4.1. Evaluation of the new RMP particles: We used injection molded plastic polymer blocks of five different polymers to manufacture columns of a defined size using CNC milling. Subsequent gelatin embedment and cryosectioning resulted in gelatin slices with a defined number of RMP particles. By monitoring the production throughout the entire process, we observed that during CNC milling and further processing rarely any of the columns broke. However, in case of broken columns, respective column plates could still be used to produce RMP particles. Only the efficiency in terms of the number of particles obtained is lowered by each broken column and the number of RMP particles per finished gelatin slice should be recorded accurately using a stereo microscope. Similarly, in terms of quality assurance a final check of the number of RMP particles in the embedded gelatin slice intended for usage can be done very fast by microscopy and photo documentation ensures traceability and highest reliability.
For proving the concept of cutting polymer column plates using cryomicrotomy, we tested the cutting process for the column sizes 1000x1000, 500x500, 250x250 and 125x125 µm x-y-edge length. Our results show that using our approach RMP particles down to an x-y-size of 125x125 µm could be produced with low size deviations. Furthermore, we tested to produce RMP particles in even smaller sizes down to 50 µm. However, the number of defect columns during production increased with decreasing size classes below 125 µm. For smaller RMPs more methodological adjustment is needed, and we decided to not report results here. Nevertheless, we are aware that there is even better CNC milling equipment than the standard CNC mill we used. CNC milling can reach accuracies of up to 1 µm23 when specialized CNC-mills are used, which potentially allows for the production of even smaller RMPs in high precision. However, the lower threshold in terms of particle size realizable with such specialized equipment has to be determined in future.
Our results show little standard deviations for the produced sizes throughout all manufactured RMP particles for all three dimensions (x, y, z). However, LD-PE-RMP particles tended to have protruding chippings and round edges. It is likely that this is due to wear of the material during milling and cutting23,24. This in turn is likely due to the lower tensile strength of LD-PE compared to PP, PET, PS and PLA (Table 1)25, where not such chippings have been observed. Tensile strength has an impact during the milling process23,24. With increasing tensile strength the shear strength increases as well and thus also the cutting force, which on the one hand results in good RMP milling results, however on the other hand enhances the stress to the milling tools and potentially leads to surface changes of the workpiece23. As we observed shape deviations and chippings only for lower tensile strength LD-PE with an average of 10.8 MPa25 we assume that polymers with lower tensile strengths are harder to process for RMP particle production via CNC-milling than polymers with higher tensile strengths. On the other hand, we expect polymers like Polymethyl methacrylate (PMMA) with a tensile strength between 30 and 50 MPa and an average of 48 MPa25 (Table 1) to be easily processable since high tensile strength PET with an average of 50 MPa for instance was processed successfully. Tire rubber with a lower tensile strength of < 18 and an average of 14 MPa (Table 1) would be harder to process by CNC milling under ambient temperature conditions25. A solution to this problem could be working under temperature conditions below the glass transition temperature of such low tensile strength polymers (Table 1). Cryogenic CNC-milling is possible and could potentially be used in such cases26. There the milling process is cooled with − 196°C liquid nitrogen applied through a nozzle directly to the milling tool and the work piece26. This could also improve the milling results for LD-PE with a glass transition temperature of around − 115°C27 (Table 1) and similar polymers.
Table 1
Average tensile strengths and glass transition temperatures for the 5 polymer types we processed in this study (LD-PE, PP, PET, PS, PLA) and for the examples PMMA and tire rubber. The specified data are from matweb25 and Greene (2021)27
Polymer | Average tensile strength | Average glass transition temperature |
LD-PE | 10.8 MPa | -115°C |
Tire rubber | 14.0 MPa | 20°C |
PP | 33.1 MPa | -5°C |
PS | 41.0 MPa | 100°C |
PMMA | 48.0 MPa | 105°C |
PET | 50.0 MPa | 75°C |
PLA | 59.5 MPa | 60°C |
Most polymer column plates embedded in gelatin could easily be cut during cryo-microtomy. However, problems occasionally occurred when cutting 1000x1000 µm x-y-edge length PLA columns. This is most probably due to material characteristics of PLA (Table 1). It was sometimes difficult to cut through the entire embedded column block, without having errors like skipping of the blade. Using a fresh blade solved this issue but only for a certain number of sections. As we did not face this problem for smaller RMP particle sizes, this is most probably related to the larger area that has to be cut per column when producing 1000x1000 µm x-y-edge length columns in combination with the tensile strength of the comparably hard polymer PLA. Thus, for production of larger RMP particles made of harder polymer types we suggest processing smaller blocks of embedded column plates (x-y-wise) with a smaller number of columns so that less force is needed to cut through the block. Furthermore, fresh blades should always be used for each column plate when critical polymers have to be cut. We could frequently observe grooves in the cut RMP particles. These occurred regardless of whether a blade was used for the first time or already for other blocks. Therefore, we assume that this phenomenon cannot be prevented by fresh blades.
