi et al., 2016; Jung et al., 2018
9	Polycarbonate (PC)	2966 1768 1503 1409 1364 1186 1158 1013 828	CH stretching C = O stretching Aromatic ring stretching Aromatic ring stretching CH3 bending C–O stretching C–O stretching Aromatic CH in plane bending Aromatic CH out-of- plane bending	Verleye et al., 2001, Noda et al., 2007, Asensio et al., 2009, Jung et al., 2018
10	Cellulose acetate (CA)	1743 1368 904 600	C = O stretching CH3 bending Aromatic ring stretching or CH bending O–H bending	Ilharco and Brito de Barros, 2000; Verleye et al., 2001; Noda et al., 2007; Jung et al., 2018
11	Acrylonitrile butadiene styrene (ABS)	2922 1602 1494 1452 966 759 698	C–H stretching Aromatic ring stretching Aromatic ring stretching CH2 bending =C–H bending Aromatic CH out-of-plane bending, =CH bending Aromatic CH out-of-plane bending	Verleye et al., 2001; Jung et al., 2018
12	Polytetrafluorethylene (PTFE)	1201 1147 638 554 509	CF2 stretching CF2 stretching C–C–F bending CF2 bending CF2 bending	Coates, 2000; Verleye et al., 2001; Jung et al., 2018
13	Poly(methyl methacrylate) (PMMA or acrylic)	2992 2949 1721 1433 1386 1238 1189 1141 985 964 750	C–H stretching C–H stretching C = O stretching CH2 bending CH3 bending C–O stretching CH3 rocking C–O stretching CH3 rocking C–H bending CH2 rocking, C = O bending	Verleye et al., 2001; Jung et al., 2018
14	Ethylene vinyl acetate (EVA)	2917 2848 1740 1469 1241 1020 720	C–H stretching C–H stretching C = O stretching CH2 bending, CH3 bending C(=O)O stretching C–O stretching CH2 rocking	Verleye et al., 2001; Asensio et al., 2009; Jung et al., 2018
15	Nitrile	2917 2849 2237 1605 1440 1360 1197 967	=C–H stretching =C–H stretching CN stretching C = C stretching CH2 bending CH2 bending CH2 bending =C–H bending	Coates, 2000; Verleye et al., 2001; Jung et al., 2018
16	Latex	2960 2920 2855 1167 1447 1376	C–H stretching C–H stretching C–H stretching C = C stretching CH2 bending CH3 bending	Guidelli et al., 2011; Jung et al., 2018

Table 4. ATR-FTIR spectroscopic characterization and identification of MPs extracted from different environmental matrices in select locations.

