Paper on the developed characterisation methods for SMPs (< 10 μm) and NPs (< 0.1 μm), and in terms of: (i) chemical identity of the SMPs/NPs polymer type; (ii) physical particle characterisation and quantification, size distribution and particle morphologies; and (iii) quantification of the mass fraction in complex matrices. Uncertainty evaluation and traceability statements will be included

Confidentiality Status: PU - Public, fully open Deliverable Cover Sheet Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or EURAMET. Neither the European Union nor the granting authority can be held responsible for them. The project has received funding from the European Partnership on Metrology, co-financed from the European Union’s Horizon Europe Research and Innovation Programme and by the Participating States. 1 of 2 21GRD07 PlasticTrace Deliverable 6 Paper on the developed characterisation methods for SMPs (< 10 μm) and NPs (< 0.1 μm), and in terms of: (i) chemical identity of the SMPs/NPs polymer type; (ii) physical particle characterisation and quantification, size distribution and particle morphologies; and (iii) quantification of the mass fraction in complex matrices. Uncertainty evaluation and traceability statements will be included Organisation name of the lead participant for the deliverable: LGC Limited (LGC) Due date of the deliverable: 01.07.2025 Actual submission date of the deliverable: 30.09.2025 2 of 2 Executive Summary Reliable reference and quality control materials are essential for advancing the measurement and risk assessment of nanoplastics. However, materials with physicochemical properties comparable to those of real environmental and food contaminants are still lacking. Within this work, presented in the paper “Multiparameter characterisation of a nano-polypropylene representative test material with fractionation, light scattering, highresolution microscopy, spectroscopy, and spectrometry methods”, we describe the development and comprehensive characterisation of a polypropylene nanoplastic material produced through a top-down mechanical fragmentation process. The material is homogeneous, stable in suspension, and representative of environmental nanoplastics. A wide range of analytical techniques — including AF4, cFFF, PTA, (MA)DLS, MALS, SEM, AFM, TEM, STEM, EDS, Raman, ICP-MS and py-GC/MS — were employed to determine its particle size and distribution, particle number concentration, polymer identity, mass fraction, and inorganic impurity content. Results show a broad size distribution (50–200 nm), consistent polypropylene composition with signs of surface oxidation, and trace levels of inorganic impurities. The study demonstrates how complementary measurement approaches can be combined to improve the reliability, uncertainty evaluation, and traceability of nanoplastics characterisation. This work directly contributes to the PlasticTrace deliverable “Paper on the developed characterisation methods for SMPs (<10 μm) and NPs (<0.1 μm)”, addressing the chemical identification, physical characterisation, and quantification of nanoplastic materials, and providing a reference framework for harmonised and comparable measurements across laboratories. 1 Multiparameter characterisation of a nano-polypropylene 1 representative test material with fractionation, light scattering, 2 high-resolution microscopy, spectroscopy, and spectrometry 3 methods 4 5 Dorota Bartczak1, Aneta Sikora1, Heidi Goenaga-Infante1, Korinna Altmann2, Roland 6 Drexel3, Florian Meier3, Enrica Alasonati4 Marc Lelong4, Florence Cado 4, Carine Chivas7 Joly 4, Marta Fadda5, Alessio Sacco5, Andrea Mario Rossi5, Daniel Pröfrock6, Dominik 8 Wippermann6, Francesco Barbero7, Ivana Fenoglio7, Andy M. Booth8, Lisbet Sørensen8, 9 Amaia Igartua8, Charlotte Wouters9, Jan Mast9 , Marta Barbaresi10, Francesca Rossi11, 10 Maurizio Piergiovanni10, Monica Mattarozzi10, Maria Careri10, Thierry Caebergs12, Anne11 Sophie Piette12, Jeremie Parot13 and Andrea Mario Giovannozzi5, * 12 13 1National Measurement Laboratory, LGC Limited, 10 Priesley Road, Guildford, UG2 7XY, United Kingdom 14 2Bundesanstalt für Materialforschung und – prüfung (BAM), Unter den Eichen 87, Berlin, Germany 15 3Postnova Analytics GmbH, Rankinestr. 1, 86899 Landsberg a. Lech, Germany 16 4Laboratoire National de Métrologie et d’Essais (LNE), 1 rue Gaston Boissier, 75015 Paris, France 17 5Istituto Nazionale di Ricerca Metrologica (INRiM), Strada delle Cacce 91, 10135 Torino, Italy 18 6Helmholtz-Zentrum Hereon, Institute of Coastal Environmental Chemistry, Inorganic Environmental 19 Chemistry, Max-Planck Str. 1, 21502 Geesthacht, Germany 20 7Department of Chemistry, University of Torino, Torino, Italy 21 8SINTEF Ocean, Department of Climate and Environment, Trondheim, Norway 22 9Trace Elements and Nanomaterials, Sciensano, Groeselenberg 99, 1180 Uccle, Belgium 23 10University of Parma, Department of Chemistry, Life Sciences and Environmental Sustainability, Parco Area 24 delle Scienze 17/A, 43124 Parma, Italy 25 11IMEM, CNR Institute of Materials for Electronics and Magnetism, Parco Area delle Scienze 37/A, Parma, 26 43124, Italy 27 12FPS Economy, DG Quality and Safety, Metrology Division (SMD), Bd du Roi Albert II, 16 – 1000 Brussels, 28 Belgium 29 13SINTEF Industry, Department of Biotechnology and Nanomedicine, Trondheim, Norway 30 31 *Corresponding author: [email protected] 32 33 34 35 36 2 Abstract: Reference and quality control materials with comparable physicochemical properties to 37 nanoplastic contaminants present in environmental and food nanoplastics are currently lacking. Here 38 we report a nanoplastic polypropylene material prepared using a top-down approach involving 39 mechanical fragmentation of larger plastics. The material was found to be homogeneous and stable 40 in suspension and has been characterised for average particle size, size distribution range, particle 41 number concentration, polypropylene mass fraction and inorganic impurity content using a wide 42 range of analytical methods, including AF4, cFFF, PTA, (MA)DLS, MALS, SEM, AFM, TEM, STEM, 43 EDS, Raman, ICP-MS and pyGC-MS. The material was found to have a broad size distribution, 44 ranging from 50 nm to over 200 nm, with the average particle size value dependent on the technique 45 used to determine it. Particle number concentration ranged from 1.7 - 2.4 E10 g-1, according to PTA. 46 Spectroscopy techniques confirmed the material was polypropylene, with evidence of aging due to 47 an increased level of oxidation. Measured mass fraction was found to depend on the marker used 48 and ranged between 3 - 5 µg g-1. Inorganic impurities such as Si, Al, Mg, K, Na, S, Fe, Cl and Ca 49 were also identified at ng g-1 levels. Comparability and complementarity across the measurement 50 methods and techniques is also discussed. 