To produce the RMP particles we always used cleaned equipment as well as filtered solutions and conducted the sample preparation steps under laminar flow boxes to prevent contamination. However, for proving the concept of producing our RMP particles using our procedure, we didn’t filter the gelatine used, since it was of extra-pure grade, and we removed it after production of the RMPs for the characterisation and analysis of the RMP particles. Yet, for future applications of the RMP particles also the gelatine must be filtered to remove potential MP contamination that can potentially be present in the gelatine powder used, even if it was of extra-pure grade.
4.2. Towards quality assessment, quality control and harmonization in terms of comparability of methods:
During the recent years it became more and more evident that comparability of MP data in different studies is not comprehensive and unsatisfying, thus in order to solve this issue, standardized methods following strict QA/QC procedures and SOPs have to be set in MP research28–39. Research projects like ‘BASEMAN’, ‘EPAL-QUASIMEME/NORMANs first global interlaboratory study on MPs, or ‘EuroQcharme’ tried to harmonize sampling, sample processing and methods of identification for the sake of comparability28–30. In these projects, all relevant forms of MP particles were produced (Pellets, Fragments, Fibers, Films, Spheres) and addressed28–30. However, in general RMP particles beyond microbeads or powders are commercially not available at present18.
Suspension polymerization or better controllable emulsion polymerization can be used to create spherical RMP particles16,35,36. PS is the most used polymer used for producing such RMP beads16,35,36. However, there are also attempts to produce spherical microplastic particles using other polymer types like PE16. The basis of emulsion polymerization is that the respective polymer is dissolved at higher temperature in a fluid, dispersed via ultrasound and then cooled with the addition of a surfactant. This way the respective polymer forms spheres. The residual solutions are filtered to obtain a powder. However, these can only be applied gram wise or as suspension in experiments16,35,36.
In terms of MP fibers, Cole (2016) presented a method to prepare RMP fibers in a soluble matrix using a microtome. These fibers can be produced with a defined size (10–28 µm in diameter and 40–100 µm in length), shape and polymer type31.
For producing fragmentous shaped RMP particles, cryomilling is currently the mostly used technique10–12,32−34. The disadvantage of cryomilling is that the particles produced always show a more or less broad particle size distribution. Thus, MP fragments produced by cryogenic milling (cryomilling) are usually subsequently sieved to create RMP particles of a certain size fraction, as in the project ‘BASEMAN’28.
Generally, the addition of an exact number of bead-, fiber-shaped or cryomilled, fragmentous RMP particles to an experiment is difficult. Manual counting and addition are not a practical solution as explained before. Thus, often a defined weight is added to a sample28–30, with uncertainty of the exact particle number added. Another method for adding RMP particles to experiments is to use RMP suspensions with a defined weight-concentration of RMP particles32. However, it is not possible to achieve completely homogeneous distributions of RMP particles in a suspension. This leads to the possibility that the amounts of particles added to an experiment is different from the intended amount or differs strongly between replicates of the same experimental treatment during, e.g., ecotoxicological studies. This in turn can lead to insignificance of results. Furthermore, losses of RMPs leading to increasing variability can also occur when handling suspensions because particles can stick to the equipment, e.g., pipette tips, due to static charge. Thus, other studies like Gábor, B. (2022) or ‘EPAL-QUASIMEME/NORMANs’ used MP-containing soda tablets that were produced of a mixture of soda salt and weight-based addition of MP particles29,30. However, preparation of these tablets brings the same issues as described above. With these approaches the evaluation of number-based analytical methods like Raman or FTIR spectroscopy during the interlaboratory study ‘BASEMAN’ proved difficult and consequently the harmonization in terms of comparability of the results of number and mass-based analysis techniques was difficult as well28.
