An evaluation of microplastics fate in the wastewater treatment plants: frequency and removal of microplastics by microfiltration membrane

In this research, the fate and removal of microplastics (MPs) entering and leaving a wastewater treatment plant (WWTP) was investigated. Additionally, application of the microfiltration membrane technique for MPs removal was evaluated. In the first phase, the quantity, type and size of the MPs were studied from three different points of the WWTP. The results showed that the average amount of MPs entering into the WWTP, accumulated in the sludge matrix and discharged from the effluent were 206, 183 and 94 MP/L, respectively. The MPs were observed mainly in the forms of fibers, pellets, and fragments, with a proportion of 35%, 39%, 22%, and 34%, 22%, 38%, and 31%, 39%, 37.7% in the influent, effluent and sludge of the WWTP, respectively. It should be noted, a minor amount of foam (2%) and film (0.3%) was also observed. The particle size distribution of the MPs in the effluent of the system was almost identical (6–14% for 1–5,000 μm), while the influent mostly contained particles within 1–100 μm (above 26%) with other particle sizes within 100–5,000 μm (8–17%). Particle size within a range of 500–5,000 μm was abundant in the sludge. In the second phase, a micro-filtration (MF) membrane with a pore size of 0.1 μm was applied to eliminate the MPs from the WWTP effluent. It was observed the application of the MF membrane technique could achieve MPs removal efficiency up to 98% in the effluent of the system.


INTRODUCTION
Existence of microplastics (MPs) in wastewater streams has become a major concern for many wastewater treatment plants (WWTP) (Brepols et al. 2020;Cuartucci 2020). MPs can be classified based on their color, shape, and polymer characteristics (Jung et al. 2018). In terms of particle size classification, MPs are defended as particles with an average diameter less than 5 mm (Talvitie et al. 2017). In terms of shape, MPs are found to be in the shape of granules, fibers, films, fragments, and pellets (Patel et al. 2009;Leslie et al. 2017). On the other hand, MPs can be categorized according to their chemical composition such as applied polymers during the manufacturing process of the origin plastics. Common polymers of MPs include polyethylene (PE), polyvinyl-chlorides (PVC), polystyrenes (PS), polyurethanes (PU), polypropylene (PP), polyethylene terephthalate (PET), polyamides (PA), and other polymers (Jung et al. 2018). Although the exact point source of MPs generation is unclear, these particles are usually by-products of industrial materials and household wastes like personal care products (i.e. soaps, face cleansers, lotions, etc) as well as food industry products (i.e. food containers, plastic bags, plastic bottles, etc) as primary MPs (Reisser et al. 2013;Lares et al. 2018;Rolsky et al. 2019). The other forms of MPs, known as secondary MPs, can be generated after the breakdown of large plastic wastes or primary MPs into small pieces during a harsh environment (Michielssen et al. 1970;Claessens et al. 2013). The Environmental Protection Agency (EPA) reported that about 80,000-219,000 tons of MPs are generated in the European countries annually, and consequently they are discarded into the environment (House of Commons Environmental Audit Committee 2016). As a results, the presence of these materials can impose many adverse impacts on the environment.
Considering the impacts of MPs on the aquatic environment, it is fair to say they are responsible for the physical, biological, or chemical distortion of water quality during the MPs fragmentation phase (Lusher et al. 2013;Wright et al. 2013). It was also reported that MPs can accumulate toxic materials and heavy metals of wastewater and sewage sludge in their texture by adsorption/absorption mechanisms (Bardi & Aminirad 2020;Naqash et al. 2020) and provide a suitable environment for attachment of microbial communities, viruses, and pathogens (Browne et al. 2011). Previous reports have shown that MPs can sorb endocrine disrupting chemicals, and emit them into the aqueous environment Naqash et al. 2020). Also, MPs can be consumed by different aquatic species and induce biological disorders (Possatta et al. 2011;Lusher et al. 2013;Magnusson & Wahlberg 2014). Moreover, the association of MPs into the water body can affect the removal efficiency of MPs from water/wastewater treatment plants due to their hard separation feature, and consequently, operational procedures can be impaired (Phillips & Bonner 2015). Therefore, removal of pollutants and toxic substances from MPs is inevitable in the aquatic environment.
Wastewater treatment plants (WWTPs) are partially able to remove fragments above 300 micrometers, while the rest of the particles that leave the system are mainly in the range of the nano-scale (Carr et al. 2016). Although the WWTPs can diminish the amount of these particles from influent, the huge quantity of MPs in the effluent of the system still pour into the aquatic environment (Mason et al. 2016;Talvitie et al. 2017). It was previously reported that preliminary, primary, secondary, and tertiary treatment processes can reduce the particles up to 6, 68, 90, and 95%, respectively (Blair et al. 2019). Furthermore, applying chemical, biological, and physical treatment processes can remove MPs up to 99%, while the rest of the particles are introduced into the environment (Liu et al. 2021). A recent investigation on MPs removal showed that the advanced oxidation (UV and O 3 ), anaerobic-anoxic-oxic, and aeration grit chambers processes can reduce the MPs particles by 71.7%, 54.5%, and 58.5% respectively (Yang et al. 2019). Therefore, innovative approaches still need to be proposed in order to reduce the amount of MPs from the effluent of WWTPs (Carr et al. 2016;Lay et al. 2017;Mintenig et al. 2017;Ziajahromi et al. 2017).
Recently, MPs removal through a tertiary treatment process as a final treatment by application of membrane technology has received increased attention due to its versatility, and higher MPs removal potential (Lu et al. 2016;Salerno et al. 2016;Lares et al. 2018). Previous studies have shown that membrane treatment presents a higher MPs removal efficiency (.99%) in comparison with the preliminary, primary, and secondary treatment processes (Sun et al. 2019;Iyare et al. 2020). However, the robust performance of the membrane process depends on the wastewater, physicochemical futures of MPs, as well as the membrane applied (Wang et al. 2018). It should be noted that membrane shape, pore size, polymer type, and fouling persistence are the most important factors that affect the removal efficiency of MPs (Ma et al. 2018;Poerio et al. 2019). Therefore, a primary knowledge about types and fate of MPs in WWTPs is required prior to adopting a suitable strategy for MPs removal from WWTP streams.
The main objectives of this study are: (i) to assess the physicochemical characteristics of MPs, and its quantity in WWTP streams and (ii) to evaluate the removal efficiency of MPs by an advanced treatment technology, the microfiltration membrane. To do this, first, the quantity and quality of the MPs were evaluated from three different points of the WWTP of Sari city in the north part of Iran. Then, the application of a micro-filtration (MF) membrane with a pore size of 0.1 μm was investigated to eliminate the MPs from the effluent of the WWTP.

