Processability of Different Polymer Fractions Recovered from Mixed Wastes and Determination of Material Properties for Recycling

To achieve future recycling targets and CO2 and waste reduction, the transfer of plastic contained in mixed waste from thermal recovery to mechanical recycling is a promising option. This requires extensive knowledge of the necessary processing depth of mixed wastes to enrich plastics and their processability in polymer processing machines. Also, the selection of a suitable processing method and product application area requires appropriate material behaviour. This paper investigates these aspects for a commercial processed, mixed waste, and two different mixed polyolefin fractions. The wastes are processed at different depths (e.g., washed/not washed, sorted into polyethylene, polypropylene, polyethylene terephthalate, polystyrene/unsorted) and then either homogenised in the extruder in advance or processed heterogeneously in the compression moulding process into plates. The produced recyclates in plate form are then subjected to mechanical, thermal, and rheological characterisation. Most investigated materials could be processed with simple compression moulding. The results show that an upstream washing process improves the achievable material properties, but homogenisation does not necessarily lead to an improvement. It was also found that a higher treatment depth (recovery of plastic types) is not necessary. The investigations show that plastic waste recovery with simple treatment from mixed, contaminated wastes into at least downcycling products is possible.


Introduction
The waste management industry often talks about "plastics" as if it were a single material, but this is not the case. Plastics are an extensive family of entirely different materials. Each plastic type is designed with specific characteristics that make it ideal for its intended application. Whatever their application was, at the end of their service life, plastic materials are necessary resources that should first be recycled (upstream/downstream), and only when this is no longer technically possible and economically feasible, they should be used as an alternative energy source in energy recovery facilities [1][2][3].
In 2019, 368 million tonnes (Mt) of plastic were produced worldwide-57. 9 Mt in Europe (EU28+NO/CH) [1], and the EU converters' demand was about 50.7 Mt. "Packaging" (~40%) and "building and construction" (~20%) represent the largest end-use markets terisation included have been widely studied and validated both at the industrial level and the research stage.

Materials and Sampling
A common mixed waste, i.e., SRF (approx. 200 kg) produced from pre-treated and untreated mixed wastes (mixed municipal waste, commercial waste, etc., excluding separately collected wastes such as lightweight packaging waste) from a production facility near the city Graz in Austria, was used for the studies. In spring 2018, the sample was taken from the falling material stream according to ÖNORM S 2123-3 [11]. The material sample was taken from the SRF processing line after pre-shredding (<500 mm), magnetic separation, and PVC separation using a NIR (near-infrared) sorter and had a particle size >100 mm.
At the same time, about 20 kg of a sample of a PO-rich waste fraction (sample name: PO_A) was taken at the same plant. According to ÖNORM S 2123-3 [11], the sample was taken from the falling material stream. The sampling is carried out after pre-shredding, magnetic separation, PVC discharge, separation of heavy materials by a wind sifter, and subsequent post-shredding (<35 mm).
A further PO sample (approx. 8 kg) (sample name: PO_B) was taken out from the wet-mechanical processing unit using a centrifugal force separator [12] with a particle size <30 mm. The input material was a mixture of common SRF, mixed plastics from light-weight packaging treatment, and mixed plastics from the industry.
All samples mentioned are representative samples composed of individual increments taken continuously over several hours during the plant operation or test run. Exemplary photos of the three test materials are shown in Figure A2 in Appendix I.