Sampling location	Sample type	Shape	Method of extraction	MP particles	Abundant polymer types	Objectives and governing process	Reference
Northern coast of Taiwan	Sediment	Fragments, foams, pellets and fibers	Sieving and density separation	1097	PE and PP	Distribution and characterization of MPs	Kunz et al., 2016
Chennai coast, India	Sediment	Resin pellets	Sieving	304 (before flood), 896 (after flood)	PE and PP	Distribution and characterization of MPs	Veerasingam, Mugilarasan, et al., 2016
Goa coast, India	Sediment	Resin pellets	Sieving	1655 (SW monsoon), 1345 (NE monsoon)	PE and PP	Distribution and characterization of MPs	Veerasingam, Saha, et al., 2016
Lakshadweep Islands, India	Sediment	Resin pellets	Sieving	603	PE and PP	Distribution and characterization of MPs	Mugilarasan et al., 2017
South east coast of India	Sediment	Fragments, fibers and foams	Sieving and density separation	178 ± 261 mg/m^2 (Low tide); 1323 ± 1228 mg/m^2 (High tide)	PE, PP and PS	Distribution and characterization of MPs	Karthik et al., 2018
Coast of Guangdong, South China	Sediment	Foams, fragments, and pellets	Sieving and density separation	6675 ± 7021 particles/m^2	PS, PP, and PE	Distribution and characterization of MPs	Fok et al., 2017
Terra Nova Bay (Ross sea, Antarctica)	Sediment	Fibers, films and fragments	Sieving	0.67 to 168.36 particles/m^2	PE, PP, PVC, PS	Distribution and characterization of MPs	Munari et al., 2017
Mumbai and Tamil Nadu coasts, India	Sediment	Fibers, granules and films	Sieving and density separation	45 ± 12 to 220 ± 50 particles/kg	PE, PET, PS, PP, PVC	Distribution and characterization of MPs	Tiwari et al., 2019
Tamil Nadu coast, India	Sediment	Fragments, fibers and foams	Hydrogen peroxide treatment and density separation	33 ± 30 to 179 ± 68 particles/kg (Low tide); 119 ± 72 to 439 ± 172 particles/kg (High tide);	PE, PP, nylon, polystyrene and polyester	Adsorption behavior of chemical elements; Distribution and characterization of MPs	Sathish et al., 2019
Japan, Thailand, Malaysia and South Africa	Sediment	Fragments, fibers, films and beads	Hydrogen peroxide treatment and density separation	100 particles/kg (Thailand); 1900 particles/kg (Tokyo)	PE, PP, PS, PET, PVC, acrylics and polyamides	Distribution and characterization of MPs	Matsuguma et al., 2017
Bohai Sea, China	Sediment	Fibers	Density separation	102.9 ± 39.9 to 163.3 ± 37.7 particles/kg	PEVA, LDPE, PE and PS	Distribution and characterization of MPs	Yu et al., 2016
South Carolina estuaries, USA	Water	Fragments, fibers, foams and spheres	Sieving, H2O2 treatment	6.6 ± 1.3 to 30.8 ± 12.1 particles/L	Polystyrene, nylon, polyester, PE and PP	Distribution and characterization of MPs	Gray et al., 2018
Geoje Island, South Korea	Water	Fragments, fibers, sheets and EPS	Filtered through GF/F 0.75 µm	206 ± 117 particles/L (fragment); 0.4 ± 1.8 particles/L (EPS); 4.5 ± 4.1 (fiber)	PE and PP	Distribution and characterization of MPs	Song, Hong, Jang, Han, and Shim, 2015
Australia	Water	Fibers, particles	Filtered through GF/F 37 µm	547 items	PES, PET	Fate and transport of MPs	Jensen et al., 2019
Qatar	Water	Fragments and fibers	Centrifugation, chemical digestion followed by filtration	0 to 3 particles/m^3	PP	Distribution and characterization of MPs	Castillo et al., 2016
North Atlantic Ocean, Scotland	Water	Microfibers	Filtered through 80 µm filter paper	70.8 particles/m^3	Polyester, PET and acrylic fiber	Distribution and characterization of MPs	Courtene-Jones, Quinn, Gary, et al., 2017
Tamil Nadu coast, India	Biota (Rastrelliger kanagurta, Siganus javus, Arius arius, Leiognathus equulus, and Mugil cephalus)	Fibers and fragments	Digestion of intestines (10% KOH)	MPs found in 10.1% of samples	PE and PP	Ingestion and effects in organisms	Karthik et al., 2018
Baltic Sea and North Sea	Biota (Clupea harengus and Scomber scombrus)	Fibers, fragments, films and spherules	Visual observation of the gastrointestinal tract under the stereomicroscope	MPs found in 5.5% of samples	PE, PA, PP, PS	Ingestion and effects in organisms	Rummel et al., 2016
North Sea	Biota (Clupea harengus, Sprattus  species,Limanda limanda, and Merlangius merlangus)	Spherules	Digestion of intestines (10% KOH)	MPs found in 0.25% of samples	Polymethylmethacrylate (PMMA)	Ingestion and effects in organisms	Hermsen et al., 2017
Coast of Easter Island	Biota (Decapterus muroadsi)	Fragments	Visual inspection	MPs found in 80% of samples	PE, PP	Ingestion and effects in organisms	Ory et al., 2017
Adriatic Sea, Italy	Biota (Mullus barbatus)	Fragments, films, pellets and lines	Digestion (30% H2O2) of whole organism	MPs found in 64% of samples	PE	Ingestion and effects in organisms	Avio et al., 2015
England	Biota (Platichthys flesus, Osmerus eperlanus)	Fibers and fragments	Visual inspection	MPs found in 75% of samples	PET, PE, Polyester	Ingestion and effects in organisms	McGoran et al., 2017
UK	Biota (Mytilus edulis)	Fibers and fragments	Treated with Trypsin solution and filtration	0.678 ± 0.044 to1.582 ± 0.448 particles/g	Acrylic	Ingestion and effects in organisms	Courtene-Jones, Quinn, Gary, et al., 2017
China	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	120–718	Nylon, EVA, PE, PET, PP, PU, PVC	Quantification of MPs in salt	Kim et al., 2018
China	Lake salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	28	PE, PET, PP, PS, Teflon	Quantification of MPs in salt	Kim et al., 2018
China	Rock salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	0–14	PET, PP, Teflon	Quantification of MPs in salt	Kim et al., 2018
USA	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	300	PE	Quantification of MPs in salt	Kim et al., 2018
India	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	50–600	Nylon, PE, PET, PP, PVC	Quantification of MPs in salt	Kim et al., 2018
Italy	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	5–50	Nylon, PE, PET, PP	Quantification of MPs in salt	Kim et al., 2018
UK	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	120	PP, PE, PVC	Quantification of MPs in salt	Kim et al., 2018
Germany	Rock salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	2	PET	Quantification of MPs in salt	Kim et al., 2018
Korea	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	100–300	Arcylic, Neylon, PE, PET, PP	Quantification of MPs in salt	Kim et al., 2018
Thailand	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	80–600	PE, PET, PP, PVC	Quantification of MPs in salt	Kim et al., 2018
Vietnam	Sea salt	Fragments, fibers, sheets	Digestion with 17.25% H2O2 and filtration	100–200	Arcylic, PE, PP, PW	Quantification of MPs in salt	Kim et al., 2018

