Introduction

Mechanical recycling of post-consumer plastics is a possibility to contribute to the transformation of the linear economy to a circular economy. Considering the worldwide plastic production of 367 Mt [1] in 2021, it is important to increase recycling capabilities and try to recycle as many plastic types as possible. A reasonable association when thinking about plastics is the occurrence of microplastics (MP). MP are found in water bodies, soil, and the air [2,3,4]. The influence of MP on the environment and human has been investigated [5, 6] as well as possible sources of MP. To the latter, a variety of sources are identified such as tire abrasion, cosmetic industry, spills, etc. [1].

However, MP emission originated during post-consumer plastic recycling of polypropylene packaging waste has not been sufficiently investigated. Suzuki et al. [7] identified 5.80 and 0.13 tmicroplastics/year during the mechanical recycling of electronic plastic waste and polyethylene terephthalate bottle waste, respectively.

During the mechanical recycling of plastics, the collected post-consumer waste needs to be pre-treated before the re-granulation step. One pre-treatment step is washing, which typically is applied after sorting at object level and the subsequent shredding. Washing of post-consumer waste is necessary to remove dirt and other residues on the plastics, which possibly interfere with downstream processing and to increase the product quality [8, 9]. Washing of different plastics can be conducted at different operational parameters. For instance, washing temperature can vary between room temperature and 80 °C [10,11,12]. Another important parameter in washing is residence time, which is always a subject of discussion, as an optimum balance should be sought between cleaning performance and economical aspects.

Therefore, this study aims to close the knowledge gap by investigating the potential occurrence of MP during the pre-treatment (washing) of polypropylene. For this purpose, the washing process was simulated with an industrial washing machine in order to be able to investigate the effects of the operational parameters. The hypothesis underlying this study is that longer residence times have a stronger influence on the formation of microplastics than higher temperatures.

Results and discussion

In Fig. 1, the filter residues are depicted in mgMP/gfeedstock. Looking at a residence time of 5 min, no clear differentiation between the filter residue at different temperatures is possible with 0.048 mgMP/gfeedstock at 30 °C and 0.053 mgMP/gfeedstock at 60 °C with a standard deviation of 0.010 and 0.017, respectively. A similar behavior can be seen when looking at a residence time of 15 min. Here no significant differences between the 30 °C and 60 °C washing trials are identifiable. However, the standard deviation of the 30 °C trial is considerable higher than the standard deviation of the 60 °C washing trial, with 0.036 and 0.007, respectively. The higher standard deviation might originate in the possibility in different flake geometries. In shredding of post-consumer plastics waste, a certain flake size distribution occurs. Smaller particles are not homogeneously distributed and hence behave differently in terms of abrasion. Another possible reason for the variation could be residues of MP in the washing machine, even though it was rinsed and cleaned before each trial. A washing machine is no precise characterization instrument with a number of joints and edges that could be the reason for material carryover. Comparing the impact of different resistance times, a significant difference between the residue generation at 5 min and 15 min is depicted. Approximately two-fold higher mass of residue is ascertained when washing 10 min longer.

Fig. 1
figure 1

Filter residue after washing of PP at different temperatures (30 °C and 60 °C) and residence times (5 min and 15 min)

Full size image

The DSC thermograms in Fig. 2 show the curves of (a) polypropylene (PP) flake, (b) filter residue from a washing reference (empty washing machine), and c) filter residue from washing PP sample. The flake is showing a melting point at around 160 °C, which is consistent with PP values in literature [13,14,15]. For both filter analysis, the melting point occurs also at around 160 °C, which indicates the presence of PP. This leads to the conclusion that when analyzing the washing water of an empty washing machine PP particles can be found. A possible reason could be that there are residues from the washing machine from previous washing cycles, which confirms the conclusion made for the standard deviations. However, these measurements are only conducted to qualitatively show the occurrence of PP in the samples. The significant difference in peak size under stable and consistent operational parameters with the only difference that during the sample trials PP is washed, indicates the occurrence of MP originating from the washed PP flakes. These results suggest a release of MP during the pre-treatment (washing) of PP and a release of MP during mechanical recycling of plastic waste was also shown in Suzuki et al. [7].

Fig. 2
figure 2

DSC thermograms of a PP flake, b filter residue from a washing reference (empty washing machine), and c filter residue from washing a PP sample

Full size image

Conclusion

Washing PP flakes in an industrial washing machine to simulate a pre-treatment step during the mechanical recycling of plastics reveals MP emissions. Results suggest a higher influence on MP emission due to the residence time in the washer rather than temperature. An approximately two-fold higher weight was shown when residence time was increased from 5 to 15 min.

Experimental

To test our stated hypothesis, an industrial washing machine was used, which allows the simulation of the impact of residence time on MP generation, by varying the time spans. In addition, it was also possible to adjust the temperature. Considering an industrial washing process which differs from the used washing machine, it was reasonable to use it for testing the impact of friction between flakes and surroundings at various time spans and temperatures.

For the trials, around 0.4 kg of pre-treated (washed) and dry polypropylene flakes (> 4 mm) were washed in a nylon washing bag in an industrial washing machine (Miele Professional PW818) with only water to simulate a washing process in plastics recycling. These flakes were washed 3 times before the washing effluent was analyzed to remove possible plastic residue and other debris on the flakes. As a reference, the washing water of the empty washing machine was used. For all measurements, triplicates were conducted. The reference samples show the residues originating from the washing machine, piping, etc. during all trials. Filter residues from washing PP samples show a combination of the reference and MP residues from the PP flakes. Hence, it is assumed when subtracting the average reference from the average PP sample the potential MP release from the flakes is indicated. After this subtraction, the results are presented on a relative basis, considering the weighing of the PP flakes.

For the washing trial, two different temperatures (30 °C and 60 °C) and two different residence times (5 min and 15 min) were tested. The rotating movement of the washing drum was set to 30 rpm and started to rotate clockwise for 58 s followed by a 2 s break and then started to rotate counterclockwise for another 58 s.

The whole washing effluent (around 12 dm3) was stepwise vacuum-filtered using a 500 µm and a 45 µm woven wire mesh (Retsch), whereby the residue on the 45 µm mesh was used for further analysis. Before transferring the residue into a Petri dish, the mesh was placed for 5 min in an ultrasonic bath to loosen all particles which potentially could adhere to the surface based on the applied vacuum. Samples were then dried at 50 °C in a drying oven (Heraeus Type T50 0.7) and subsequently the dry weight of transferred particles was determined using a scale (Satorius QUINTIX224) to quantify potential MP emissions (solids). Quantified samples were then qualitatively analyzed using the differential scanning calorimeter (DSC) DSC PerkinElmer 4000, which can be used for the identification of polymers [16, 17]. The sample is heated from 30 °C to 200 °C at a rate of 10 °C min−1 under nitrogen atmosphere. The melting temperature of the resulting DSC thermograms was used to qualitatively identify the prevailing polymer.