Methods
In this section, the experimental and analytical procedures are described. Figure 1 (I. Plant set up for investigations) shows the modular plant configuration for material preparation, as it could also look like in real processing plants. The investigation method is divided into three areas: A.) Mechanical pre-processing of the input materials; B.) Polymer processing consisting of material homogenisation, a compression moulding process, and test specimen preparation; and C.) Material characterisation with thermal, mechanical, and rheological material testing.
The mechanical pre-processing consists of a drum screen for the separation of fine material (<20 mm), a manual sorting station for the removal of non-plastics and other materials, a double shaft pre-shredder to reduce the average particle size of plastics below 100 mm, a cold washing aggregate (a self-built stirred washer), a thermal drying cabinet (drying at 105 • C up to constant weight according to ONR CEN/TS 15414-1), a sensor-based sorting system (near-infrared) for the manual sorting of the standard plastic types (PE, PP, PET, and PS), and a post-shredder (cutting mill) to reduce the particle size to <4 mm.
In the polymer processing step, one-half of the shredded plastic flakes per plastic type were fed to a counter-rotating parallel twin-screw extruder TSE 42/7D (screw diameter (D): 42 mm; screw length: 7D; model no.: 8324; type: Plasti-Corder PL2000 from Brabender ® GmbH & Co. KG, Duisburg, Germany) with a three-zone screw for thermoplastics. This equipment is used for material homogenisation, e.g., thermoplastic multicomponent systems, polymer blends, or composite materials. The product (filament) was cooled in a water bath and granulated afterwards. The other half of the materials were directly processed into plates (dimensions: 160 mm × 160 mm × 4 mm) with a hot vacuum compression moulding process (vacuum press type P200PV, Dr. Collin GmbH, Maitenbeth, Germany). The material-specific four-zone temperature profile for material homogenisation is given in Table A3 in Appendix C. All materials were homogenised at a screw speed of 110 rpm. The material-specific, five-stage press profiles (temperature, pressure, and time) were determined empirically. The press profiles are presented in Table A4 in Appendix C. Test specimens used for material characterisation were stamped or cut (CNC milling machine) from the plates. Extensive tests were carried out for this purpose: thermal characterisation with differential scanning calorimetry (DSC), melt mass flow rate (MFR), determination of ash content, mechanical characterisation with tensile tests, (notched) impact strength, and determination of the bulk density from plastic flakes after shredding and granulates after homogenisation.
Crystallinity (X C ) is calculated from the melting enthalpy (∆H m ) measured with DSC and the approximated melting enthalpy of totally crystalline material (∆H 0 ) from the literature according to Equation (1) [13,14].
Figure 1 (II. Material flow "waste to recycling material") shows all material flows generated during the investigations from waste to finished test specimens.
The mixed waste, i.e., SRF, was the only input material screened and manually sorted into six material fractions (wood, paper, and cardboard (P&C), plastics, inert, metals, and other materials). Exemplary photos of the manually sorted material fractions are shown in Appendix I in Figure A3. Only the plastic fraction was further processed and investigated. The other fractions were discarded and, therefore, not relevant for further investigation. The plastic fraction was divided into three similar parts. One part was dried and shredded without NIR sorting. The second part was dried and sorted by NIR into five plastic types and the rest. Exemplary photos of plastic types sorted with NIR are given in Appendix I in Figure A4. The third SRF part was washed, dried, and NIR-sorted. The sorted plastic types were post-shredded separately.
The two PO materials were not screened and not manually sorted due to their small grain size (i.e., <35 mm and <30 mm).
The input material PO_A was divided into two similar parts. One part was washed, dried, and shredded. The other part was dried and shredded without a washing step. As the input material, PO_B came from wet-mechanical processing, it was only dried and shredded.
All material flows were divided after shredding. Half of each material went into the extruder for homogenisation, and the other half was compression moulded directly into test plates without homogenisation.