The µ-FTIR method allows chemical imaging of larger areas of membrane filters with high resolution using focal plane array (FPA) detectors (Bergmann et al., 2017, 2019; Cincinelli et al., 2017; Liu, Olesen, et al., 2019; Mintenig et al., 2017; Olesen et al., 2019; Primpke et al., 2017, 2018, 2019; Primpke, Cross, et al., 2020; Simon et al., 2018; Vianello et al., 2019; Zarfl, 2019). µ-FTIR technique is extensively used for chemical imaging (Table 5) of different modes—coupled with ATR (Muhlschlegel et al., 2017; Pegado et al., 2018; Song, Hong, Jang, Han, & Shim, 2015; Wagner et al., 2017), transmission (Compa et al., 2018; Ding et al., 2018; Frias et al., 2010; Kuhn, Schaafsma, et al., 2018; Liu, Lu, et al., 2018; Loder et al., 2015; Lourenco et al., 2017) and reflection (Kanhai et al., 2018; Lares et al., 2018; Phuong, Poirier, et al., 2018; Tagg et al., 2015; Vianello et al., 2013; Wang, Peng, et al., 2017). Though µFTIR imaging is an effective method for MP characterization, it needs long measurement time (often greater than 20 h) of samples of MP particles on a filter (Kappler et al., 2016; Loder et al., 2015; Primpke et al., 2017). Moreover, it also poses the risk of sample contamination or loss. µFTIR in reflection mode is required a good reflection property and it is less suitable for colored or small MP particles, whereas transmission mode may show total absorption for large or thick particles (Primpke, Christiansen, et al., 2020).

Table 5. µ-FTIR imaging characterization and identification of MPs extracted from different environmental matrices in select locations.