51 52 Keywords: nanoplastic, particle size, number concentration, polypropylene identification, 53 polypropylene mass fraction 54 55 3 1. Introduction 56 Over 350 million tons of plastic waste are produced globally every year, of which nearly two-thirds 57 are estimated to be released into the environment as plastic waste [1]. Numerous international 58 organisations, including the United Nations (UN), World Health Organization (WHO), and 59 Organisation for Economic Co-operation and Development (OECD) have called for actions to 60 increase our understanding and to propose effective mitigation measures to protect the public and 61 the environment from plastic pollution. In Europe, the European Commission (EC) has responded 62 through policy documents, including the Green Deal and the EU Plastics Strategy. The EC has also 63 advanced legislation, including the Registration, Evaluation, Authorisation, and Restriction of 64 Chemicals (REACH) and the Drinking Water Directive (DWD, 2020/2184 [2]). In the United Kingdom, 65 the Government has set a target of eliminating avoidable plastic waste by end of 2042 to prevent 66 further pollution with plastics. 67 Despite joint international efforts to reduce the amount of plastic waste, it continues to be released 68 into the environment at an unprecedented scale [3]. Once in the environment, plastics undergo 69 fragmentation upon exposure to UV radiation and through mechanical stress [4]. Larger pieces of 70 plastic are fragmented into microplastics (MP), defined as particles in the size range from 1 µm to 1 71 mm [5] or 5 mm [6], and eventually to nanoplastics (NP), with sizes below 1000 nm [7]. MP and NP 72 have been reported in all environmental compartments, accumulating in soils and sediments, and 73 considered critical persistent pollutants of an increasing global concern owing to their high durability 74 and long-life [8]. 75 The potential long-term impacts on biota and human health arising from MPs and NPs present in the 76 environment are still unknown, and there are no defined maximum daily exposure limits due to the 77 lack of robust toxicological data [9]. This is, in part, due to the absence of metrologically validated, 78 harmonised and standardised measurement methods, as well as the lack of consensus with regards 79 to the typical quantities (per size class, especially for particles with sizes below 1000 nm) and 80 physicochemical characteristics of various types of plastics typically occurring in the environment. 81 There remains a need to develop robust analytical methods for plastic characterisation, especially 82 at the sub-micron and nanoscale. 83 The characterisation and quantification of plastic particles <1000 nm is challenging due to their 84 multimodal and polydisperse character and irregular shape, but also due to the limitations of 85 analytical techniques routinely used for the characterisation of other types of particles in the size 86 range from 1 – 1000 nm. For instance, commonly used light scattering techniques, such as dynamic 87 light scattering (‘DLS’ or Multi-Angle DLS ‘MADLS’ [10]), multi-angle static light scattering (MALS 88 [11]) or Particle Tracking Analysis (PTA [12]), as well as high-resolution microscopy methods, such 89 as transmission electron microscopy (TEM [13]), Scanning electron microscopy (SEM [14]) or atomic 90 force microscopy (AFM [15]) are compatible with particles in that size range, but they cannot easily 91 distinguish between plastic particles and other types of particles that might also be present in the 92 sample [16]. Conversely, spectrometry-based methods, such as pyrolysis gas chromatography mass 93 spectrometry (pyGC-MS), can distinguish plastic particles but do not provide information about 94 particle size or morphology [16]. Moreover, most of the imaging spectroscopy methods, such as 95 micro-Raman (µRaman) or micro-Fourier transform infrared spectroscopy (µFTIR), have size 96 detection limits >1000 nm and are therefore unable to characterise and quantify particles in the 97 nanoscale [17]. 98 These limitations with individual techniques highlights the need for the development of alternative 99 multi-technique, hyphenated (e.g. multidetector Field-Flow Fractionation ‘FFF’ [16]) or hybrid 100 approaches (e.g. SEM/RAMAN [18], Dielectrophoresis (DEP)-Raman [19, 20]) involving a 101 combination of measurement methods, allowing simultaneous characterisation of multiple 102 parameters at the nanoscale, including chemical identity, particle size, size distribution, as well as 103 4 mass and number concentration. Such combined approaches will be invaluable in supporting 104 policymakers, international standardisation efforts and testing laboratories, and will enable the 105 development and characterisation of future NP quality control (QC) measures and reference 106 materials (RMs), which are currently unavailable. 107 This work describes a systematic evaluation of the applicability of selected light scattering, 108 fractionation, high-resolution microscopy, spectroscopy and spectrometry methods, as standalone 109 techniques and in combination (including hyphenated and hybrid), for characterising plastic particles 110 <1000 nm. Assessment was conducted using a combination of commercially available polystyrene 111 (PS) microspheres and more environmentally relevant polypropylene (PP) nanoplastic particles 112 produced through a top-down method. The complementarity and comparability of the different 113 analysis techniques are discussed, including when applied in combination, along with main sources 114 of measurement errors for selected techniques and measurands. 