C.) Material Characterisation
For thermal characterisation of the materials, DSC measurements were performed using a DSC1 (Mettler-Toledo GmbH, Urdorf, Switzerland) in a temperature range from 0 to 230 • C for PE, PP, and PS materials, and from 0 to 200 • C for PO and P materials with a heating rate of 10 K/min in a nitrogen atmosphere (nitrogen flux rate 50 mL/min). The cooling rate was 20 K/min. To make the thermal history the same for all materials, a measuring program with one heating, one cooling, and second heating was chosen. Only the cooling and second heating curves were used for analysis. In advance, for checking the thermal stability, measurements up to 300 • C with a heating rate of 20 K/min in a nitrogen atmosphere were carried out for each material. This was used to determine the range of measurement itself. Seven reproducibility measurements for the heterogeneous and three for the homogeneous materials were carried out according to DIN EN ISO 11357-1 [15]. Standard 40 µL aluminum crucibles with pierced lids were used.
Charpy impact tests and notched impact tests (Ceast Resil 25, INSTRON/Ceast, Pianezza, Italy) according to DIN EN ISO179-1 [16] were performed at room temperature using a pendulum with 2 J (unnotched) and 0.5 J (notched) for P_PE, P_W,PE, P_W,C,PE, P_W,PP, P_W,C,PP, PO_A,W, and PO_B,C; a pendulum with 0.5 J (notched and unnotched) for P_PP, P_C,PP, all PS materials, PO_A, PO_A,C, PO_A,W,C, PO_B, and P; and a 7.5 J pendulum (unnotched) for P_W,C,PE. Tensile tests (Zwick Z010, Zwick/Roell GmbH & Co. KG, Ulm, Germany) were performed at room temperature according to DIN EN ISO 527-1 [17] and EN ISO 527-2 [18]. The ash content was determined according to DIN EN ISO 3451-1 [19]. Due to the heterogeneity, three reproducibility measurements were carried out, and the mean values were calculated for the discussion. The bulk density was determined for the plastic flakes after shredding and for the granulates after compounding, respectively, according to DIN EN ISO 60 [20]. Five measurements per material were carried out. The MFR (Modular Melt (Mass) Flow Tester, INTERON/Ceast, I) was determined according to DIN EN ISO 1133-1 [21]. The test conditions were set to 190 • C and 2.16 kg for all materials except for the PS materials. For PS, the test conditions were set to 200 • C and 5 kg.

Results
For the examined plastic materials, the following properties were investigated: The composition of mixed waste, i.e., SRF and its plastic type content, thermal and mechanical properties, characterisation of the flow behaviour (MFR), bulk density, and ash content.
All stated values are wt.% DS (DS-dry substance), given in full percent only for clarity purposes.
All PE, PP, PO, and P materials could be processed without any major problems. The PS materials emitted much gas in both processing variants, and several test runs were necessary to find a stable processing method. The PET materials could neither be homogenised nor compression moulded due to excessive contamination. Possible impurities are multilayer bottles, residual label material (PO), different non-compatible PET grades [10], diffused substances, etc. The reasons for non-processability were not further investigated in this paper. All other materials could be processed. The plates made out of the heterogeneous materials showed flow directions (see Appendix I: Figures A7-A9). These are due to material accumulations in the compression moulding process.

Total Composition of the Mixed Wastes and Plastic Type Content
The composition of the input materials does not influence the subsequent investigations and is given here only to complete the information. Further extensive and current data on typical SRF composition are given by [22]. The detailed data of the investigated material are given in Appendix A.
Of the fine material (<20 mm), 8.5% OS (OS-original substance) was separated by pre-screening and discarded from SRF. The subsequent manual sorting analysis revealed the following composition of SRF > 20 mm. The plastics represented the largest material fraction with 86.5% OS . The other fraction (sorting residue and composites) represented 6.5% OS . The share of P&C was 5.8% OS . The share of metals and inert materials was 0.6% OS each, and 0.1% OS was the content of wood. The mass losses caused by material drying during storage, sorting losses (mobile organic material, dust formation, etc.), and screening losses were not taken into consideration for calculation here. These are in the range of 3% OS of the total sample.
The sorted out plastic fraction (86.5% OS , see above) consisted of the following plastic types. The PE fraction represented the largest share with 36.4% OS . The other fraction (black and other plastics as well as unidentified objects) represented 21.2% OS . The PET share was 20.7% OS, and PP was contained with 15.7% OS . The smallest fraction was PS with 6% OS . Table 1 displays the evaluations of the DSC measurements. Evaluated were the crystallisation temperature (T C ) with respective crystallisation enthalpy (∆H c ), melting temperatures (T m1 and T m2 ) with respective melting enthalpy (∆H m1 and ∆H m2 ), and glass transition temperature (T g ). A representative cooling curve and the second heating curve per analysed material for the respective material group (PE, PP, PS, PO, and P) are shown in Appendix B for better illustration. Table 1. Results of the DSC measurements: crystallisation temperature (T C ), crystallisation enthalpy (∆H c ), melting temperatures (T m1 and T m2 ), melting enthalpy (∆H m1 and ∆H m2 ), and glass transition temperature (T g ).