Materials and sample collection
In this study, the wastewater treatment plant (WWTP) of Sari city in the north part of Iran was evaluated as a case study. This plant was designed with an average capacity of 18,000 m 3 /day. The collected sludge was attained from the influent and effluent of the secondary settling tank in different seasons. Three samples were collected using an electric pump (Godwin-GSL80) from the different levels of the settling tank (i.e. ∼1 m from the surface, middle, and bottom of the tank), as well as in the in/effluent of the settling tank,. Due to the separate system of wastewater collection in the study area, where wastewater sources are mainly from domestic and limited commercial sources and observing low variability in amount of MPS through random sampling, it was decided to take three samples with three replicates, which showed a reasonable consistency. Therefore, three samples were collected per season and labeled as Sample 1, Sample 2, and Sample 3, and collected in autumn, winter, and summer, respectively. The collected samples were taken to the laboratory and kept at 4°C prior to analyses. Then the collected samples were filtered to giving particles above 1 μm.

MPs extraction and detection assay
In order to extract MPs from the collected samples, 200 mL of hydrogen peroxide (H 2 O 2 ) solution (35%) was added to 500 mL of each sample to eliminate the organic matter. Then they were kept at room temperature for 7 days prior to detection by a microscope analyzer (OPTO-EDU A10.1018) (Foekema et al. 2013;Mathalon & Hill 2014;Wesch et al. 2016). The extracted samples were filtered by Whatman filter paper (grade 42, 2-μm pore size), and then they were kept at room temperature for 2 days prior to checking with an integrated HD stereomicroscope camera (model AC 230 V 50 Hz; LABEX, UK). According to volatile solids and total solids tests (Water Environment Federation 1999), above 90% of filtered samples were MPs. ATR-FTIR spectroscopy (model WQF510A, China), in the wavelength region of 400-4,000 cm À1 with 16 scans, was used to identify the chemical characteristics of MPs. The polymer type of filtered samples was identified based on the absorption frequencies at a specified bound. ImageJ software (2016), was used to present the size of MPs. Table 1 shows the characteristics of MPs, which were observed according to their size and type.