Sampling location	Sample type	Shape	Method of extraction	MP particles	Abundant polymer types	Objectives and governing process	Reference
Beijiang River, China	Sediment	Fragments and films	Density separation and floatation	178 ± 69 to 544 ± 107 particles/kg	PP and PE	Adsorption behavior of metals; distribution and characterization of MPs	Wang, Peng, et al., 2017
Denmark	Water	—	Filtered through filter paper 10 µm, H2O2 treatment	490–22,894 items/m^3	PP, PVC, PES, PE and PS	Distribution and characterization of MPs	Liu, Olesen, et al., 2019
Northern Gulf of Mexico	Water	Fragments, fibers and beads	Filtered through Whatman glass microfiber filter 0.70 µm	4.8 to 8.2 particles/m^3 (Bongo net); 5.0 to 18.4 particles/m^3 (Bongo net)	Cellophane, Alkyd resin and PE	Distribution and characterization of MPs	Mauro et al., 2017
Taihu lake, China	Water	Fragments, fibers, films and pellets	Filtered through filter paper 5 µm	0.01 × 10^6 to 6.8 × 10^6 items/km^2(plankton net); 3.4 to 25.8 particles/L	Cellophane, PE, PS and PP	Distribution and characterization of MPs	Su et al., 2016
Tamar estuary, England	Water	Fragments, sheets, fibers and pellets	Sieving with various mesh sizes (3 mm, 1 mm and 270 µm)	204	PE, PS and PP	Distribution and characterization of MPs	Sadri and Thompson, 2014
Wind farm, Yellow Sea, China	Water	Fibers, granules and films	Sieving, H2O2 treatment	0.330 ± 0.278 particles/m^3	PE, PET and Cellophane	Distribution and characterization of MPs	Wang, Zhou, et al., 2018
Ross Sea, Antarctica	Water	Fibers and fragments	Filtered through glass fiber filter 1µm	0.0032 to 1.18 particles/m^3	PE and PP	Distribution and characterization of MPs	Cincinelli et al., 2017
France	Air	Fibers	Filtered on quartz fiber GF/A Whatman filters (1.6 μm)	2 to 355 particles/m^2/day	PET, Polyamide	Distribution and characterization of MPs in atmospheric fallout	Dris et al., 2016
France	Air	Fibers	Density separation and filtration through GF/A Whatman filters (1.6 μm)	1586 to 11,130 particles/m^2/day	PP, Polyamide (nylon), PE	Distribution and characterization of MPs in indoor and outdoor environments	Dris et al., 2017
China	Air	Fibers	Filtered on glass fiber GF/A Whatman filters (1.6 μm)	0 to 4 particles/m^3	PET, PE, PES, RY	Distribution, characterization and risk assessment of suspended atmospheric MPs	Liu, Wang, Fang, et al., 2019
Finland	Waste water treatment plant	Fibers	H2O2 treatment and filtration	0.1 to 124.7 particles/L	PS, PE, PA, PP	Quantification and characterization of MPs in wastewater effluent	Lares et al.,2018
Northern Ireland	Biota (Mesoplodon mirus)	Fibers, fragments, and films	Digestion of intestines (10% KOH)	MPs found in 85% of samples	PP, acrylic, PE, polyester	Ingestion and effects in organisms	Lusher et al., 2015
Portugal	Biota (Boops boops, Trachurus picturatus Scomber japonicas, and Trachurus trachurus)	Fibers and fragments	Visual inspection	MPs found in 19.8% of samples	PP, PE, polyester, Rayan	Ingestion and effects in organisms	Neves et al., 2015
Northeast Atlantic around Scotland	Biota (Pleuronectes platessa, and Micromesistius poutassou)	Fibers	Washed with doubly distilled H2O	MPs found in 47.7% of samples	PA, PET, Arcylic, PP	Ingestion and effects in organisms	Murphy et al., 2017
China	Biota (Hyporhamphus intermedius, Larimichthys crocea, Mugil cephalus, Callionymus planus,and Cyprinus carpio)	Fibers, framents, pellets, sheets and films	Digestion (30% H2O2) of whole organism	MPs found in 36.8%–92.3% of samples	Cellophane, PET, PS	Ingestion and effects in organisms	Jabeen et al., 2017
Mondego estuary, Portugal	Biota (Dicentrarchus labrax, Diplodus vulgaris, and Platichthys flesus)	Fibers	Digestion of intestines (10% KOH)	MPs found in 38% of samples	Polyester, PP, PE, Polyacrylonitrile	Ingestion and effects in organisms	Bessa et al., 2018
Shanghai, China	Biota (Alectryonella plicatula, Cyclina sinensis, Meretrix lusoria, Mytilus galloprovincialis, Patinopecten yessoensis, Ruditapes philippinarum, Scapharca subcrenata, Sinonovacula constricta, and Tegillarca granosa)	Fibers, fragments and pellets	Digestion (30% H2O2) of whole organism	4.3 to 57.2 particles/individual	PE, PET, PA	Ingestion and effects in organisms	Li et al., 2015
French Atlantic coast	Biota (Mytilus edulis and Crassostrea gigas)	Fragments	Digestion (10% KOH and 50% KI) of whole organism	0.61 ± 0.56 to 2.1 ± 1.7 particles/individual	PE, PP	Ingestion and effects in organisms	Phuong, Poirier, et al., 2018
Qingdao, China	Biota (Chlamys farreri and Mytilus galloprovincialis)	Fibers, fragments and granules	Digestion (10% KOH and 30% H2O2) of whole organism	5.2 to 19.4 particles/individual; 1.9 to 9.6 particles/individual	Cellophane, PP, PTFE	Ingestion and effects in organisms	Ding et al., 2018
UK	Biota (Mytilus edulis)	Fibers, fragments	Digestion (30% H2O2) of whole organism	1.1 to 6.4 particles/individual	PS, PP, PE	Ingestion and effects in organisms	Li, Green, et al., 2018
English channel (UK)	Biota (Merlangius merlangus, Micromesistius poutassou, Trachurus trachurus, Trisopterus minutus, and Zeus faber)	Fibers, fragments, films and beads	Inspection of the digestive tract with a dissecting microscope	MPs found in 36.5% of samples	PA, polyester, PS	Ingestion and effects in organisms	Lusher et al., 2013
Norway	Biota (Mytilus spp.)	Fibers, fragments	Digestion (10% KOH)	1.5 ± 2.3 particles/individual	Cellophane, EVA, PET, PP, PE	Ingestion and effects in organisms	Brate, Hurley, et al., 2018
China	Biota (Mytilus edulis, and Perna viridis)	Fibers, fragments, and pellets	Digestion (30% H2O2) of whole organism	0.77 to 8.22 particles/individual	Rayon, PET, PE, PVC, PP	Ingestion and effects in organisms	Qu et al., 2018
Mediterranean coast of Turkey	Biota (Argyrosomus regius, Diplodus annularis, Lithognathus mormyrus, Mullus barbatus, Mullus surmuletus, Nemipterus randalli, Pelates quadrilineatus, Saurida undosquamis, Sparus aurata, Trachurus mediterraneus, and Upeneus pori)	Fibers	Digestion (35% H2O2) of stomach and intestines	MPs found in 34% of samples	PE, PP	Ingestion and effects in organisms	Guven et al., 2017
China	Sea salt	Fragments and fibers	Digestion with 30% H2O2 and filtration	550–681	CP, CL, PAN, PB, PE, PES, PET, PP, POM, PMA, poly(vinyl acetate:ethylene)	Quantification of MPs in salt	Yang et al., 2015
India	Sea salt	Fibers, fragments	Digestion with 30% H2O2 and filtration	56–103	PA, PE, PES, PET, PS	Quantification of MPs in salt	Seth and Shriwastav, 2018
Italy	Sea salt	Fibers, fragments, fi	Filtration through filter (0.45 µm) by vacuum filtration	22–594	PE, PP	Quantification of MPs in salt	Renzi and Blaskovic, 2018
Croatia	Sea salt	Fibers, films, fragments, granules, pellets, and foams	Filtration through filter (0.45 µm) by vacuum filtration	13,500–19,800	PE, PP	Quantification of MPs in salt	Renzi and Blaskovic, 2018
Spain	Sea salt	Fibers	Centrifugation followed by filtration	50–280	PE, PET, PP	Quantification of MPs in salt	Iniguez et al., 2017
Spain	Well salt	Fibers	Centrifugation followed by filtration	115–185	PE, PET, PP	Quantification of MPs in salt	Iniguez et al., 2017