115 116 2. Experimental Section 117 2.1 Materials 118 2.1.1 Monodispersed polystyrene spheres 119 A single batch of PS microspheres of approximately 200 nm (202 ± 4 nm), k=2 (Duke 3200A) was 120 purchased form Thermo Fisher (Fremont, CA) to use as a quality control material (QC). The material 121 was characterised by the manufacturer for size using TEM and was supplied with indicative 122 information on the solid content. In addition, 60 nm PS beads with a certified size (60 nm ± 4 nm, 123 Nanosphere™ Size Standard 3060A) were purchased from Thermo Fisher Scientific (Waltham, MA, 124 USA). 125 126 2.1.2 Polydispersed polypropylene particles 127 A polydispersed nanoPP test material containing particles <1000 nm in size was produced by 128 fragmenting PP pellets with an UltraTurrax and subsequently applying filtration (Figure 1). 129 Approximately 25 g of PP pellets were placed into a glass beaker containing 250 mL acetone. The 130 beaker was cooled on ice for ~30 min, after which the suspension was filtered through a pleated filter 131 to remove larger particles. The volume of the resulting filtrate was then reduced to 10% (ca. 25 mL) 132 by rotary evaporation and diluted with ultrapure water. The suspension was subjected to rotary 133 evaporation again to remove the remaining acetone (~30 min at 100 mbar). The resulting aqueous 134 suspension was filtered again. The final filtrate was divided into 890 brown glass vials, each 135 containing ~2 mL of PP suspension. 136 The homogeneity of the nanoPP material was investigated by analysing the content of 12 different 137 bottles in duplicate by PTA to determine particle size and number concentration (as described in 138 Section 2.2). Material stability was tested periodically using batch DLS analysis (as described in 139 Section 2.2) at time 0, 1 month, 3 months and 6 months, covering the period over which all 140 measurements described in the work reported here were performed. Three bottles were measured 141 in duplicate at each time point. All materials were diluted, as required by each technique, in ultrapure 142 water (18.2 MΏ cm, 25 ⁰C, Elga Purelab Chorus, Viola Water Technologies, Buckinghamshire, UK) 143 prior to analysis, unless stated otherwise. 144 5 145 Figure 1: Fragmentation of PP pellets with UltraTurrax (left), filtration of fragmented PP for fractionation 146 (centre) and the final bottled PP suspension (right). 147 148 2.2 Instruments and Methods 149 2.2.1 PTA 150 A NanoSight NS300 (PTA; Malvern Panalytical Ltd., UK) system equipped with a laser module with 151 a wavelength at 405 nm, high sensitivity scientific complementary metal-oxide-semiconductor 152 (sCMOS) camera, a syringe pump and Low-Volume-Flow-Cell (LVFC), was used by Postnova 153 Analytics and LGC. UNITO used a ZetaView® PMX-120 PTA (Particle Metrix GmbH, Germany), 154 equipped with a light source wavelength of 488 nm and a 90° laser scattering video microscope with 155 10× magnification. The instruments were switched on 30 min prior to measurement. The flow156 through cells were cleaned with ultrapure water before each new sample injection, as well as 157 between individual aliquots of the same sample and at the end of all measurements, until no more 158 particles are detected. For the NanoSight NS300, each sample was diluted gravimetrically ~100 x in 159 ultrapure water to a final particle number concentration of approximately 20-100 particles per frame 160 and shaken manually prior to analysis. Measurements were performed in a flow mode, and the 161 camera settings and focus were optimised manually. All measurements were conducted at room 162 temperature, allowing the instruments to automatically determine the actual temperature in the flow 163 cell and assign an associated water viscosity value for data analysis. Captures with a duration of 60 164 s were recorded and repeated 5 times per sample. For analysis by PMX-120, each sample was 165 diluted ~700 times (gravimetrically). The sensitivity and shutter were set to 80 and 100, respectively, 166 with a frame rate of 30 fps and a minimum track length of 15 frames. For each sample, 3 sets of 33 167 videos (1 second each) were recorded, analysing a minimum of ∼ 2500 NPs per measurement. 168 169 2.2.2 DLS (MADLS) 170 A MADLS Zetasizer Ultra instrument (Malvern Panalytica, UK)l equipped with 173⁰, 90⁰ and 13⁰ 171 angles and a low-volume, high-performance black Quartz Cuvette (ZEN 2112) was used by LGC 172 and INRIM. The MADLS instruments were switched on at least 30 mins prior to analysis. At LGC, 173 the sample was analysed both undiluted and following an ~10x dilution in ultrapure water. Individual 174 samples were measured 3-5 times under the following repeatability conditions: analysis temperature 175 25 °C, with media viscosity set to 0.8872 mPa.s and the refractive index (RI) set to 1.33, while the 176 material RI was set to 1.49 and the absorbance set to 0.01. Multiple narrow peak mode was used. 177 At INRIM, 5 runs per measurement were conducted on each undiluted sample. The operating 178 temperature was maintained at 25 °C. The average hydrodynamic diameter (z-average) and 179 6 polydispersity index (PDI) were obtained from the correlation function fitted according to ISO 180 22412:2017. 181 182 2.2.3 FFF-MALS 183 At Hereon, centrifugal FFF (cFFF) measurements were performed using a CF2000 system 184 (Postnova Analytics, Landsberg a. L., Germany) equipped with an autosampler, a degasser unit and 185 a UVD disinfection unit. The system was equipped with an analytical fractionation channel with a 186 thickness of 231 µm, a channel area of 100 cm2 and a void volume of 2 mL. A typical injection volume 187 of 20 µL was used for the measurements. The method employed a start speed of 3500 rpm and 188 maintained a flow rate of 1.5 mL/min. The injection time was set to 38 seconds, followed by a 189 relaxation time of 5 min. The carrier liquid was 0.2% (v/v) NovaChem100. The cFFF system was 190 connected to a MALS detector (PN3621). After each run, a rinse step of 10 min was conducted to 191 overcome potential carry over effects and ensure reproducible conditions for subsequent 192 measurements. At the end of measuring each batch of samples, a blank run was performed injecting 193 only MilliQ. In total, 3 samples with 4 replicates each were fractionated and characterised. Samples 194 were diluted five-fold in MilliQ prior to analysis. Evaluation of the data obtained from the MALS 195 detector was conducted by applying a sphere model to the scattering data. Typically, the range of 196 12° to 156° was used for data evaluation, where 68° and 132° were excluded from evaluation as 197 these angles did not provide sufficient data owing to the detectors needing to be replaced. 