Operation condition of microfiltration membrane
After the phase of extraction and detection of MPs, an advanced treatment technology was applied to evaluate the MPs removal in the effluent of the WWTP. In this regard, a submerged microfiltration membrane (MFM) with 0.1 μm pore size was applied to separate the MPs from the effluent of the wastewater. The characteristics of MFM are shown in Table 2. The effluent of the system was manually pumped into the microfiltration tank with a working volume capacity of 50 L (i.e. depth-60 cm, length-30 cm, width-28 cm). The MFM system was operated semi-continuously to treat 2 m 3 of wastewater per day. A digital process controller was installed to monitor the performance of the microfiltration membrane through detecting the displacement of the water level of the tank. In the case of membrane fouling, the process controller orders the influent and effluent of the system to stop and the backwash process begins. Moreover, an aeration system was installed to provide agitation and prevent the membrane fouling as well as the flux declining.
The sludge at the bottom of the filtration tank was pumped out as a collected sample. Three samples were taken at different levels (bottom, middle and top) of the treated effluent tank, and then the samples were mixed as a unique collected sample. The collected samples were taken to the laboratory and kept at 4°C prior to analyses. The detection and separation of MPs from aqueous solution were similar to the previous section (see MPs extraction and detection assay). The collection, separation and detection of samples was done in triplicate. Figure 1 shows the schematic microfiltration membrane set-up.   Table 3 describes the MPs particles in the in/effluent and sludge of the secondary settling tank. As can be seen from Table 3, the highest amount of MPs was observed in the influent of the system with an average of 206 MP/L. The results of analyses showed that MPs were accumulated in the sludge at 183 MP/L. As previously reported, the accumulation of MPs in the sludge could have been due to the adsorption of these materials by the sludge matrix ). Furthermore, from Table 3, it can be found that the quantity of MPs in the influent and effluent of the WWTP varied corresponding to the season. It can be noticed that the number of MPs increased in the influent of samples 1, and 2 rather than sample 3 (Table 3). This phenomenon could have been due to climate change, by which rainfalls wash out the MPs into WWTPs (Xia et al. 2020). However, the settleability of MPs increased in sample 1 with an efficiency of 90.91%. The settleability of samples 1 and 2 was 88.9%, and 86.4% respectively (Table 3). From Table 3, the amount of MPs in the sludge samples of 1-3 was 200, 192, and 159 MP/L respectively. It was reported that the temperature and operational conditions (i.e. pH, dissolved oxygen, etc) have an indispensable role in sludge flocculation (Nouha et al. 2018). Thus, at the higher temperature, the sludge floc disintegrates and consequently results in low settleability of flocs. As a result of this phenomenon, the amount of MPs in the sludge matrix was lower in the summer (sample 3) due to sludge disintegration than that of autumn and winter (i.e. samples 1 and 2). From Table 3, it can be estimated that about 1,123.5 billion MPs enter into the WWTP annually. However, about 617.6 billion MPs are discharged into the aquatic environment annually from the effluent of the WWTP. Therefore, the post-treatments need to be applied after effluent of the system before the land application or direct discharge. Moreover, the incorporation of MPs into  the sewage sludge is above 1,202 billion MPs annually, hence the application of sewage sludge must be banned from the land application or should be treated prior to use. Taking the types of MPs into the account, from Figure 2 and Table 4, it can be seen that all types of MPs are found in the in/effluent and sludge of the secondary settling tank. The most abundant MPs is fibers, fragments, and pellets followed by a minor amount of foam and film in the sludge. The amount of film in the in/effluent of the system was negligible (i.e. almost 0.00%). Also, in the sludge of the system, a minor percentage of film (i.e. 0.3%) was observed. The same was also observed for foam with 2%. Thus, it can be concluded that the film and foam are either removed in the previous unit prior to the secondary settling tank or the existence of these MPs in the influent of WWTP is trivial.  On the other hand, the percentage of fragments, pellets, and fibers in all collecting stations (e.g. in/ effluent and sludge), was partially the same, ranging from 31-39%, 22-38%, and 22-39%, respectively (Table 5). Since these materials are heavier than film and foam, they tend to settle or stick to the sludge matrix. Consequently, the presence of these materials impairs the subsequent operational units such as anaerobic digestion, when the digesters are fed with MPs-concentrated sludge (Li et al. 2020;Aminzadeh et al. 2021). Therefore, an appropriate post-treatment should also be adopted to eliminate the MPs from sewage sludge (Noguchi et al. 2019). Figure 2(b) shows the different sizes and types of MPs particles in the WWTP. From the figure, it was observed that the current size of pellets, fibers, fragments, and foam was 500 μm, 1,151 μm, 368 μm, and 197 μm, respectively. Also, the typical MPs size was classified into six groups, namely 1-100 μm, 100-300 μm, 300-500 μm, 500-1,000 μm, 1,000-2,000 μm, and 2,000-5,000 μm (Figure 3). Figure 3 shows that the distribution of MPs according to the classified size is almost the same in the effluent. Therefore, MPs particles such as fragments, pellets, and fibers, which are abundant in the effluent of the WWTP, can be found in a wide range of sizes (i.e. 1-5,000 μm) (Table 4 and Figure 3). However, in the influent, the abundant particles are within the range of 1-100 μm (above 26%). It can be noticed from Figure 3, as the particle size of MPs increases, the distribution of MPs decreases in the influent of the system. This effect could have been due to the large MPs being removed in the previous WWTP units (Blair et al. 2019). At higher particle size (within the range of 100-5,000 μm), MPs are mostly accumulated in the sewage sludge due to their higher density and specific weight (Carr et al. 2016;Lares et al. 2018). The most abundant particle in the sludge was within a range of 500-5,000 μm with a proportion of 25-29%. Thus, it can be assumed that the WWTPs do not have an intrinsic potential to eliminate particles greater than 1 μm; therefore, a new approach for removal of MPs particles greater than 1 μm is proposed through incorporating a microfiltration membrane system, which will be discussed further.   The polymer features of collected MPs from in/effluent and sludge of settling tank were analyzed using an FTIR spectrometer with 16 scans and the spectra from 4,000 cm À1 to 450 cm À1 (Figure 4). A minimum of three adsorption bands was detected within 1,300-1,690 cm À1 in the collected MPs from the influent (Figure 4(a). These functional broad bands were found to be abundant in methyl group in branched low-density polyethylene, and polyvinyl chloride (Nishikida & Coates 2003;Noda et al. 2007;Asensio et al. 2009;Brandon et al. 2016). The broad band within the region of 1,330-1,400 cm À1 represents C-H stretch, and CH 2 bend and was found to be polypropylene (Asensio et al. 2009). The functional region at 2,260 cm À1 shows the CN stretch group and can be assigned to nitrile (Verleye et al. 2001). Also, from Figure 4, broadband within 2,970-3,400 cm À1 , with a maximum of 3,434, is presented. This band can be attributed to aromatic C-H stretch, implies the presence of polystyrene (Verleye et al. 2001;Noda et al. 2007;Asensio et al. 2009).