Automatic FTIR imaging such as FPA (focal plane array) is used for fast acquisition of several spectra within an area through one single measurement. Liu, Olesen, et al. (2019) used μFTIR imaging with a 125 × 125 Mercury-Cadmium-Telluride (MCT) FPA detector (spectral resolution of 5.5 μm) to identify the polymer types of MPs. Bergmann et al. (2019) applied µFTIR equipped with a FPA (64 × 64 detector elements) for scanning an area of 20 × 20 FPA (14.1 by 14.1 mm), which allowed the detection of MP particles down to 11 μm in 4.5 h. Various sophisticated softwares are used for spectral correlation analysis in MP research. Approximately, 1.8 million spectra, which take 48 hrs data analysis time using the Bruker OPUS software, and take only 4 hrs time using the siMPle software to complete the spectral correlation (Peeken et al., 2018; Primpke, Cross, et al., 2020). Though mercury cadmium telluride (MCT) detector is used in µFTIR chemical imaging of MPs, it is a time-consuming method for larger filter areas (Harrison et al., 2012; Vianello et al., 2013; Yu et al., 2016). Loder et al. (2015) used ten filters with different materials, pore size and thickness to test their applicability for FPA based MP characterization experiment using µFTIR (both reflectance and transmittance modes). Out of eight tested filters, only two filters, viz., polycarbonate and the aluminum oxide, were suitable for FPA based µFTIR measurements of MPs. The problems of remaining six filters were their IR window range, i.e., IR transparency was either too narrow, IR characteristics had high diffractive error (reflectance mode) or absorbance values much higher than 0.5, resulting in unclear IR spectra. Small MP particles were often concentrated on various filters (e.g., aluminum oxide filters, and metal covered PC filters) (Bergmann et al., 2017; Loder et al., 2015; Mani, Primpke, et al., 2019; Mintenig et al., 2017, 2019; Peeken et al., 2018; Primpke et al., 2017, 2018, 2019; Primpke, Christiansen, et al., 2020), silicon membranes (Kappler et al., 2015, 2016), slides (Kunz et al., 2016; Wagner et al., 2017), compression cells (Cai et al., 2019; Frias et al., 2014) and windows made of infrared transparent or reflective materials (Loder et al., 2015; Pham et al., 2017; Primpke, Christiansen, et al., 2020). The cost of aluminum oxide filter is cheaper than other filters (Bergmann et al., 2017; Mani, Primpke, et al., 2019; Mintenig et al., 2017, 2019; Primpke et al., 2017, 2018, 2019). However, the aluminum oxide filters have wavenumber limitation toward 3600–1250 cm⁻¹ compared to other filters (FPA: 3600–900 cm⁻¹ and even lowest for MCT (Primpke, Christiansen, et al., 2017, 2020). However, characterization of MPs using FTIR measurements cannot be performed in the presence of water as its spectrum will overlay the target spectra (Primpke, Christiansen, et al., 2020; Primpke, Cross, et al., 2020).

Within a FTIR spectrum of MP, not every spectral region contains relevant information (Renner, Nellessen, et al., 2019). Therefore, the spectrum can be reduced to a more compact and highly characteristic sub-spectrum, which can increase the selectivity of library searching (Hendrickson et al., 2018; Kappler et al., 2015; Kroon, Motti, Talbot, et al., 2018; Renner et al., 2017). Usually, FTIR spectra of MPs can be divided into three regions: 4000–2750, 2750–1850, and 1850–700 cm⁻¹. Renner, Nellessen, et al. (2019) found that the variance (intensity) and specificity (number of individual signals) are very high in the fingerprint region (1850–700 cm⁻¹) (Figure S1). Therefore, it is suggested that the fingerprint region is suitable for the identification of MPs, whereas the middle region (2750–1850 cm⁻¹) is not appropriate for this purpose due to its low variance and specificity (Cabernard et al., 2018; Kappler et al., 2016). FTIR spectra for various types of polymers are shown in Figure S1. These spectra have been evaluated using five different spectral ranges, such as stretching vibration of CH/CH₂/CH₃ groups (2980–2780 cm⁻¹), CH₂ bending vibration (1480–1400 cm⁻¹), C = O stretching vibration (1760–1670 cm⁻¹), C = O stretching vibration (1800–1740 cm⁻¹) and CF₂ stretching vibration (1174–1087 cm⁻¹). FTIR spectra of unknown MPs can be effectively identified through comparing to a commercially available polymer library of known recorded spectra of polymers (Renner, Nellessen, et al., 2019).

