Discovery and Quantification of Microplastic Generation in the Recycling of Coated Paper-Based Packaging

Abstract

Microplastics (MPs) are found almost everywhere in the environment and the food chain. The long-term effects of MPs on living organisms are still unclear, so preventing anthropogenic MP generation has become crucial. Fibre-based packaging recycling is investigated here, shedding light on possible MP generation and its consequences. As a typically overlooked source of secondary MPs, cellulosic packaging often consists of thin polymeric coatings that can fragment during recycling. Dispersion coating technology for paper substrates is considered here. The coating formulation was tagged with rhodamine-B and investigated using semi-automatised techniques, including fluorescence microscopy, optical microscopy, and Raman spectroscopy. The results raise concerns as the coating under investigation (8 g/m²) broke into more than 75,000 secondary MPs, whose equivalent diameter and particle count density in the recycled material averaged 75.4 µm and 4.7 particles/mm², respectively. Wastewater analysis found finer particles (average equivalent diameter: 51.4 µm) with a higher particle count density (6.7 particles/mm²). Overall, 72% of the retrieved particles were smaller than 100 µm. Without proper wastewater screening, such particles (representing 87% in the wastewater filter) may enter the environment, hence representing a hazard for living organisms including humans.

1. Introduction

The issue of microplastics (MPs) and their environmental leakage has gained significant attention [1,2,3]. MPs are particles ranging from 1 μm to 5 mm [4]; much interest is on secondary microplastics generated due to breakdown processes [5,6]. Secondary MPs arise from chemical and mechanical weathering, abrasion, or photodegradation, primarily from land-based debris [6]. Such fragmentation occurs due to natural processes or human-related activities, leading to toxic accumulation in living organisms [7,8,9]. Preventing MP production, leaching, and transportation is essential; hence, different waste sources of MPs require exploration [10,11,12].

Globally, and in the European Union, packaging waste has consistently increased since 2010 [13,14]. In 2021, around 60% of the total packaging waste was cellulosic (34.0 Mt) and plastic (16.1 Mt) packaging [13].

Secondary MPs can be generated during various packaging industry stages, including production, transport, recycling, and disposal [1,15]. A previous review highlighted the need for advanced wastewater treatment methods for the effective removal of MPs [16], e.g., membrane bioreactors, electrocoagulation, or rapid sand filtration that can reach >98% MP removal efficiency. In one study [17], the analysis of plastic recycling facility wastewater samples detected MPs as small as a few micrometres, supporting previous findings; similarly, most MPs have been found to be smaller than 300 µm in water treatment plants [18].

Given the industry’s shift towards fibre-based packaging—known as the “paperisation” process—high-performance polymeric, inorganic, or hybrid coatings are being developed [19,20,21,22,23]. Among these, dispersion coatings are taking a share of the market. The introduction of non-fibrous materials in recycling streams must be managed to ensure secondary raw material quality and prevent anthropogenic MPs from entering the environment through wastewater. These MPs can enter the food chain via aquatic organisms and ultimately affect human health [24,25,26] by entering the body [9]. To the authors’ knowledge, what happens to polymeric coating particles during different stages of cellulosic packaging recycling remains undocumented.

Dispersion-coated fibre-based packaging is generally considered recyclable due to the low reject rates observed during recycling [27,28,29,30]. Coated substrates should be included in discussions on MP generation since coating fragments may bypass mechanical separation at paper mills, explaining the low amounts of coarse and fine rejects during lab testing. Current technical standards for assessing fibre-based product recyclability include the Italian recycling standard [31] and the European harmonised methodology [32]. The methodologies include the analysis of coarse and fine rejects, and of macrostickies separated by mechanical means. Moreover, parameters such as recycled material stickiness and optical inhomogeneities are assessed. Such methods are representative of the paper mill process at a laboratory scale. On the contrary, to date, no standard or technical procedure exists to measure or assess secondary MP release.