Parameters
T The measured T C for the investigated PE materials is, on average, 108 • C. The calculation of the crystallinity for the PE materials according to [13,14] with 293 J/g for totally crystalline PE resulted in values between 36% and 38% for a cooling rate of 20 K/min. This is a comparatively low crystallinity for PE and corresponds to that for virgin (v)LLDPE (10-50%) [23]. vLDPE typically has a crystallinity in the range of 45-55% and vHDPE in the range of 70-80% [23].
Two melting temperatures were determined for the PE materials. The primary melting point (T m1 ) is that most of the material melts are between 125 and 129 • C. The measured secondary melting point T m2 is about 162 • C and is due to contained impurities (higher melting foreign plastics such as PP, for example). Only P_W,PE shows a more distinct secondary melting point at approx. 110 • C. In the literature, melting temperature ranges for different vPE types are given as follows: 120-130 • C for LLDPE, 105-115 • C for LDPE, and 128-136 • C for HDPE [24]. It is interesting to note that the DSC curves are very similar, especially for the heterogeneous PE materials. As was to be expected, these become even more similar through homogenisation, which is evident in the smaller fluctuation margins.
The measured T C for the investigated PP materials is 117 • C on average. According to [13,14], the calculation of the crystallinity with 207 J/g for totally crystalline PP yields values between 55% and 58% for a cooling rate of 20 K/min. This is a relatively high crystallinity for PP. Isotactic vPP has a crystallinity of 70-80%, syndiotactic PP of 30-40%, and atactic PP is amorphous and has no crystallinity [23,25]. The T m1 at approx. 165 • C was determined for the PP materials. Only P_PP and P_C,PP show a distinct T m2 at about 128 • C, which is due to contamination with foreign material, which can be removed by washing. In the literature, melting temperature ranges for vPP types are given between 161 and 186 • C [26]. Likewise, the DSC curves of the heterogeneous PP materials are very similar and, after homogenisation, even closer to each other.
Both heating curves of all investigated PS materials show a continuous decrease over the measured temperature range. This corresponds to the literature, as PS has low heat resistance, and from 55 • C onwards, an acceleration of ageing starts, which is why PS is usually only used up to 70 • C [27]. The measured T g is about 99 • C on average, which corresponds to the literature value of about 80-100 • C for vPS [24,25]. The vPS types predominantly used are atactic and are, thus, in amorphous form and, therefore, have neither a T C nor a T m [24]. Therefore, it is remarkable that both a T C (113-120 • C) and a T m1 (161-164 • C) were measured for the PS materials. The melting temperature is 240 • C [28] for isotactic vPS and 270 • C [28] for syndiotactic vPS. The heterogeneous PS materials' curves are much more heterogeneous compared to those of PE and PP and show more fluctuations and deviations from each other. Due to the homogenisation, these are smoothed considerably and are more similar to each other.
For the PO materials, a T C at approx. 108 • C, a distinct T m1 at approx. 125 • C (∆H m = 50 J/g), and a T m2 at approx. 163 • C (∆H m = 15 J/g) are measured. Furthermore, a further secondary melting temperature is measured at approx. 110 • C. This has already been observed with P_W,PE. With PO_B,C, it is evident compared to the other PO materials that T m1 is more distinct and the secondary melting temperature at 110 • C is hardly present. Additionally, with PO_B,C, a second crystallisation peak at approx. 120 • C becomes clear from the HDPE content [24]. The comparison of the curves of the PO_A materials shows major deviations only for the cooling curves. The other curves are very similar, especially those of the homogeneous PO_A materials. For PO_B, the 2nd heating curves also show major deviations from each other.
The mixed plastic fraction (P) curves are surprisingly similar and show a high degree of similarity with those of PO materials. This is especially true for P_C and PO_B,C.