4. Weathering and aging

Understanding the surface degradation of MPs in the environment increases our knowledge of the pollutant-plastic interaction. In addition, knowing the weathering of MPs is important for understanding the ecological impacts of the most common type of plastic debris. The environmental degradation of MPs via photo-, bio-, oxidative-, hydrolytic and altogether were studied using FTIR techniques by many researchers (Andrade, Fernandez-Gonzalez, et al., 2019; Auta et al., 2018; Bond et al., 2018; Brandon et al., 2016; Costa et al., 2018; Nazareth et al., 2019; Rajakumar et al., 2009; Ren et al., 2019; Tang et al., 2019; Tofa et al., 2019; von Friesen et al., 2019). FTIR results highlight three likely areas of weathering related changes in MPs: hydroxyl groups (broad peaks from 3100 to 3700 cm⁻¹, centered at 3300–3400 cm⁻¹), alkenes or carbon double bonds (1600–1680 cm⁻¹) and carbonyls (1690–1810 cm⁻¹). FTIR spectroscopy is used to measure the changes in chemical bond structures (carbonyl groups, hydroxyl, and carbon-oxygen) with weathering (Brandon et al., 2016). Carbonyl index (CI) is commonly used to measure the light induced photo-oxidation of polyethylene MPs, since it normally increases with increasing exposure time, i.e., increasing degree of photo-degradation (Endo et al., 2005; Veerasingam, Saha, et al., 2016). Carbonyl index is defined as the absorbance of carbonyl bond peaks relative to the absorbance of reference peaks (methylene bond). The carbonyl absorption bands are considered in the region 1760–1690 cm⁻¹ and the methylene absorption bands are taken in the region 1490–1420 cm⁻¹ (methylene scissoring peak) (Halle et al., 2016). Hydroxyl index (HI) is calculated as the ratio of the absorbance peak at 3340 cm⁻¹ (3300–3400 cm⁻¹) to the absorbance at reference peak. It is found that several reference peaks are used, including 974 and 2720 cm⁻¹ for PP and 1465 and 2020 cm⁻¹ for PE (Brandon et al., 2016). Rajakumar et al. (2009) noted significant changes of HI only at the time of rapid growth of CI value. Carbon-oxygen index (COI) is measured based on the ratio between absorbance peak in the region 1000 to 1200 cm⁻¹ and reference peak. Reference peak regions for PE and PP are 2908–2920 cm⁻¹ and 2885–2940 cm⁻¹, respectively. Changes in HI and COI values could be readily diagnosed by FTIR, followed by CI value (Brandon et al., 2016).

Recently, Andrade, Winemiller, et al. (2019) found that FTIR could be used as a low-cost technique to monitor changes in different polymer types (polyamide, polypropylene and polystyrene) of MPs during oceanic aging. In polyamide MPs, the photo-oxidation and/or thermo-oxidation processes increase the intensity of 1150 cm⁻¹ band in FTIR spectra. The IR spectrum of polypropylene MP changes substantially with weathering due to the appearance of following functional groups: 3370 and 3240 cm⁻¹ (hydroxyl groups), 1640 cm⁻¹ (C = O and double bonds), 1530 cm⁻¹ (C = O ketones), 1440 cm⁻¹ (carboxylic acids), 1140 cm⁻¹ (alkanes), 1100 cm⁻¹ (C–O bond), and 720 cm⁻¹ (CH₂) (Andrade, Fernandez-Gonzalez, et al., 2019; Tang et al., 2019). The FTIR spectra of weathered polystyrene MP showed spectral changes correspond to the formation of new bands at 3360–3240 cm⁻¹ (hydroxyl group), 1640 cm⁻¹ (double bond or C = O groups), and 1100 cm⁻¹ (C-O bonds) (Andrade, Fernandez-Gonzalez, et al., 2019). The degree of crystallinity of polymer is a key variable in defining the morphological, physical and mechanical properties of the MP. Crystallinity is calculated by the method used in Stark and Matuana (2004). The doublet peaks observed at 1474–1464 cm⁻¹ and 730–720cm⁻¹ correspond to PE crystalline content (1474 and 730 cm⁻¹) and amorphous content (1464 and 720 cm⁻¹), respectively. The percentage of the crystalline content, X, can be calculated using the following equation: $$X=100-\frac{(1-I_a/I_b)/1.233}{1+I_a/I_b}\mathrm{ }\left(100\right)$$ where I_a and I_b can be determined from the bands at either 1474 and 1464 cm⁻¹ or 730 and 720 cm⁻¹, respectively. Crystallinity increases when chain scission occurs in MPs. The shorter chains produced by chain scission events can crystallize more readily than the crosslinks. The chain scission affects the tie molecules, and thereby the crystallinity is decreased. Holmes et al. (2014) found that the weathering process is more influenced in the absorption of trace metals by MPs than its polymer types. It is found that during the aging process, the absorbance ratio between contaminant vibrational band and polymer vibrational band is increased. However, aging of polymers is a very complex process and it is nearly impossible to estimate average duration of existence of MPs in an environmental compartment. Instead, aging extent of several MPs can be compared with one another (Renner et al., 2017). Wang, Zhou, et al. (2019) reported that the probable sources of major MPs can be identified based on polymer types, shape and topographic features.