From a paper mill perspective, possible MP accumulation in water increases suspended solids and burdens on flocculation and sedimentation [33,34], possibly necessitating micro-filtering treatments to prevent environmental leakage. Before reaching the flocculation and sedimentation stage, MPs may contaminate the secondary raw material, i.e., recycled paper, which may raise concerns about particle migration in food contact applications. MPs might also be generated during the composting of packaging, further leaking into and contaminating the soil. Previous studies and standards [35,36] introduced multiple methodologies to identify and track MPs and determine their constituents in polymeric composites, including Raman spectroscopy, micro-FTIR, the fluorescent dyeing of polymeric particles, fluorescence microscopy, and image analysis. In one study, Tong et al. [37] investigated the best solvent for rhodamine-B (RhB) to observe and track MPs with fluorescence microscopy. RhB was physically adsorbed on the surface of multiple polymers, including polyethylene, polypropylene, and polystyrene, and did not fade after rinsing with water. In another study, Purington et al. [38] focused on a waterborne styrene–butadiene coating using RhB to understand binder migration in the cellulosic substrate through confocal laser scanning microscopy and fluorescence microscopy. Brown et al. [17] and Erni-Cassola et al. [39] employed Nile red dye for detecting MPs in environmental and industrial samples and quantifying common packaging polymer particles in the 20–1000 µm range, respectively. Nile red is similar to RhB as it is fluorescent and capable of visibly colouring the sample. However, it is poorly soluble in water, often requiring methanol or acetone as a solvent. Kwon et al. [40] tracked hydroxyethyl methacrylate using fluorescein o-methacrylate to tag latex in dispersion coatings.

The novel hypotheses of this work are that thin coatings on cellulosic substrates generate large amounts of micrometric secondary MPs during paper recycling, and that fluorescent species can help track, count, and measure them at different stages of the recycling testing methodology.

The focus in this work was on a styrene–butadiene dispersion coating tagged with RhB to provide contrast for detection and measurement purposes. As dispersion coatings are mainly water-based products, RhB was adopted as a tagging species due to its higher solubility. The aim of this work was to advance the knowledge of MP generation for dispersion-coated paper substrates, suggesting a method and comparing multiple analysis techniques that may be used to define a shared method to assess MPs in the paper recycling stream that integrates with current recycling testing methodologies. Recycling was tested according to the UNI 11743:2019 [31] standard at the laboratory scale. The study was limited to a specific family of polymeric dispersion coatings tagged before application, which may limit is applicability to already-coated substrates or products. Additionally, the experimentation was carried out at the laboratory scale, based on the hypothesis that current methodologies are representative of industrial-scale reprocessing. The adopted analysis techniques were suitable for extensive investigations. The authors wanted to provide preliminary data to raise possible concerns on the topic.

2. Materials and Methods

MetsäBoard (Espoo, Finland) Natural WKL Bright (130 g/m²), a fully bleached uncoated white kraftliner, was used as a substrate. A commercial waterborne styrene–butadiene latex similar to DL966—a grade broadly investigated in previous literature [41,42]—was used. It is highly cross-linked, showing a solid content of 53% and a T_g ≅ 15 °C. Rhodamine-B powder was bought from Sigma-Aldrich (St. Louis, MO, USA).

2.1. Rhodamine-B-Tagged Latex Preparation and Application

The preparation of RhB-tagged latex followed the description provided by Purington et al. [38]. Briefly, 0.03 wt% of RhB (on a dry weight basis) was added to the latex while on a magnetic stirrer that operated at 300 rpm for at least 16 h; while stirring, the beaker was kept closed using a paraffine film.

The substrate was coated with a K-Bar number 2 (12 μm wet film thickness) at a speed of 40 mm/s. Next, it was dried in a VWR (Radnor, PA, USA) VentiLine 112 Prime oven at 120 °C for 90 s and conditioned to 23 °C and 50% relative humidity for at least 24 h before further testing.

2.2. Physical and Mechanical Testing

The dry coat grammage was determined by the grammage difference between the coated and uncoated substrates, measured according to ISO 536:2019 on 6 measures. A Lorentzen & Wettre (Kista, Sweden) micrometre was used for the thickness measurement according to ISO 534:2011, averaging the result of 10 measures. An Andersson & Sørensen (Ishøj, Denmark) model 6 Bendtsen Tester was used to measure the surface roughness according to the Bendtsen method, ISO 8791-2:2013, averaging the result of 6 measures.

A Lorentzen & Wettre Gurley densometer was used to measure air permeance according to the Gurley method, ISO 5636-5:2013, averaging the result of 6 measures. Both coated and uncoated substrates were tested. An Acquati (Arese, Italy) AG/MC 1 Electronic Dynamometer measured the tensile strength and strain at break according to ISO 1924-2:2008 at a constant rate of elongation (20 mm/min), averaging the result of 5 measures. Both coated and uncoated substrates were tested.