For the production of a defined number of RMPs, Bamshad and Cho (2021) proposed a procedure where they used a laser jet printer13 for printing particles. These RMP particles only showed a 5.2% deviation in the x-y-sizes which is lower than the average x-y-size deviation of 9.67% we measured across all size classes and polymer types of our RMP particles. However, the particles that were produced by laser jet printing were black, made of PS and contained multiple additives to make the polymer printable13. Thus, the original polymer properties are altered strongly. Next to potential different behavior and effects in ecotoxicological experiments compared to original polymers, chemical differences of these printed RMP particles might complicate identification via µFTIR and Raman spectroscopy.
As shown above, all mentioned methods have their drawbacks. Either the difficulty in handling of the references, the unknown size distribution or the characterization of the particles is difficult. Thus, there is still need for RMP particles of a defined chemical composition / polymer type, size and shape that can be handled easily and added in a defined number.
Our new RMP particles can contribute to solve most of these issues since they allow for the addition of an exact number of RMP to a sample as internal standard for the calculation of recovery rates after sample preparation without manual counting or use of suspensions. This in turn, for the first time, allows for a number-based validation of sample transfer processes, extraction, purification, subsampling etc. of all common MP analysis methods with all relevant polymer types. Here the suitability of RMPs includes analysis of environmental samples of virtually all matrices since the defined shaping of our RMP particles ensures differentiability between MP particles occurring in environmental samples. Because of their defined shape and weight with low standard deviation, our RMP particles also allow an assessment of the mass per number of RMP particles. This in turn allows for the use of our RMPs as internal standard for evaluation and quality control in mass-based analysis techniques and thus comparability between number-based techniques like FPA-µFTIR or Raman and mass-based analysis techniques like TED- or pyrolysis-GS/MS.
4.3 New RMP particles and possibilities beyond evaluation of analytical methods
Next to evaluation of environmental contamination, incomparability of data also influences important MP research fields such as ecotoxicological effects or environmental transport studies. For ecotoxicological effect studies, it is a prerequisite to use environmentally relevant concentrations of MP particles to ensure that the results are comparable to environmental conditions39. So far MP concentrations exceeding contamination levels in the environment are frequently used to set up toxicological experiments potentially resulting in artefacts 39. The same is true for concentrations below contamination levels where no effects are observed39. In this context our RMP particles can be added as internal standards to all kinds of samples for evaluation of analytical methods for establishing a reliable data basis on the environmental contamination. This creates a basis for ecotoxicological effect studies under realistic MP concentration levels.
Next to issues with environmentally relevant concentrations of MP particles in laboratory studies, MP particles in the environment do not share the same traits as well1,16,17,39,40. These are shape, size, surface charge, density, polymer type, additives, residual monomers and environmentally mediated properties like surface coating with an ecocorona, surface fouling by a biofilm, ageing and degradation1,16,17,39,40,41. These factors lead to potential differences of ecotoxicological effects and also differences in the distribution of MP particles in the enviroment1,4,5,39,42. To enhance quality of MP effect studies it is thus also important, to use RMP particles that help to disentangle the effects of the different particle properties. Therefore, accurate physical/chemical characterization of MP particles is of utmost importance8,9,36–38. However, there is a lack of in-depth RMP particle characterization in MP ecotoxicological research. Despite the heterogeneity of MPs found in environmental samples laboratory experiments so far have often mainly been conducted using PS beads/spheres34,39,42. In this context our newly developed RMP particles can be produced using various polymer types, with/without additives or residual monomers, in different sizes and potentially shapes etc. However, it is conceivable that freshly manufactured RMP particles may show different characteristics than aged ones. To solve this issue artificial aging can be applied to our RMP particles. Other chemical modifications can be applied as well. Additionally, the polymer melt used for creating the polymer blocks we use for CNC-milling before injection molding can be spiked with e.g., fluorescent dye or other markers if needed for certain experiments.