198 199 At Postnova Analytics, multi-detector (MD)-AF4 experiments were performed on a AF2000 MT 200 system (Postnova Analytics, Landsberg a. L., Germany). An analytical AF4 channel with a tip-to-tip 201 length of 277 mm, a width of 20 mm and hip width of 5 mm was equipped with a regenerated cellulose 202 membrane (RC) with a molecular cut-off of 10 kDa and a 350 µm spacer height. The temperature 203 during fractionation was kept constant at 25 °C using a channel thermostat. The samples were 204 injected using an autosampler. The fractionation system was directly coupled to a UV/Vis detector 205 and a MALS detector (21 active angles, laser wavelength 532 nm). The UV absorbance was 206 measured at 254 nm. The MALS detector was normalised using fractionated 60 nm PS beads and 207 a spherical fit model. A volume of 20 uL of a 40 ppb PS bead suspension was injected as received, 208 with no sample preparation performed. The carrier liquid consisted of 0.2% (v/v) NovaChem100 209 (PN). For the fractionation, a detector flow rate of 0.50 mL/min, an injection flow rate of 0.20 mL/min 210 and an injection time of 5 min was applied. The initial cross flow rate was set to 1.20 mL/min. After 211 a transition time of 0.2 min, the cross-flow rate was kept constant for 0.2 min and then decreased 212 within 40 min using a power decay (exponent = 0.2) to 0.10 mL/min. This last cross flow rate was 213 kept for 10 min, followed by a rinse step of 5 min. All measurements were conducted in duplicate. In 214 total, 12 vials (each with 2 aliquots) were fractionated and characterised. MALS data evaluation was 215 performed within an angular range of 12° to 156°, the scattering intensities were evaluated by fitting 216 a sphere model to the angular dependent scattering data to obtain size information (radius of 217 gyration, Rg). This type of fit model yielded results with low deviations across the complete size range 218 with squared correlation coefficients above 0.98. The sphere model represented the data points 219 accurately around the peak maximum. Slight deviations for the smallest and largest size fractions 220 were observed, but comparable results were derived from a fourth order polynomial fit. The system 221 was controlled by the NovaFFF Software (version 2.2.0.1) and the data evaluation was performed 222 in the NovaAnalysis software (version 2408). 223 At LNE, MD-AF4 analysis was performed on an AF4 system (AF2000 Postnova Analytics) coupled 224 to MALS (DAWN HELEOS II, Wyatt Technology) equipped with 18 angles and UV (SPD-20A, 225 Shimadzu) detectors. A metal-free analytical AF4 channel (tip-to-tip length of 277 mm, 20 mm width 226 and 5 mm hip width) (Postnova Analytics) was used. The carrier liquid was prepared by dissolving 227 Novachem100 (Postnova Analytics) in ASTM Type I ultrapure water to a final concentration of 228 7 0.0125% (v/v) and passing through a 0.1 μm filter (RC, Postnova Analytics). The channel out-let flow 229 rate was 0.5 mL/min. MALS data treatment was performed using the Berry model of second degree, 230 11 angles and the Astra Software (version 6.1.7, Wyatt Technology). The UV detector was used for 231 MALS data treatment and for recovery calculation. The uncertainty associated with the Rg was 232 established as the combination of the repeatability, the average MALS model uncertainty and the 233 size bias between the certified value and the measured value of a 200 nm PS standard. Blanks 234 consisting of pure carrier liquid were injected between samples and no carry over was observed. 235 Aliquots of the nanoPP suspension were characterised without any dilution or further sample 236 treatment (e.g. filtration or ultrasonication). 237 At SMD, MD-AF4 analysis was performed on a Wyatt Eclipse DualTec, equipped with a UV/Vis 238 detector (Agilent G7114A), a MALS detector (DAWN HELEOS II, Wyatt Technology), and an inline 239 Malvern Zetasizer Nano ZS (Malvern Panalytical, UK). The carrier was SDS (0.01% m/v) in ultrapure 240 water, filtered through a 0.1 µm filter (RC, Millipore). The separation method was an isocratic 241 program for most of its duration and separation. The UV/Vis wavelength was set to 200 nm, and 242 blank subtraction applied. 243 244 2.2.4 TEM and STEM-EDX 245 At Sciensano, conventional TEM imaging was performed using a Tecnai G2 Spirit 12 (120 kV, 246 Thermo Fisher Scientific, Eindhoven, The Netherlands) with BioTwin lens configuration equipped 247 with a 4X4K Eagle CCD camera and using TIA software (Thermo Fisher Scientific). In addition, a 248 Talos F200S G2 (200kV, Thermo Fisher Scientific) equipped with a Ceta 16M camera, high angle 249 annular dark field (HAADF) detector, Super-X detector and Velox software (Version 3.8, Thermo 250 Fisher Scientific) was used for scanning TEM coupled with energy dispersive x-ray spectroscopy 251 (STEM-EDX) to perform chemical mapping. The 200 nm PS QC material was diluted 10 times using 252 MilliQ water and deposited on Alcian blue pre-treated pioloformand carbon-coated copper grids 253 (Agar Scientific, Essex, England) by grid-on-drop deposition (10’ contact) [21]. The size properties 254 of the particles were measured semi-automatically from TEM images using the ParticleSizer plugin 255 [22] in the ImageJ software. The total measurement uncertainty was determined by a validation 256 study, following an approach similar to Verleysen et al. [23], consisting of three replicate 257 measurements per day for 5 consecutive days and measuring at least 500 particles per 258 measurement (see detailed methodology in Electronic Supplementary Information, section E). The 259 nanoPP suspension was used undiluted and different methods for the TEM grid preparation were 260 tested: untreated, Alcian blue pre-treated and glow discharged pioloformand carbon-coated copper 261 grids (Agar Scientific, Essex, England). Sample deposition on the grid was done by either grid-on262 drop deposition (10’ contact), drop-on-grid deposition followed by evaporation drying [21] (overnight), 263 or on-grid ultracentrifugation. Size properties of particles were measured manually from TEM images 264 in ImageJ. 