All results related to substrate physical and mechanical testing are reported in the Supporting Information (SI).

2.3. Recycling

The recyclability test was carried out according to the UNI 11743:2019 standard with a few deviations from the specified test procedure (Figure S4), as follows. Due to material limitations, each test was carried out in a single replica using 25 g dry weight sample, maintaining the pulped batch consistency to 2.5%, as per the standard. Consistency was coherent to UNI 11743:2019 to ensure consistent results. The use of a single replica represents a limitation of this work; however, the repulping process with standardised equipment and operating parameters provides reproducible results, with limited uncertainty. The following procedure was added to the methodology defined in the standard to obtain samples for further analysis.

Visible and UV light analysis occurred before and after inking the macrostickies’ filter, as described in Section 2.5. Before sheet formation, the secondary raw material (fine screening accept pulp) was filtered on a 200-mesh wire and the wastewater was recovered in a dedicated container. The recovered wastewater was filtered on a filter paper of grade 1289 with a diameter of 240 mm (basis weight 84 g/m², filtration speed 20 s/10 mL, deposition range 8–12 μm) with the Lab Sheet Former. The filter was dried for 10 min at 93 °C, and sheet images were acquired and analysed to identify possible microplastics. The filter deposition range limits the minimum particle size retained on the filter, though solid buildup on the filter may enhance filtration efficiency initially by trapping finer particles.

In accordance with the testing method, the equipment included a Haage Anagramm Technologien GmbH (Peißenberg, Germany) AG 04 Disintegrator for the pulping stage, a FRANK-PTI (Birkenau, Germany) Somerville analyser for the coarse and fine screening, a Haage Anagramm Technologien GmbH (Peißenberg, Germany) Lab Sheet Former BBS-3 to produce recycled sheets, macrosticky filters, and wastewater filters.

2.4. Raman Spectroscopy and Fluorescence Mapping

A Horiba Jobin-Yvon (Kyoto, Japan) LabRam HR800 coupled with a Peltier-cooled charge-coupled device was used to subject samples to Raman spectroscopy. An Olympus (Tokyo, Japan) BX41 microscope with a 50× objective was coupled with the Horiba Jobin-Yvon (Kyoto, Japan) LabRam HR800 spectrometer to focalise the laser in a small sample region. The spatial resolution of the setup was 1 µm.

The analysis was performed for near-infrared (785 nm) and green (532 nm) laser sources. Figure S5 in the Supporting Information shows the Raman shift for latex, rhodamine-B, and cellulose fibres. Laser excitation at 785 nm emphasises vibrational bands and avoids fluorescence. Here, it was used to detect potentially tagged MPs and to confirm that cellulose fibres were not tagged with RhB.

Due to the low concentration of RhB relative to latex, laser excitation at 532 nm was also applied to detect the RhB fluorescence peak. The low concentration of RhB was not sufficient to observe characteristic Raman bands for this species; however, the excitation at 532 nm enabled the fluorescence identification of latex particles.

RhB-tagged particles and cellulose fibres were analysed from wastewater filters and the recycled sheet to confirm proper latex tagging and physical RhB adsorption stability.

The wastewater sample was also mapped at both 785 nm and 532 nm over a 100 × 100 µm area (10 µm measurement point spacing for 785 nm and 1 µm for 532 nm) to support RhB tagging and MP identification.

2.5. Microscopy

The substrate fibre constituents were investigated with an Olympus (Tokyo, Japan) BX-51 optical microscope after fibre reaction with Herzberg and Graff C reactive. The presence of microscopic coating fragments or fibre staining was assessed with an Olympus (Tokyo, Japan) SZ-61 stereomicroscope at all stages of the recycling methodology.

A Tescan scanning electron microscope (SEM) equipped with a Bruker (Billerica, MA, USA) energy-dispersive X-ray (EDX) probe was used to analyse coating fragmentation and its interaction with cellulose fibres. Before recycling, the coating thickness was measured. SEM analysis was performed on a 2 × 2 cm specimen for the recycled paper, macrosticky filter, and wastewater filter. The specimens were sputtered with a gold–palladium alloy and investigated for morphological (secondary electrons), constitutive (backscattered electrons), and chemical (EDX) information. EDX was used to confirm Raman analysis.