MPs removal by microfiltration membrane (MFM)
MFM technique is a versatile technology for water and wastewater treatment, and it is estimated that the application of different membranes for MPs removal will increase in the following years Poerio et al. 2019;Chuanwen et al. 2020;Lorain et al. 2020). Since the selection of MFM is based on the composition of treated wastewater, and the particle size of MPs, deep knowledge about types and size distribution of MPs are required (i.e. it was evaluated in the previous section) (Cerón-Vivas & Noyola 2017; Li et al. 2018). From Table 4, it can be seen that higher MPs are entered into the MFM tank (sample 1) before discharge with 106 MP/L. However, only 2 MP/L was observed   Table 4, it can be concluded that the number of MPs affects the efficiency of MFM since, at a higher amount of MPs, the removal efficiency of MFM declined. Also, the detected particles after the MFM system were below 0.1 μm, which implies that the particles were within nano-scale. It should be noted that our findings are fitted by other investigations (Carr et al. 2016). Mintenig et al. (2017) reported that the application of the post-filtration unit (e.g. pile cloth media filtration) could completely remove the MPs particles greater than 500 μm, and particles within the range of 20-500 μm with a removal efficiency of 95%. Also, Ziajahromi et al. (2017) found that the ultrafiltration membrane can significantly reduce the MPs to 0.28 MP/L in the effluent of the system.

CONCLUSIONS
The quantity of MPs was evaluated in the in/effluent and sludge of WWTP. The results showed that about 1,123.5 billion MPs are entered into the WWTP, and about 55% leave the WWTP, annually. Also, the results showed that 1,202 billion MPs accumulate in the sludge, annually. Most of the MPs were fibers, fragments, and pellets, and only a minor percentage of foam and film was observed.
The heavier particles are trapped in the sludge matrix, although the smaller particles within the range of 1-300 μm are observed in the influent of the system. The trapped particles in the sludge were in the range of 500-5,000 μm, while in the effluent of system the distribution of particle size is identical with a proportion of 6-14%. In this study, the advanced MFM could achieve MPs removal up to 98% in all samples. Therefore, this study provides knowledge for evaluating the fate of MPs in the WWTPs and possibly eliminates MPs in the effluent of WWTP prior to discharge.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.