Mechanical Material Properties
The measured MFR for the PE materials is between 1.8 and 2.6 g/10min. These are very low values. In the literature, MFR values between 0.5 and 25 g/10 min (test conditions: 190/2.16) are given for vLDPE and 0.35-17 g/10 min for vHDPE [24]. These are surprisingly good values, which indicate low material damage and, thus, good processability. No influence of the washing process can be seen.
PP materials show a significant increase in MFR due to washing. Thus, the MFR of P_PP is increased by a factor of 12 for P_W,PP. The MFR for P_C,PP is higher by a factor of three than for P_PP. The homogenised PP materials also show that a 40% higher MFR is achieved by washing. During all PP sample measurements, outgassing of volatile components was observed, which pushed the sample upwards [29,30]. This leads to certain measurement uncertainties. The reasons for this have not been further investigated in this paper. In the literature, MFR values of 0.5-65 (test conditions: 190/5) are found for vPP [24].
The large fluctuation ranges of all PS materials can be explained by the measurement uncertainties caused by clogging of the nozzle after a certain time. This was observed in all PS samples and can be explained by contaminants with a higher melting temperature [29]. The two heterogeneous PS materials show low MFR values compared to the homogenised PS materials. P_W,C,PS shows a lower MFR than P_C,PS. In the literature, MFR values between 1.5 and 18 g/10 min (test conditions: 200/5) are achieved for vPS [24]. Thus, the heterogeneous PS materials can be classified as very easy flowing and the homogeneous materials as normal flowing.
The heterogeneous PO_A materials show the lowest MFR values (below 1 g/10 min). There is no influence of washing on the MFR of the PO_A materials seen. Due to the homogenisation, the MFR rises to the MFR level of the PE materials. PO_B has a mean MFR of 3.3, which is reduced to 2.7 by homogenisation.
P has a mean MFR of 2.3, which is increased to 3.7 by homogenisation. The fluctuation ranges of the mean values can be explained by the measurement uncertainty caused by the outgassing of volatile components after a certain time [29,30]. The reasons for this have not been further investigated in this paper.
The tensile parameters, the Charpy impact strength (a cU ), and the Charpy notched impact strength (a cN ) are reported in Figure 3 for all materials. The results of the impact tests of all investigated materials are given in Appendix E, and the results of the tensile tests are given in Appendix F. Only P_W,C,PE displays plastic deformation with a well-defined yield point in the stress-strain curves [31]. P_PS, P_W,C,PP, P_PE, PO_A, and PO_A,C showed this behaviour only with single test specimens. The other materials showed mainly brittle behaviour [31].
The Young's modulus (E) is for all PE materials in a similar range around 530 ± 40 MPa, and no significant influences due to washing or homogenisation can be detected. Comparison with data from the literature (vLDPE:~200 MPa; vHDPE:~1000 MPa) [24,31] shows that the achievable values are acceptable. According to the literature, vPE has the following tensile strengths (σ M ) and elongations at the yield point (ε M ): vLDPE-8-15 MPa at~20%; and vHDPE-20-30 MPa at~12% [24,31]. The literature gives elongations at break (ε B ) of 400-800% [24]. The measured values for σ M correspond to those for vLDPE. However, both ε M and ε B are far below the literature values. The notched impact strength for all PE materials is in the range between 6 and 15 kJ/m 2 . In the literature, values of about 6 kN/m 2 or without a break are given for vPE [24,31]. For the impact strength, the literature predominantly states "no break". The examined PE materials are mostly only partially broken, and P_W,C,PE is not broken at all (see notes in Tables A6-A25 in Appendix E) [32].