An EPSON (Suwa, Japan) V750 PRO flatbed colour scanner equipped with a colour charge-coupled device line sensor was used to retrieve sample scansions at 600–1800 dpi resolution and to measure the macrosticky area after recycling, alongside the image analysis software PTS DOMAS 3.02. Particle density was calculated as the total number of MPs divided by the scanned area, i.e., 80 × 80 mm².

Reflectance fluorescence analyses were performed on an Olympus (Tokyo, Japan) BX51WI fixed-stage upright microscope, mounting a U-MNIBA3 filter (excitation filter: 470–495 nm, emission filter: 510–550 nm). The exposure time was 20 ms. A 40× lens maximised the scanned field (2.2 × 1.7 mm). At least 20 random measurements were acquired for each sample to determine the average particle area of the scanned surface.

Filtered wastewater, the recycled material sheet, and the macrosticky filter were analysed.

The scanned and fluorescence microscopy images were analysed using ImageJ v.1.54g, computing the data with Microsoft Excel v.2409. Raman spectra were analysed using the HORIBA (Kyoto, Japan) Scientific LabSpec 6 spectroscopy suite.

3. Results

3.1. Material Preparation

RhB was successfully tagged on the latex particles, providing a bright-coloured dispersion formulation, as reported in Supplementary Section A. The coated substrates were characterised by the typical RhB intense colour, as shown in Figure S1. The possible presence of untagged RhB in the dispersion that could tag cellulose fibres upon coating application should be avoided. A first indicator was the absence of residual colour on the rod coater after tagged dispersion application, suggesting no free RhB molecules that could stain the equipment. To ensure that no untagged RhB was present in the waterborne dispersion, an optical microscopy observation of the pulped samples was carried out, qualitatively observing that the fibres maintained their original colour (see Supplementary Section E). Additionally, Raman spectroscopy analysis was conducted, as described in Section 3.4.

3.2. Coated and Uncoated Substrate Characterisation

RhB was successfully tagged on the latex particles, providing a bright-coloured dispersion formulation, as reported in Supplementary Section A. The coated substrates were characterised by the typical RhB intense colour, as shown in Figure S1. The possible presence of untagged RhB in the dispersion that could tag cellulose fibres upon coating application should be avoided. A first indicator was the absence of residual colour on the rod coater after tagged dispersion application, suggesting that no free RhB molecules were present that could stain the equipment. To ensure that no untagged RhB was present in the waterborne dispersion, the optical microscopy observation of the pulped samples was carried out, qualitatively observing that the fibres maintained their original colour (see Supplementary Section E). Additionally, Raman spectroscopy analysis was conducted, as described in Section 3.4.

3.3. Recycling

The results of the laboratory recyclability tests are reported in

Table 1. Recycling test results according to the modified methodology of UNI 11743:2019.

Parameters	Uncoated Substrate	Coated Substrate
Coarse Rejects [%]	<0.1	<0.1
Area of Macrostickies < 2000 µm [mm2/kg]	250	272,740
Fiber Flakes [%]	0.2	3.8
Recycled sheets—Adhesion test	Not detectable	Not detectable
Recycled sheets—Optical inhomogeneity	Level 1	Level 2 1
Assessment of recyclability according to Aticelca 501:2023 evaluation system	LEVEL A+	NOT RECYCLABLE with paper

¹ Due to the presence of rhodamine-B.

, which also reports the level of recyclability according to the Aticelca 501:2023 [43] evaluation system. The results indicate that the coated substrate investigated here can be considered non-recyclable; the outcome is solely due to the high levels of macrostickies, which depend on the coating formulation. Macrosticky analysis, related to adhesive particles that can endanger the continuous process at a paper mill, is currently optional in a recycling methodology similar to the one adopted in this work (i.e., Cepi); recent methodological advancement means that the role of this analysis is crucial. According to the results, and differently to the statements made in previous literature, not all dispersion coatings should be considered recyclable within paper recycling streams, especially when based only on the coating grammage or technology used. Very small amounts of coating (5.8% in this case) can render the paper technically non-recyclable.

The sample provided no or negligible coarse rejects and some fine rejects (3.8%), composed of fibre flakes and coating particles (Figure S7). This finding highlights the fragmentation tendency of the coating, the particles of which mostly pass through a 5 mm wide hole plate and are likely to pass through a 0.15 mm wide slot plate. Specifically, the RhB-tagged coating fragments were visible in the fine reject and the secondary raw material, possessing a slightly elongated shape.