5. Ecotoxicology

Globally, environmental-vector of MPs have been studied in various marine and freshwater biota, especially in fish, bivalves and birds (Anbumani & Kakkar, 2018; Chang et al., 2020; Du et al., 2020). Several toxicology studies (organismal-vector) including uptake of MPs and associated pollutants have been conducted in laboratories (Browne et al., 2013; Cole et al., 2013; Koelmans et al., 2016). Fishes and bivalves, which are used as bio-indicators to assess the health of aquatic ecosystems, are also used as the most common organisms to study ecotoxicology of MPs. The major toxic effects of MPs on biota can be categorized as follows: (1) physical toxicity, (2) chemical toxicity, (3) cellular toxicity, and (4) Toxicity due to pathogenic microbes (Bhattacharya & Khare, 2020). Following are the adverse effects of ingested MPs in aquatic and terrestrial organisms reported in the publications: (1) feeding activity and behavioral changes (Bour et al., 2018; Scherer et al., 2017; Wang, Mao, et al., 2019; Wu, Tao, et al., 2019), (2) inhibition of growth, reproduction and fecundity (Garrido et al., 2019), (3) histopathological damage and/or death (Lei et al., 2018), (4) Oxidative stress (Yu et al., 2018), (5) neurotoxicity (Barboza et al., 2018), (6) immune system responses (Avio et al., 2015), (7) changes in gene expression (Rochman et al., 2014, 2015), and (8) energy deficiency (Rodriguez-Seijo, 2018). Prokic et al. (2019) found that in biota, MPs can have the following effects: (1) induce oxidative damage (increased lipid peroxidation and DNA strand breaks), (2) alter anti-oxidative system (superoxide dismutase, catalase and glutathione peroxidase were parameters with the highest and significant changes in activities) and metabolism (isocitrate dehydrogenase and lactate dehydrogenase activity), and (3) have neurotoxic effects (inhibition of acetylcholinesterase activity). These effects depend on size and dose of used MPs, and/or their interaction with other xenobiotics.

FTIR spectroscopic imaging (µFTIR) is found to be label-free, nondestructive chemical analysis and powerful tool for studying live biological cells (Chan & Kazarian, 2013; Baker et al., 2014). Recently, µ-FTIR spectroscopy was used to describe the distribution of MPs in different fish organs and the possibility of bio-accumulation in fish tissues (Su, Deng, et al., 2019). Sun, Chen, et al. (2018) used FTIR technique to investigate the toxic effects of polystyrene nano- and micro-particles on the marine bacterium Halomonas alkaliphila. FTIR absorption bands for carbohydrate, polysaccharides, and amide were shifted to a higher wavenumber when Halomonas alkaliphila was exposed to nanoplastics, but not in MPs. Thus, the size of plastics plays an important role in the alteration of the bacterial chemical composition (Sun, Chen, et al., 2018). Therefore, FTIR technique is not only used to identify the polymer types of MPs ingested by biota, but also used to investigate the toxicity effects in the biota.

6. Analytical methods and future challenges

The quality, spectral contrast and quantitative precision of FTIR spectrum recorded from a MP sample depends on the choice of sample preparation method and/or sampling technique. Transmission, attenuated total reflectance, diffuse reflectance, reflectance (mapping) and reflectance (focal plane array mapping) techniques are used extensively in MP research, and every technique has its own advantages and disadvantages (Table 6) depending on analytical issue and sample properties (Renner et al., 2016). Three steps are followed to characterize the MPs using FTIR spectra: (1) inspection of data quality, (2) evaluation of FTIR spectra to identify the corresponding MPs, and (3) identification using spectral library search—calculation of a similarity value called “Hit Quality Index” (HQI).

Table 6. Techniques and accessories used in FTIR for analysis of MPs based on their merits and demerits (modified from Renner et al., 2016).

Sampling technique	Handling	Penetration depth (µm)	Size of MPs (µm)	Unsuitable
(1) Transmission
Transmittance of thin polymer films	Very simple	<300	>13,000	MPs, polyamides
Transmittance of MPs within KBr pellets	Complex	<300	<1000	Large MPs, individual MPs
µ-Transmittance	Very complex	<300	>10	Polyamides
(2) Reflectance
Attenuated total reflectance (ATR)	Very simple	∼2	>1000	Small MPs, convex particles, severely aged or contaminated samples
µ-ATR	Complex	∼2	>100	Convex particles, severely aged or contaminated samples
(3) Diffuse reflectance
DRIFTS	Complex	<50	<50	Large MPs, individual MPs

Inspection of data quality has two steps—looking at the characteristic shape of the whole spectrum and characterizing individual vibrational bands. However, this inspection method requires much experience and it is very time consuming. Smoothing and baseline correction are important to improve the quality of FTIR spectra of MPs (smoothing to increase signal-to-noise ratio). The moving average and Savitzky-Golay smoothing are well established and both are useful techniques to enhance the certainty of results. The common baseline correction methods used in FTIR spectroscopy are Savitzky-Golay differentiation (based on Savitzky-Golay smoothing), rolling-circle filter, asymmetric least squares, and adaptive iteratively reweighted penalized least squares (Renner, Nellessen, et al., 2019).

Evaluation of FTIR spectra is the most critical step in MP analysis. Manual interpretation of relevant absorption peaks based on reference tables and comparison of complete spectra with a reference spectral library are the two common approaches in the evaluation. Manual interpretation is time consuming and requires expert knowledge, and thus, it suits only for a low throughput quantity. Spectra obtained from analysis of MPs are typically compared and matched with those of model samples from library databases. For example, in one study, the spectra matched with HQI ≥0.7 were accepted, those with HQI <0.6 were rejected and spectra with HQI = 0.6 were individually interpreted (Woodall et al., 2014).