The PP materials have Young's moduli (~1500 MPa) almost three times higher than PE. Interestingly, P_W,C,PP is the lowest value of this material series at 923 ± 19 MPa. Again, a comparison with the literature values (E: 1300-1800 MPa) [23,33] shows that these values are acceptable. For vPP, σ M between 25 and 40 MPa at ε M, around 20% [24], depending on the type, can be found in the literature. Depending on the vPP type, ε B of 200-900% is possible [24]. The measured σ M for PP materials is below 25 MPa, and only an ε B of 2.5 ± 0.4% was measured. As with the PE materials, this indicates significant material embrittlement. The impact strength determined for the PP materials is below 20 kJ/m 2 , which corresponds to the literature values [24]. The a cN determined is between 2 and 5 kJ/m 2 , which is slightly below the literature's values (vPP: 4-12 kJ/m 2 ) [24].
The Young's modulus of P_PS (~955 MPa) shows a sharp increase to~2300 MPa for P_W,PS and about 2400 MPa for P_C,PS and thus, are the highest values of all materials examined. However, the combination of washing and homogenisation causes E (~923 MPa) to drop even below the initial value of P_PS. In the literature, values between 2200 and 3300 MPa [24] are given for vPS. For vPS, σ B between 45 and 65 MPa/mm 2 and 3 and 4% for ε B are found in the literature [24,33]. The measured a cU of the PS materials is below 5 kJ/m 2 and below the values found in the literature (5-20 kJ/mm 2 ) for vPS [31,33]. The a cN of the PS materials are between 1 and 2.5 kJ/m 2 , and this is in the field of the literature values (vPS:~2.0 kJ/m 2 ) [24].
Except for PO_A (1,053 ± 56 MPa), the PO materials have very similar Young's moduli between 830 and 900 MPa. Significant influences due to washing or homogenisation are not recognisable. The Young's moduli of the P and P_C material (~940 MPa) are also very similar, and no influence of homogenisation can be seen.
The values of σ M and ε B increases due to the homogenisation of PE, PP, and PO_B materials. With PS, a significant increase is measured of σ M and ε M by washing or compounding, but in combination, no significant change to P_PS is observed. The PO_A materials all show very similar values for σ M and ε M, with the higher values for PO_A,W,C being achieved. The AC for the PE materials decreases for both the heterogeneous and the homogeneous fraction from about 4% to 2.4% by about 40% due to the washing process. For the two heterogeneous PP materials, no influence of the washing on the AC of about 2.4% was observed. In the homogenised PP fraction, the AC decreases by approx. 40% from 2.7% to 1.7% due to washing. The AC of the PS and PO_A materials has been reduced by approx. 30% for both the heterogeneous and the homogeneous fraction by washing. The unwashed PO_A materials have the second-highest AC of all investigated materials. The average AC for the PO_B is 2.2%, and for PO_B,C 2.5%. The lower contents compared to PO_A can be explained by the cleaner input materials (e.g., pre-sorted mixed plastic fraction) used for PO_B production. As expected, the unwashed, unsorted mixed plastic fraction (P) has the highest AC (approx. 8.4% for P and 6.2% for P_C) since there was no surface cleaning by washing or losing fine material, e.g., by sorting.