265 At the University of Parma, TEM analysis was performed using a JEOL JEM-2200FS field-emission 266 microscope equipped with an EDX detector (Oxford Xplore), operated at an acceleration voltage of 267 200 kV [24, 25]. Ultra-thin carbon-coated copper grids (200 mesh, Electron Microscopy Society) 268 were selected to provide optimal support. The images were recorded in both TEM imaging mode 269 using a Gatan UltraScan US1000 camera and in STEM imaging mode using a HAADF detector. For 270 each grid, 20 micrographs were acquired in random sampling mode, ensuring coverage of both the 271 border and centre regions of the grid. Ultrapure water blanks were analysed to detect any potential 272 contamination introduced during sample preparation and handling. Particle counting and sizing were 273 performed using the open-source ImageJ software. The intensity range was adjusted to isolate 274 particles from the background. To improve particle separation and minimise noise, manual 275 thresholding and morphological filtering were applied. Sample preparation was performed in a 276 14 488 489 Figure 3. Average AF4-MALS elution fractogram obtained with 90° angle by Postnova. 490 491 At LNE, six replicates of each nanoPP vial were analysed on two different days to include daily 492 variability in the measurement precision. The results of the characterisation show excellent 493 repeatability, weight-average Rg,w of 86 nm ± 4 nm and a recovery of about 74% (Table 3). Similar 494 results were obtained at SMD, with slightly higher Rg,z and Rg,w and recoveries higher than 70% 495 (Table 3 and Electronic Supplementary Information, Section E). 496 Taking in account the different nature of the measurands, the associated uncertainties (k=1), and 497 the different model used for fitting MALS signal intensities, the values measured by the participants 498 using flow-FFF are in good agreement among them, with FFF theory (Rh,w of 103 ± 2 nm, Table 3) 499 and PTA, DLS and MADLS results (Tables 1 and 2). 500 Furthermore, at Hereon, four replicates of each nanoPP vial were analysed using cFFF coupled to 501 MALS (Table 3). Rg,z and Rg,w values obtained with this instrumental set-up were lower than values 502 reported for AF4 (see Electronic Supplementary Information, Section F, for details), but closer to 503 reported PTA values (Table 1). Although the explanation of these inter-instrumental and inter504 detector differences is analytically interesting, it was beyond the scope of this work. In general, MD505 FFF methods produced comparable size characterisation data consistent with PTA, DLS, and 506 MADLS results. It is also worth noting that FFF channel recoveries of >70% were achieved in all 507 laboratories and with all systems, meeting method acceptance criteria described in ISO 21362 [28]. 508 SEM-EDX was used to obtain size and size distribution data on nanoPP, as well as elemental 509 information. Representative SEM images shown in Figure 4 reveal that the nanoPP material 510 comprised particles with irregular shapes and a broad size distribution, ranging from ~50 nm to ~160 511 nm, and in agreement with size values reported by the numberand intensity-based analysis 512 techniques. The EDX spectrum of the imaged particle population in the nanoPP sample revealed 513 clearly visible peaks relating to the oxygen O Ka line (Figure 5). The observable S Ka signal 514 potentially comes from the isotactic form of the PP polymer, with cross linkage in the presence of 515 sulphur [29]. The copper peak Cu La and the carbon peak C Ka come from the substrate (carbon 516 and pioloform coated Cu grid) used to deposit the particles, whilst the Al Ka signal comes from the 517 sample holder. 518 519 15 520 Figure 4: Representative SEM images of particles present in the nano-PP sample with 521 magnification x10 000 for image on the left and magnification x100 000 for image on the right. 522 523 524 Figure 5: EDX spectrum performed on the selection of particles present in the nanoPP sample as 525 indicated on the SEM image (right) and EDX mapping in the same area (left). 526 527 STEM-EDX analysis of the nanoPP material, deposited by various methods, revealed the presence 528 of several chemical compounds (mostly salts and other inorganic compounds, see Electronic 529 Supplementary Information, Section F). Grids prepared by grid-on-drop deposition and using Alcian 530 blue grid staining resulted in the least amount of interference of various deposited compounds. From 531 such a grid, a size histogram of area-equivalent circular diameter (ECD) was constructed based on 532 the manual measurement of 74 particles, resulting in a mean ECD of 166 nm (Electronic 533 Supplementary Information, section F, Figure 1), consistent with the results from characterisation 534 techniques able to analyse the particles in suspension. The limited statistics reflect the low particle 535 concentration on the grid. While most particles were isolated, a few agglomerates were also 536 observed, in which case constituent particles that could be resolved visually were measured 537 individually. The composition of a subset of particles was verified by STEM-EDX to be sure that the 538 C signal, indicative of plastic composition, was dominating. However, since C is also present in the 539 16 supporting film, it cannot be said with certainty that the particles included in the size analysis are 540 nanoPP. 541 Electronic Supplementary Information, Section E, gives an overview on how to validate TEM for 542 reliable determination of particle size and shape based on the 200 nm PS QC material. It 543 demonstrates that accurate and precise measurements can be obtained for monomodal and 544 monodispersed nanoplastic particles, when the particle concentration is sufficiently high. In contrast, 545 attempts to measure the size of nanoPP demonstrates that detection and analysis of environmentally 546 relevant nanoplastics with TEM is much more challenging. Currently, there is no reliable method to 547 selectively separate and enrich nanoplastics with low abundancies to enable analysis in the same 548 way as QC materials. While analytical TEM, such as STEM-EDX, can be useful for detecting false549 positive plastic identifications through elemental composition analysis and thus serve as a 550 complementary technique to DLS and PTA, it cannot directly confirm the presence of plastics. 