Thanks to their bright colour, coating particles in the range of millimetres were visible to the naked eye in the recycled sheet, as well as in the macrosticky filter. The wastewater filter was characterised by a much finer particle distribution, barely visible to the naked eye. Filtering the wastewater generated some issues due to the presence of extremely fine fractions that almost clogged the pores of the filter paper, requiring long draining times. Yet, the stereomicroscope analysis highlighted an even finer RhB-tagged particle distribution, qualitatively demonstrating the presence of MPs even in the wastewater.

3.4. Raman Spectroscopy

RhB-tagged particles were identified using Raman spectroscopy; the relative spectra are shown in Figure 1a. The Raman shift of the latex (Figure S5) highlights peaks at 622 cm⁻¹ (C=C-H ring bending of styrene), 1003 cm⁻¹ (aromatic ring breathing of styrene), and 1033 cm⁻¹ (C-H in-plane bending of the ring) [44]. The characteristic peaks of RhB are not visible except for the coating particles in the wastewater sample, where a peak at 1530 cm⁻¹ can be observed. Furthermore, the presence of RhB can be inferred from the increasing fluorescence signal at lower wavenumbers, consistent with Figure S5.

Additionally, peaks associated with cellulose were observed around 390 cm⁻¹, 900 cm⁻¹, and 1100 cm⁻¹, with the latter being more defined. Indeed, cellulose fibres often surround the rhodamine-B-tagged polymer particles. The recorded Raman signal may also originate from regions beneath the tagged particle where cellulose fibres are present. The Raman spectra of fibres from a blank sample (i.e., untagged and uncoated cellulosic substrate) show spectra that overlap with the wastewater spectrum in Figure 1b, indicating that RhB does not leach during recycling.

Analysing the Raman shift for fibre samples (Figure 1b) shows that both the intensity of the signal and the peaks are consistent with those of cellulose fibres, as shown in Figure S5. No significant fluorescence backgrounds (potentially related to RhB) were observed at the analysed points, nor were characteristic wavelengths of either latex or RhB. A peak around 1083 cm⁻¹ was sometimes found in the wastewater analysis. In one case, the spectrum result was completely different from what was expected for cellulose fibres (see the cyan curve in Figure 1b). It is possible to highlight three characteristic Raman shifts for calcium carbonate: around 283 cm⁻¹, 714 cm⁻¹, and 1083 cm⁻¹ [45].

Raman spectroscopy was also used to perform fluorescence mapping of the surface, analogously to what was carried out with microscopy. The mapping procedure involves the collection of Raman spectra over a grid of points on the sample. For each grid point, we selected the 990–1018 cm⁻¹ band for latex and 1070–1159 cm⁻¹ band for the fibres. Regarding RhB, we focused on the 685–1774 cm⁻¹ band that represents a portion of its broad fluorescence emission band. For each band, a colour scale proportional to the peak area was automatically determined by the software. Figure 2a shows the mapping results obtained by excitation at 785 nm and 532 nm, displaying signals for latex and fibres (at 785 nm) and RhB (at 532 nm). The fluorescence signal for RhB is visible, demonstrating that it is directly associated with the latex particles, with a fluorescence intensity difference of over 60-fold between areas with and without particles. The dual capability of the Raman instrument to detect both Raman shifts and fluorescence enables comprehensive information to be obtained on the chemical nature of the analysis points, confirming that the observed signals correspond to tagged polymers. Tagging techniques can be effectively used for imaging, as demonstrated here, providing immediate feedback on sample characteristics. Moreover, a representative map at 532 nm is shown in Figure 2b (area: 350 × 350 µm), highlighting the presence of multiple and close MPs that were only partially detected by visible light imaging.

3.5. Visible Light Imaging

Information on the equivalent area occupied by secondary MPs according to multiple dpi resolution is reported in the Supporting Information.

Here, 1800 dpi scans were further analysed to determine the representative distribution of the MP size for each sample, as reported in Figure 3. The wastewater filter contained 42,981 particles, i.e., 6.7 particles/mm², whose average equivalent diameter was 51.4 ± 34.0 µm (median: 50.5 µm; min: 15.9 µm; max: 653.1 µm). Such results highlight the fact that secondary MPs elude mechanical screens and the sheet former mesh thanks to their reduced dimensions.