FTIR spectra acquired from different modes (transmission, reflection and diffuse reflectance) are not the same (Picollo et al., 2014). Only a few researchers are working on this topic (Xu & Gowen, 2019; Xu, Thomas, et al., 2019), and adequate attention is not given for comparing/matching the unknown spectra with literature/spectral library. Several degradation processes in MPs cause spectral alterations relative to pristine reference library spectra. Thus, library searching could be vulnerable to misidentification of MP samples. However, most studies ignored spectral changes caused by MP degradation when comparing or matching with the reference spectral library. Therefore, it is important to improve new library searching procedures, which are more robust and can handle the problem of comparing the weathered MPs with virgin polymer references. Substrate (holding MPs) has a minimum spectral interference and can immobile the MPs especially for µ-FTIR imaging, where the stage moves during scanning (Corami et al., 2020; Vianello et al., 2013; Xu, Thomas, et al., 2019). Therefore, it is necessary to address the importance of appropriate substrate material for obtaining high quality results with minimum spectroscopic interference in the FTIR spectra.

In general, two approaches are followed to analyze MPs on filter materials: (1) measurement of complete filter area spot by spot (Kappler et al., 2016; Loder et al., 2015; Primpke et al., 2017) and (2) measurement of selected points of interest (Maes et al., 2017). In complete filter measurement method, the maximum number of spots are analyzed (also high analysis time) and big data sets are needed to be evaluate. However, without either FPA detector or charge-coupled device, this approach is not recommended. In the select points of interest measurement method, the number of spots (also less analysis time) is reduced, and FPA detector is not needed. But there is a risk of overlooking MP particles and also an additional step is required to define points of interest (Renner et al., 2020). Though the automated evaluation method demonstrated that the analysis time has decreased from several days to 4–9 h for a scanning, only recently this technique is introduced in the MP research (Chen et al., 2020; Primpke, Christiansen, et al., 2020). Therefore, further improvements are needed in constructing a reference spectrum for weathered MPs, and also reducing the analytical time. Moreover, training the classifier can increase the analysis speed substantially when dealing with large datasets of FTIR spectra. For example, automated identification methods were tested based on hierarchical cluster analysis (Primpke et al., 2018), shortwave infrared imaging (Schmidt et al., 2018), identification of the most relevant bands (Renner et al., 2017; Renner, Nellessen, et al., 2019), random decision forest method (Hufnagl et al., 2019), modified chemometric identification concept (Renner, Sauerbier, et al., 2019), machine learning method (Kedzierski et al., 2019), Python based μFTIR mapping (Renner et al., 2020) and Hybrid fusion method (Chabuka & Kalivas, 2020). The analysis of FTIR spectra is time-consuming as often it is needed to compare the spectra one by one with the reference spectra. Moreover, appearance of additional bands in the spectra due to aging of MPs and/or fouling present on particles is a significant part of misinterpretation. More often, the reference spectra in the spectral library are made of new and clean plastics, and there could be a possibility that MPs may not be identified successfully with the matching algorithm and spectral library of FTIR instrument. Therefore, it is mandatory to develop additional accessories to combine with FTIR techniques for the characterization of MP particles less than 10 µm size.

7. Conclusions

MPs have been identified and reported in environmental matrices from the poles to equator and from atmosphere to deep ocean. Though the usage of FTIR techniques in MP research has increased tremendously since 2004, there are still some challenges to be overcome in the area of standardizing the operational protocols for identification and quantification. Among the reviewed research articles, ATR-FTIR has been used in 60% of the studies to characterize different polymer types at various environmental matrices. Attenuated total reflectance (ATR) technique coupled with FTIR spectroscopy is widely used to characterize the large size MPs, whereas smaller MPs require the use of µFTIR coupled with detector, especially, µFTIR coupled with focal plane array detector facilitates a much faster generation of chemical imaging of MPs by simultaneously scanning several thousand spectra within a single measurement. FTIR technique is also used to study the changes in chemical bond structures (hydroxyl, carbonyl groups and carbon-oxygen) of MPs during various weathering process. Moreover, FTIR method is used to understand the ecological effects of ingested MPs and its associated pollutants and biochemical variations at the cellular level. Following criteria are needed to be considered during data processing, evaluation and identification MPs using FTIR spectroscopy: (1) when we compare/match an unknown spectrum with literature or a commercial spectral library, it is necessary to check the mode of acquisition of FTIR spectra (acquired the same mode or different modes), (2) the spectral change caused by weathering and aging, when comparing or matching with the spectral library, and (3) desirable substrate to be used to reduce the spectral interference. Standardization of chemometric techniques, decrease in data processing time, better file handling capabilities of systems are expected to improve FTIR analysis in MP pollution research. Since identification of MP particles <10 µm using FTIR technique is a challenging task, development of novel additional accessories or proper combination use of existing techniques is needed for future MP research.

Acknowledgements

This work was carried out under the following projects: (1) QUEX-ESC-QP-TM-18/19, funded by Qatar Petroleum through ESC, Qatar University, Qatar, (2) INT/RUS/RFBR/P-339, funded by Department of Science & Technology, India, and (3) RFBR No. 18-55-45024, funded by RFBR, Russia. Open Access funding provided by the Qatar National Library.