551 Analysis of the nanoPP material STEM-EDX at the University of Parma TEM produced results that 552 were in agreement with those described above (see Electronic Supplementary Information, section 553 G). 554 With (classical) AFM performed at SMD only mechanical measurement was possible. Reliable 555 dimensional measurement was achieved by using deposition method described in the Results 556 section, which yielded a sufficient amount of particles without visible agglomeration. Overall, the size 557 range measured with AFM was in agreement with the size reported by SEM and TEM (see Electronic 558 Supplementary Information, Section H for details). 559 560 3.2.2. Particle number concentration 561 Particle number concentration was measured with PTA by LGC and Postnova (using NS300 562 instruments) and UNITO (using a PMX-120 instrument), with the results presented as average 563 particle number concentrations with associated standard deviation (Table 4). In addition, LGC also 564 calculated the associated measurement uncertainty as being ~19% (k=2, calculated according to 565 Eurachem/CITAC guidelines), with repeatability in particle counting identified as the main 566 contributing factor (~80%). The nanoPP concentrations determined across the 3 instruments range 567 from 1.7310 ± 2.479 to 2.410 ± 2.69, which are in agreement, considering the associated measurement 568 uncertainty (k=2). 569 570 Table 4. Particle number concentration obtained with PTA. 571 Institute Particle number concentration (g-1) Average (g-1) Stdev (g-1) RSD (%) Postnova 1.73 E10 (n=24) 2.47 E9 (n=24) 14.3 LGC 2.04 E10 (n=45) 1.78 E9 (n=45) 8.7 UNITO 2.4 E10 (n=15) 2.6 E9 (n=15) 10.8 572 573 3.2.3. Chemical identification 574 In Figure 6, Raman spectra are shown for bulk PP (as reference; grey), nanoPP with DEP off (red, 575 background), and nanoPP with DEP on (blue). All nanoPP peak assignments are highlighted and 576 detailed in the Electronic Supplementary Information (Section I), alongside a comparison to the bulk 577 PP. The nanoPP spectrum exhibits most of the characteristic PP peaks; however, differences were 578 observed in peak positions, as well as in their relative intensities and widths. For example, the peak 579 (d) at 2840 cm−1, assigned to the symmetric stretching of CH2, shifted to 2848 cm−1, suggesting XXX. 580 17 In addition, two peaks assigned to the stretching of carbonyl groups (C = O) appear at 1650 cm−1 581 and 1605 cm−1, possibly formed due to oxidation of nanoPP during its preparation procedure, where 582 the PP granules were soaked in acetone and then dispersed. The fingerprint region also presents 583 significant variations: the peaks of the bulk PP at 1436 cm−1, 1254 cm−1, 1219 cm−1, 1168 cm−1, 1153 584 cm−1, and 940 cm−1 are shifted to 1445 cm−1, 1277 cm−1, 1202 cm−1, 1157 cm−1, 1127 cm−1, and 929 585 cm−1 in nanoPP, respectively. Also, the two evident peaks in bulk PP at 842 cm−1 and 810 cm−1 are 586 not present in the nanoPP, with the formation of an intermediate peak at 830 cm−1. Despite these 587 small differences between the spectra of the bulk PP and the nanoPP, the chemical identity of the 588 nanoPP material was confirmed as polypropylene, with evidence of aging/oxidation having occurred. 589 590 591 Figure 6: Raman spectra (from bottom to top) of bulk PP, nanoPP with DEP off (background), and nanoPP 592 with DEP on. 593 594 Correlative SEM/Raman analysis was also performed on a drop of the nanoPP suspension 595 deposited on a silicon wafer to assess the global chemistry. As chemical analysis of individual nano596 scale particles by Raman is not possible due to the limited size resolution of this technique in the 597 nanometer size regime, bulk analysis can be used to generate an average spectrum. The collected 598 spectra confirmed the material was PP and were in agreement with data obtained by INRiM (see 599 Electronic Supplementary Information, Section I for details). 600 601 3.2.4. Polypropylene mass fraction quantification 602 The concentration data for nanoPP determined by pyGC-MS are presented in Table 5. Laboratory 603 blanks showed a background ranging from 3-30% of the PP, which was not subtracted from the 604 reported concentrations. The calculated concentration depended on the selected markers. This was 605 due to the discrepancy in relative peak response of the individual peaks in the duplet and triplet 606 compounds between the PP reference material and the nanoPP. The racemic 2,4,6-trimethyl-1607 nonene and heterotactic 2,4,6,8-tetramethyl-1-undecene peaks were significantly higher in relative 608 abundance compared to the reference material (and to what is commonly observed in isotactic PP). 609 The data generated using these markers are therefore likely to overestimate the nanoPP 610 concentration in the samples. The concentration of nanoPP would then be in the range 3-5 µg mL-1, 611 depending on the applied marker. These differences in chemical composition between the PP 612 reference material and the nanoPP (produced from a different PP source material) highlights a 613 potential challenge with PP quantification by pyGC-MS. In environmental or human samples the type 614 of PP present cannot be known and maybe a mixture of different PP particles, meaning the use of a 615 18 single PP refence material for generating calibration curves could easily overor underestimate the 616 true PP concentration. 