The analysis of the recycled material identified 30,166 particles, i.e., 4.7 particles/mm²; the average equivalent diameter was 75.4 ± 110.6 µm (median: 31.9 µm; min: 15.9 µm; max: 1527.7 µm), highlighting a broad particle size distribution. Compared to the observations for the wastewater filter, the lower median value highlights the fact that the recycled sheet holds several fine MPs. However, many large particles were mechanically separated in phase 4 (as in Figure S4) and collected by the macrosticky filter (total particle count: 4713; average equivalent diameter: 209.9 ± 222.3 µm; median: 107.0 µm; min: 16 µm; max: 1547.2 µm).

It should be considered that the counted particles are the ones visible on a specific side of the samples. Possible additional particles, partially or totally covered by fibres or located on the opposite side of the samples, are not included, implying that the generated MP number represents a minimum value. Moreover, the identified MPs with an equivalent diameter <20 µm could not be correctly determined due to resolution limitations, as their signal was within the image analysis baseline; therefore, their fraction was neglected in the calculations. However, there is indeed evidence of their presence when other techniques are used (see Figure S11e and Figure 2b).

The average MP size distribution shown in Figure 4, neglecting the fraction <20 µm because it is too close to the lower measurement range boundary. The growing fraction of finer particles moving from macrostickies to wastewater is explained by the mechanical filtering system sequence that intrinsically involves finer slots, as per the adopted recycling methodology. Therefore, from a broader perspective, Figure 4 also gives indication of the overall particle size distribution of the analysed macrostickies, recycled material, and wastewater filter samples.

3.6. Fluorescence Microscopy

Fluorescence microscopy analyses were performed randomly on previously analysed surfaces to achieve a more detailed reading thanks to a better resolution. The percentage of the total scanned area occupied by RhB-tagged particles is represented in

Table 2. Images acquired with fluorescence microscopy. Average area (± standard deviation) covered by RhB-tagged particles.

Sample	Area (%)
Macrostickies—Not inked	5.67 ± 4.98
Macrostickies—Inked	4.28 ± 4.04
Recycled material sheet	10.01 ± 2.81
Filtered waters	4.59 ± 0.95

.

Fluorescence microscopy returned values around twofold higher than visible light imaging (Figure S9). The reason for this is the higher magnification and radiation penetration of the former method. Higher magnification is associated with increased resolution (619 px/mm), whereas radiation penetration is associated with the signal acquisition of particles even partially covered by fibres.

The further manual measurement of 230 particles from wastewater micrographs (using ImageJ) returned an average length of 68 ± 38 µm (minimum: 14 µm; maximum: 202 µm). The size distribution of the manual measurement was coherent with the wastewater distribution shown in Figure 4. The selected particles were limited by the low visibility in the micrographs due to possible overlying fibres. Nevertheless, the result is coherent with the particle equivalent diameter measured and reported in Figure 3.

The measured standard deviation reported in Table 2 is high for the macrostickies (both inked and not inked) because the surface features large particles distributed on a broad surface and sometimes only partially present in the image. This is coherent with the sampling methodology.

The inking of the macrosticky filter hinders microplastic detection (Figure S11).

3.7. SEM

A low-magnification observation of the recycled sheet revealed difficulties in the identification of coating particles for both the secondary and backscattered electron detectors. The fibre braiding of the sheet held and hid some coating particles, preventing the clear identification of their borders.

The higher magnification of backscattered electrons highlighted brighter deposits (Figure 5c) on the coating particles’ surface. EDX analysis identified the presence of calcium, aluminium, silicon, and magnesium. The presence of CaCO₃ is coherent with the Raman analyses (Figure 1b) and is related to common pigments used in paper production. Aluminium, magnesium, silicon, and sulphur are associated with the presence of kaolin and talc, as well as other substances commonly involved in paper production.

The SEM observation of the MS filter (Figure 5d) allowed the authors to identify the perimeter of the coating particles on the filter surface that were not hidden by the fibre braiding. Imaging obtained with secondary electrons highlighted the different morphologies and fragmentation of the coating particles, with diameters ranging from 200 μm to 1 mm.

The presence of coating particles is also evident in the wastewater sample. Morphological images highlighted a greater coating fragmentation with particle diameters smaller than 100 μm.