Bulk Densities
The determined bulk densities of all materials are shown in Figure 5 before (flakes <4 mm) and after homogenisation (granulates). The results of the bulk densities of all investigated materials are given in Appendix G. Exemplary photos of the produced flakes ( Figure A5), and granulates ( Figure A6) are given in Appendix I.

Flakes
As expected, the bulk densities of all flakes are lower than those of the granules. The PE flakes have a bulk density of 0.082 ± 0.0023 g/cm 3 (P_PE) and 0.072 ± 0.0022 g/cm 3 (P_W,PE). The PP and PS flakes have a bulk density of approx. 0.16 g/cm 3 , almost twice as high. For PP, this can be explained by the higher proportion of compacted, threedimensional particles, although vPP (0.895-0.91 g/cm 3 ) is in the same material density range as vPE (0.87-0.97 g/cm 3 ) [24]. The PE flakes consist mainly of flat, thin, twodimensional particles, although vPS has a higher material density (0.104-0.109 g/cm 3 ) [24] than vPE or vPP, and 2D particles are less common. Nevertheless, PS can occur in an expanded form (EPS: 0.0015-0.009 g/cm 3 ) [31], which would significantly reduce the bulk density. The PO_A flakes have a bulk density of about 0.07 g/cm 3 . This suggests that the flakes contain a high proportion of mainly PE films. The bulk density of PO_B is 0.092 g/cm 3 and of P is 0.113 g/cm 3 . No feeding problems (e.g., bridge formation) were observed for the flakes of all materials.

Granulates
The homogeneous PE granulates have the highest bulk density with 0.49 g/cm 3 . The PP granulates have a bulk density of 0.44 g/cm 3 . The PS granulates have the lowest bulk density of the homogeneous materials with approx. 0.27 g/cm 3 . This can be explained by the fact that degradation processes caused increased outgassing during extrusion, which could not be sufficiently removed (see Figure A6 in Appendix I). Additionally, the different bulk densities of PO_A,C (0.4 g/cm 3 ) and PO_A,W,C (0.46 g/cm 3 ) can be explained in this way. The bulk density of PO_B,C was the second highest with 0.48 g/cm 3 . P_C had a bulk density of 0.44 g/cm 3 . Commercially available plastic granulates have a bulk density between 0.5 and 0.9 g/cm 3 [32]. No feeding problems (e.g., bridge formation) were observed for the granulates produced from all materials.

Discussion
The degree of crystallinity of polymers is directly related to their material properties: the more crystalline a polymer is, the harder and more brittle it is, and dimensional stability and melting point or softening point increase because intermolecular forces can act more effectively due to the more uniform arrangement of the molecules [13,22]. Despite the low crystallinity (36-38%), the PE materials examined, except for P_W,C,PE, show predominantly brittle material behaviour. Since a washing process somewhat improves the mechanical properties, it can be concluded that these are mainly impurities that negatively influence the material properties and that material ageing plays a subordinate role. Likewise, the distinct, second melting temperature at about 128 • C for P_PP and P_C,PP can be explained by the presence of organic impurities (e.g., other plastics with a density >1 g/cm 3 ), which can also be removed by washing. Therefore, T C and T m1 of the PS materials can only be explained by contained organic impurities, e.g., PP.
The DSC curve progressions of the investigated PO materials as well as the determined values of T C and a distinct T m1 at approx. 125 • C and a T m2 at approx. 163 • C indicate a higher PE than PP content of the PO materials. The melting temperature at approx. 110 • C of some PO materials and P_W,PE can be attributed to organic impurities or a higher LDPE content. A second crystallisation peak at approx. 120 • C for PO_B,C becomes clear from the HDPE content [34]. This and the higher ∆H m of T m2 allow the assumption of a somewhat higher PP content than in PO_A. The similar DSC curves of mixed plastics (P) to the investigated PO materials lead to the hypothesis that most investigated P materials consist of PO. This indicates that a separation, and separate PE, PP, and PO processing is not necessary since the thermal properties do not change significantly.
The MFR results are surprisingly good for almost all materials examined, which indicate low material damage and thus, good processability. Depending on the material (high or low viscosity), suitable processing methods must be selected. The investigated PO and P materials have similar MRF values to the investigated PE materials. From this, it can be concluded for the MFR that a separation of the PE materials out of mixed plastics is not mandatory.
Except for the very brittle PS materials, all other materials investigated have surprisingly good mechanical properties. The mechanical characteristics show that wet processing, combined with a homogenisation step, does not necessarily lead to an improvement in mechanical properties. Most of the investigated materials show a clear decrease in mechanical properties compared to virgin homopolymers known from the literature. This indicates the existence of organic and inorganic impurities as well as material degradation due to ageing [35].
The investigations on the ash content of the materials show high inorganic contents, which are mostly significantly reduced by washing. These inorganic impurities are a plausible explanation for the observed deviations between the washed and unwashed materials. Additionally, a short service life (<1 year) is to be expected for the plastics in the wastes examined. It is therefore assumed that material ageing plays only a minor role. A part of the AC is due to inorganic fillers (e.g., glass fibres, silicates, oxides, and hydroxides) in the polymer matrix. Thus, despite the comparatively high ash contents (3-8%), the PO and P materials investigated show a good mechanical property profile. This suggests that the inorganic impurities contained act to a certain extent as a reinforcing material.
Knowledge of the bulk density of free-flowing materials is an essential parameter for the design of storage, transport, and dosing equipment. The bulk density is also essential for the material feed behaviour and the pressure build-up in solid conveying areas of extruders or injection moulding machines [36]. It should be noted that the pelletising system's settings and the melt strand temperature have a significant influence on the pellet geometry and, thus, on the bulk density [31]. Commercially available plastic granulates have a bulk density between 0.5 and 0.9 g/cm 3 [33]. As the granulate bulk densities determined are only slightly lower, with the exception of P_C,PS, and P_W,C,PS, it is assumed that these materials have good conveying and feeding properties. No feeding problems (e.g., bridging) occurred with the flakes and granulates produced from all materials.