617 618 Table 5. Polypropylene mass fraction obtained with pyGC-MS. 619 Marker nanoPP (µg mL-1) Blank (µg mL-1) Average Stdev RSD (%) Average Stdev RSD (%) Sum 2,4,6-Trimethyl-1-nonene 5.0 1.1 22 0.71 0.02 2 Sum 2,4,6,8-Tetramethyl-1-undecene 4.3 1.0 24 0.66 0.01 2 2,4-Dimethylhept-1-ene 2.7 0.5 19 0.79 0.01 1 2,4,6-Trimethyl-1-nonene (meso) 3.5 0.7 20 0.75 0.01 1 2,4,6-Trimethyl-1-nonene (race) 9.5 2.1 22 0.73 0.01 1 2,4,6,8-Tetramethyl-1-undecene (iso) 2.8 0.6 23 0.74 0.01 1 2,4,6,8-Tetramethyl-1-undecene (het) 17 4 22 0.5 0.2 30 2,4,6,8-Tetramethyl-1-undecene (syn) 3.7 0.8 22 0.76 0.01 1 620 3.2.5. Determination of inorganic impurities 621 Inorganic impurities present in the nanoPP material (detected qualitatively using EDS) were 622 analysed using microwave assisted total ICP-MS. Results of the multielement characterisation 623 showed a variety of elements in the nanoPP suspension, with Si being the most abundant, followed 624 by Al, Mg, K and Ca, which supports elemental mapping obtained with SEM/EDS. Elements detected 625 with ICP-MS were distributed heterogeneously in the sample (as indicated by relatively high stdev 626 of the replicate measurements), likely due to adsorption of these elements to the surface of the 627 nanoPP particles. Hyphenated AF4-ICP-MS was attempted by Postnova to verify if the elements 628 identified were associated with PP particles, but the ICP-MS signal measured in the fractionated 629 peaks was too low to enable on-line quantification. 630 631 4. Conclusions 632 The work present here describes the multiparameter physicochemical characterisation of a novel, 633 more environmentally relevant nanoPP test material prepared using a top-down approach employing 634 mechanic fragmentation of PP pellets. The material was characterised for particle size, size 635 distribution, shape, number concentration, polypropylene mass fraction, chemical identity and the 636 content of inorganic impurities using a wide range of analytical methods. The nanoPP material was 637 found to be polydispersed in size, exhibited irregular particle morphologies, and presented signs of 638 oxidative aging. Despite the materials’ challenging characteristics, good agreement between the 639 employed techniques was achieved for the key measurands, paving a way to future development 640 and commercialisation of comprehensively characterised nanoPP Reference Materials that are more 641 representative of critical environmental and food particle-based plastic pollutants of increasing global 642 concern. 643 644 5. Authorship contributions 645 Aneta Sikora: Writing – original draft, review and editing, formal analysis, visualisation, investigation, 646 methodology and validation Dorota Bartczak: Conceptualisation, writing – review and editing, 647 supervision, methodology and validation; Heidi Goenaga-Infante: Writing – review and editing, 648 19 supervision. Marta Barbaresi: investigation, formal analysis, writing – review and editing. Francesca 649 Rossi: methodology, investigation, formal analysis. Maurizio Piergiovanni: investigation, writing – 650 review and editing. Monica Mattarozzi: writing – review and editing. Maria Careri: writing – review 651 and editing, supervision. Andy M. Booth: data validation, writing – review and editing, Lisbet 652 Sørensen: investigation, methodology, formal analysis, writing. Jeremie Parot: methodology, writing 653 – review and editing, Amaia Igartua: investigation, methodology, formal analysis, writing-review and 654 editing., Thierry Caebergs: investigation, methodology, formal analysis, writing – review and editing, 655 Anne-Sophie Piette: investigation, formal analysis, Roland Drexel: investigation, methodology, 656 formal analysis, validation, writing – review and editing, Florian Meier: validation, writing – review 657 and editing, supervision, Francesco Barbero: writing – review and editing, investigation, formal 658 analysis, visualisation, methodology and validation; Ivana Fenoglio: writing – supervision. Charlotte 659 Wouters: methodology, formal analysis, validation, writing, review and editing. Jan Mast: review and 660 editing, supervision. Enrica Alasonati: investigation, methodology, formal analysis, validation, writing 661 – review and editing; Marta Fadda: investigation, formal analysis, writing – review and editing; 662 Alessio Sacco: investigation, formal analysis, writing – review and editing; Andrea Mario Rossi: 663 supervision, funding; Andrea Mario Giovannozzi: Conceptualisation, writing – review and editing, 664 supervision, methodology, validation and funding. 665 All authors reviewed and approved the final version of manuscript. 666 There are no conflicts to declare. 667 668 6. Acknowledgements 669 - The project 21GRD07 PlasticTrace (Funder ID: https://doi.org/10.13039/100019599) has received 670 funding from the European Partnership on Metrology, co-financed from the European Union’s 671 Horizon Europe Research and Innovation Programme and by the Participating States. 672 - METROFOOD-IT project has received funding from the European Union -NextGenerationEU, 673 PNRR—Mission 4 “Education and Research” Component 2: from research to business, Investment 674 3.1: Fund for the realisation of an integrated system of research and innovation infrastructures— 675 IR0000033 (D.M. Prot. n.120 del 21/06/2022). 676 - Research and innovation network on food and nutrition Sustainability, Safety and Security–Working 677 ON Foods” (ONFOODS) project which received funding from the NRRP and the Mission 4 678 Component 2 Investment 1.3-Call for tender No. 341 of 15/03/2022 of the Italian Ministry of 679 University and Research funded by the European Union–NextGenerationEU. Award Number: 680 Project code PE0000003 681 - This publication has been funded by the Italian Ministry of University and Research (MUR) in the 682 framework of the continuing-nature project “NEXTGENERATION METROLOGY”, under the 683 allocation of the Ordinary Fund for research institutions (FOE) 2023 (Ministry Decree n. 789/2023). 684 - 685 7. References 686 [1] Ritchie H. ‘How much plastic waste ends up in the ocean?’ OurWorldinData.org 2018, 687 'https://ourworldindata.org/how-much-plastic-waste-ends-up-in-the-ocean' , last accessed 18/09/25. 688 [2] EU (2020) Directive 2020/2184 of the European Parliament and of the Council of 16 December 2020 on 689 the quality of water intended for human consumption (recast). Official J EU 23.12.2020.435 690 [3] Pilapitiya, P.N.T. and Ratnayake A.S. ’The world of plastic waste: A review’ Cleaner Materials, 2024, 11, 691 p.100220. 