4. Discussion

From a methodological perspective, the technologies adopted in this work proved adequate to identify the presence of RhB-tagged MPs in different fractions of paper-based material recycling. Due to its intrinsic characteristics, RhB behaved similarly to Nile red, the latter having been broadly discussed in recent literature [25,39,46]; both provide a characteristic colour under visible light radiation and feature fluorescence under specific laser wavelengths. Therefore, both document-scanning equipment and fluorescence microscopy—to be used alongside image analysis software, e.g., ImageJ—can spot the presence of coating fragments. Raman analysis confirms that the RhB-coloured fragments are actual anthropogenic MPs. Due to the high fluorescence of RhB at the given concentration under excitation at 532 nm, short scanning times (around 0.2 s) allow the determination of high-density (i.e., high-definition) maps. However, when analysing large areas—in the order of squared centimetres and decimetres—fluorescence imaging becomes helpful to achieve larger mapping. This is particularly true for the analysis time. The fastest technique for analysing the scanned area was visible light imaging, which also involves the use of affordable low-tech devices; however, it necessitates tagging species that provide high contrast to the fibres of the substrate. The possible use of UV scanners could be investigated to explore fast image acquisition.

Table 3. Advantages and disadvantages of the investigation techniques involved in this work.

Technique	Pros	Cons
Visible scanning imaging	- Fast and cheap - Analysis of large areas (up to dm2)	- Suitable only for coloured (visible light) tagging species - Quite low resolution—depends on device specs - Fibres might partially or totally cover MPs
Fluorescence microscopy	- Fast analysis - MPs underneath fibres are identified	- Suitable only for fluorescent species - Scanned area up to cm2
Raman spectroscopy	- MPs underneath fibres are identified - Maximum resolution - Chemical species can be identified - Also suitable for fluorescence analysis	- Time-intensive for mapping at 785 nm - Scanned area up to hundreds of µm2
SEM-EDX	- Broadly variable scanned area - Possible tagging substance identification	- Need to know the chemistry of the system

presents the main considerations about the techniques involved in this study.

Given the nature of the samples involved in this study, i.e., cellulose and polymeric coatings, the proposed method enables the analysis of MPs without the need for density separation or peroxide oxidation and enzymatic digestion, as in other standard practices used to detect MPs [35,36] that are not related to fibre-based materials. This type of workflow, besides neglecting covered or partially covered MPs in the recycled material or wastewater samples, may provide a fast and non-destructive method that can be integrated into the already-defined recyclability testing methodologies.

The obtained results are alarming. Despite the laboratory scale of this study, anthropogenic secondary MPs were produced and detected in large amounts. Due to their dimensions and considering the absence, or limited abundance, of filtration systems in current plants, MPs elude mechanical filtering systems and enter the environment and possibly the food chain [25,47]. As highlighted elsewhere [16], proper wastewater treatment should become a mandatory prevention mechanism. A recent study involving paper mills in Germany [48] confirmed that tertiary filtration can drastically reduce the volume of MPs entering the environment, otherwise reaching 10⁶–10⁸ MPs/m³. Another study related to Finland [49] reported 900–1600 MPs/kg in primary sludges. Moreover, previous studies typically assessed outbound process wastewater (

Table 4. Sample preparation, analysis techniques, and MP dimension of outbound waters or sludges at recycling facilities.

Ref.	Focus	Sample Preparation	Analysis Techniques	MP Dimensions
This work	Tracking MPs generated at lab scale during recycling process	Coating tagging with RhB	Visible light imaging, Raman spectroscopy and fluorescence mapping/microscopy	Mainly <100 μm; Lower limit: 15 μm
[48]	Outbound wastewater analysis of paper mills in Germany	Oxidative treatment + density separation	μ-Raman spectroscopy	Mainly <100 μm; Fraction >500 μm was negligible Lower limit: 20 μm
[49]	Sludges from a multi-purpose paper mill in Finland	Chemical-enzymatic digestion + density separation	Raman spectroscopy	Mainly <200 μm Lower limit: 20 μm
[17]	Outbound wastewater analysis of plastic recycling facilities in the United Kingdom	Digestion + Nile Red tagging	Fluorescence microscopy	Mainly <10 μm Lower limit: 2.6 μm
[50]	Outbound wastewater and sludge analysis of plastic recycling facilities in Türkiye	Density separation	Optic microscope	Mainly <250 μm Lower limit: 45 μm

). Some discussion in the industry is related to MPs present in the inbound water from paper mills. Previous research on industrial case studies did not shed light on how many MPs are generated during recycling compared to those already there and not separated in the process. This work provides further insights, since only the MPs generated during recycling were assessed, excluding possible interference from MPs already present in the process.