Conclusions
The investigations have shown that all waste materials could be processed into plasticrich fractions with a grain size < 4 mm using simple waste treatment without any significant problems. The processing of the different plastic fractions with simple compression moulding showed that all PE, PP, PO, and P materials could be processed without any major problems. This, together with the low MFR values, suggests that conventional extrusion into semi-finished products such as pipes or plates could be technically possible. A list of potential products for the materials investigated is given in Appendix J. The injection moulding process must be tested, and investigations must be carried out with a high-pressure capillary rheometer. Furthermore, thermogravimetric (TGA) and Fouriertransform infrared (FTIR) analysis to determine chemical structure changes possible for polymeric waste during the technological process of the materials is recommended for future investigations. The PS materials emitted gas in both processing variants, and several test runs were necessary to find a stable processing method. Therefore, an evaluation of volatile organic compound emissions from the materials is necessary if they would be implemented in industrial processes. The PET materials could neither be homogenised nor compression moulded due to excessive contamination.
In some cases, the material properties determined are (Young's modulus, impact strength) clearly below those of virgin polymers. This limits the product range that can be manufactured and its range of applications. The results also show that an upstream washing process improves the achievable properties, but homogenisation does not necessarily improve properties. It was also found that a higher treatment depth (recovery of plastic types) from mixed wastes is not necessary since the PO and mixed plastics fractions showed similarly good material data with good processability.
In summary, the investigations show that the recovery and simple treatment of plastics from mixed, contaminated wastes into at least downcycling products seems to be possible. The transfer of used plastics from thermal recovery to recycling could make an important contribution to achieving additional recycling targets, resource conservation, and CO 2 and waste reduction.

Acknowledgments:
The authors would like to thank their student Anna-Lisa Lanthaler who actively supported the execution of experiments. We express our thanks also to the proofreader, Thomas Lucyshyn and Lisa Kandlbauer for the graphic support.

Conflicts of Interest:
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. We express our thanks also to the project partners who contributed significantly to the successful completion of the test series with their equipment and infrastructure.                   Table A53. Possible applications for the materials investigated.