692 [4] Andrady, A.L., Barnes, P.W., Bornman, J.F., Gouin, T., Madronich, S., White, C.C., Zepp, R.G. and Jansen, 693 M.A. ‘Oxidation and fragmentation of plastics in a changing environment; from UV-radiation to biological 694 degradation’. Science of The Total Environment, 2024, 851, p.158022. 695 20 [5] ISO/TR 21960:2020 ‘Plastics—environmental aspects—state of knowledge and methodologies’ 696 [6] Commission Regulation 2023/2055 of 25 September 2023 amending Annex XVII to Regulation (EC) n. 697 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, 698 Authorisation and Restriction of Chemicals (REACH) as regards synthetic polymer microparticles. Official 699 Journal of the European Union; 2023. 700 [7] Gigault, J., Ter Halle, A., Baudrimont, M., Pascal, P.Y., Gauffre, F., Phi, T.L., El Hadri, H., Grassl, B. and 701 Reynaud, S. ‘Current opinion: What is a nanoplastic?’ Environmental pollution, 2018, 235, p.1030-1034. 702 [8] Wagner, M., Hartmann, N.B., Verschoor, A., Hüffer, T., Hassellöv, M. and Thompson, R.C. ‘Are we speaking 703 the same language? Towards a definition and categorization framework for environmental plastic debris’ 704 Society of Environmental Toxicology and Chemistry, SETAC Europe 28th Annual Meeting, 2018, p.35. 705 [9] Alqahtani S., Alqahtani S., Saquib Q., and Mohiddin F. ‘Toxicological impact of microplastics and 706 nanoplastics on humans: understanding the mechanistic aspect of the interaction’ Front Toxicol. 2023, 5, p. 707 1193386. 708 [10] Coones R. T., Kestens V. and Minelli C. ‘A comparison of hydrodynamic diameter results from MADLS 709 and DLS measurements for nanoparticle reference materials.’ Journal of Nanoparticle Research 2025, 27, p. 710 170. 711 [11] Zhao, H., Wang, X., Wang, R., Hua, D., Li, K. and Ji, F., Optimal static light scattering detection angle for 712 particulate matter size and concentration measurement. Measurement Science and Technology, 2023, 34, 713 p.125802. 714 [12] Gallego-Urrea J. A., Tuoriniemi J. and Hassellöv M. ‘Applications of particle-tracking analysis to the 715 determination of size distributions and concentrations of nanoparticles in environmental, biological and food 716 samples’ TrAC Trends in Analytical Chemistry, 2011, 30, p. 473. 717 [13] Wouters C., Morales D., Tsilikas K., Verleysen E. and Mast J. ‘A validation methodology for size and 718 shape measurement of nanoplastics by transmission electron microscopy’ In BIO Web of Conferences, EDP 719 Sciences, 2024,129, p. 26039. 720 [14] Stine J. S., Aziere N., Harper B. J., and Harper S. L. ‘ A novel approach for identifying nanoplastics by 721 assessing deformation behavior with scanning electron microscopy’ Micromachines, 2023, 14, p.1903. 722 [15] Galluzzi M., Lancia M., Zheng C., Re V., Castelvetro V., Guo S. and Viaroli S. ‘Atomic Force Microscopy 723 (AFM) nanomechanical characterization of micro-and nanoplastics to support environmental investigations in 724 groundwater’ Emerging Contaminants 2025, 11,p. 100478. 725 [16] Huber M. J., Ivleva N.P, Booth A.M., Beer I., Bianchi I., Drexel R., Geiss O. et al. ‘Physicochemical 726 characterization and quantification of nanoplastics: applicability, limitations and complementarity of batch and 727 fractionation methods’ Analytical and bioanalytical chemistry 2023, 415, p. 3007. 728 [17] Megha S., Unnimaya S., Mithun N., Chidangil S., Kumar S. and Lukose J. ‘Raman Spectroscopy Based 729 Approaches for Microplastics Investigations’ In Microplastics in African and Asian Environments: The 730 Influencers, Challenges, and Solutions, Cham: Springer Nature Switzerland, 2024, p. 647. 731 [18] Gang Li., Yang Z., Pei Z., Li Y., Yang R., Liang Y., Zhang Q. and Jiang G. ‘Single-particle analysis of 732 micro/nanoplastics by SEM-Raman technique’ Talanta 2022, 249, p.123701. 733 [19] Fadda M., Sacco A., Altmann K. et al. ‘Tracking nanoplastics in drinking water: a new frontier with the 734 combination of dielectrophoresis and Raman spectroscopy’ Micropl .& Nanopl. 2024, 5, p. 24. 735 [20] Placci A., Fadda M., Coralli I., Wang J., Zattoni A., Costa A.L., Portela R., Giovannozzi A.M., Fabbri D., 736 Melucci D. and Giordani S.’Multitechnique characterization of eco-corona formation on airborne nanoplastics’ 737 RSC advances. 2025, 28, p. 30849. 738 [21] Mast J. et al., NANoREG D2.10 SOP 01 Preparation of EM-grids containing a representative sample of a 739 dispersed NM, 2018. 740 [22] Wagner, T. ParticleSizer. https://zenodo.org/badge/latestdoi/18649/thorstenwagner/ij-particlesizer, last 741 accessed on 18/09/2025. 742 21 [23] Verleysen E., Wagner T., Lipinski H.G., Kägi R., Koeber R., Boix-Sanfeliu A., De Temmerman P.J. and 743 Mast J. ‘Evaluation of a TEM based Approach for Size Measurement of Particulate (Nano) materials’ Materials 744 2019, 12, p.2274. 745 [24] Piergiovanni M., Mattarozzi M., Verleysen E., Siciliani L., Suman M., Bianchi F., Mast J., Careri M. The 746 Combined ICP-MS, ESEM-EDX, and HAADF-STEM-EDX Approach for the Assessment of Metal Sub-Micro747 and Nanoparticles in Wheat Grain. Molecules 2024, 29, 3148. 748 [25] Riboni N., Ribezzi E., Nasi L., Mattarozzi M., Piergiovanni M., Masino M., Bianchi F., Careri, M. 749 Characterization of Small Micro-and Nanoparticles in Antarctic Snow by Electron Microscopy and Raman 750 Micro-Spectroscopy. Appl. Sci. 2024, 14, 1597. 751 [26] Foucher, J., Labrosse, A., Dervill´e, A., Zimmermann, Y., Bernard, G., Martinez, S., Gr¨onqvist, H., 752 Baderot, J., Pinzan, F. ‘The Coming of Age of the First Hybrid Metrology Software Platform Dedicated to 753 Nanotechnologies (Conference Presentation)’ https://www.spiedigitallibrary.org/co nference-proceedings-of754 spie/10145/1014507/The-coming-of-age-of-the-first-hy brid-metrology755 software/10.1117/12.2258093.short?SSO=1, last accessed 18/09/25. 756 [27] Bouzakher-Ghomrasni N., Taché O., Leroy J., Feltin N., Testard F., Chivas-Joly C. ‘Dimensional 757 measurement of TiO2 (Nano) particles by SAXS and SEM in powder form’ Talanta 2021, 234, p.122619. 758 [28] ISO/TS 21362:2018 ‘Nanotechnologies-Analysis of nano-Objects using asymmetrical-flow and centrifugal 759 field flow fractionation’. 760 [29] PCT/DZ02/00001, WO 02/085973 A1, 2002, 761 https://patentimages.storage.googleapis.com/95/a4/96/b3961c6f59549d/WO2002085973A1.pdf, last 762 accessed 18/09/2025. 763 764