Paper and board packaging waste accounts for almost 41% of the total packaging waste generated in the European Union, exhibiting a positive trend over the last decade [13]. A remarkable share is related to functionalised substrates, typically featuring polymeric layers to improve barrier performance. Conventional one-sided coatings are easily managed at paper mills, whereas liquid containers and other structures with, e.g., water barrier coatings on both sides are generally managed by specialised facilities. Recent market trends show an increase in coated paper and board packaging on the market, with a growing multitude of functional layer materials. There are several reasons for this, from new regulatory frameworks and recent changes in design for recycling guidelines to the increased attention paid to recycling performance and testing methodologies and strong material innovation. The “paperisation” process requires higher-performance coatings, as well as thinner coating layers. This is coherent with the push to develop packaging that is recyclable at scale, and the development of “monomaterial” packaging, i.e., with <5% non-cellulosic content, as opposed to composite materials [51]. The more businesses that switch from polymeric to fibre-based packaging, the higher the number of coated products that reach paper mills and possibly fragment into fine secondary MPs. Therefore, defining a tracking and quantifying methodology for MP generation during fibre-based packaging recycling can support both developers and regulators to steer the decision-making process.

The identified secondary MPs possess a dimension that differs by one order of magnitude from macrostickies to wastewater. Wastewater does contain bigger particles, but most are smaller than those found in secondary raw materials and are typically undetectable to the naked eye. Assuming a circular shape, the equivalent diameter of wastewater MPs is coherent with the smallest MPs detected or measured in other works [15,25,52]. This work supports, as in previous literature [15,18,25], the need for the further investigation and treatment of fine MPs. Of greatest interest are fractions <0.1 mm, which constitute the majority of particles found within the wastewater analysed in this work and others that focus on sludges [48,49],

Therefore, possible wastewater treatment, e.g., membrane filtering or clarification, should be required for wastewater effluents from paper mills. Similar work related to the counting and identification of MPs in living organisms and sediments [26,53] highlighted the fact that prevention and proper policies should be mandatory, particularly in the case of recycling and production plants. Indeed, it is uncommon for conventional systems to intercept smaller MP fractions [48].

The results also showed that MPs had accumulated in the secondary raw material. Hence, these generated secondary MPs might embed in new products and possibly pollute other streams. As an example, consumers may improperly handle or sort packaging waste [54] by, for example, conferring fibre-based packaging contaminated with MPs into the organic fraction meant for compost production. MPs could also persist in the cellulose-based packaging supply chain, possibly further reducing their size in subsequent recycling processes, making their interception even more difficult and their related hazard higher.

As discussed in previous research [55,56,57,58], different coating grades, whether commercial or experimental, feature variable properties and therefore potentially different fragmentation behaviour during pulping. The behaviour depends on multiple factors, e.g., filler content and dimension, latex chemistry, glass transition temperature, and degree of cross-linking; such factors can also highly impact the recyclability outcomes, as observed in this study with the macrosticky area (Table 1). However, although this work only addresses a single type of latex, considerations drawn from these results can be applied to all similar systems. Indeed, under the hypothesis of similar behaviour at the industrial scale, the present results highlight critical issues that require further understanding and investigation.

The present tagging methodology proved feasible when the coating grade is available as dispersion, i.e., before the coating application and drying; evaluating samples that are already coated requires procedure modifications and further investigation.

5. Conclusions

The present study tracked the coating fragments produced during the recycling of coated paper-based packaging material. A tagging and analysis methodology of different fractions was proposed to successfully count and measure secondary microplastics. The main conclusions are as follows:

(1)

Rhodamine-B can tag styrene–butadiene latex. No residual rhodamine-B remained untagged for concentrations of 0.03 wt%, as observed with Raman spectroscopy analysis.

(2)

Secondary microplastics were found in the macrosticky filter, recycled material, and wastewater filter. The particle number increased and average size reduced from the macrostickies to wastewater filter, up to an average of 51 µm in the wastewater.

(3)

High-resolution scanning under visible radiation proved adequate for broad area investigation; however, fractions below 20 µm require further investigation due to limits on the adopted equipment.

(4)

Raman spectroscopy and UV microscopy possess higher resolution and allow the measurement of particles partially covered by cellulose fibres (in the recycled material or in the wastewater filter) or in proximity to the surface.