From Electronic Waste to 3D-Printed Product, How Multiple Recycling Affects High-Impact Polystyrene (HIPS) Filament Performances

The rapid growth of the electronics industry is producing excessive electronic waste. One of the common types of materials in electronic waste is high-impact polystyrene (HIPS). In this study, HIPS from electronic waste was recycled through an extrusion process and used as a 3D print filament. The effects of recycling on printability, physical properties, and mechanical properties in horizontal and vertical directions were examined. It was found that until the fourth-cycle, mechanical properties such as horizontal tensile strength, horizontal flexural strength, vertical flexural strength, and vertical impact strength were comparable with virgin commercial filament. In addition, the vertical flexural modulus in the fourth cycle increased by 77.28%. However, the density of recycled HIPs’ first to the fourth cycle slightly decreased by 10.6%, and the melt flow rate increased by 20.3%. It was also observed that until the third cycle, the effect of the reprocessing steps was insignificant on the defect of the 3D-printed product. In general, the experiments show various results, mainly in mechanical properties. Nevertheless, recycled HIPS filaments are comparable to or better than commercial ones in some cases. As a result, recycled HIPS filaments hold the potential to be considered as an alternative to other types of 3D print filaments.


Introduction
In the recent century, polymeric materials experienced a surge in popularity in the electrical and electronic equipment industry (EEE) due to their natural insulator properties. Consequently, the waste from the EEE industry, namely electronic waste (e-waste), grew as old appliances were replaced by newer ones [1]. According to a recent analysis by the United Nations, only 20% of the estimated 44.7 million metric tons (MMT) of e-waste was confirmed to be collected and effectively recycled [2]. Containing an estimate of 20-35% plastics, high-impact polystyrene (HIPS) and acrylonitrile butadiene styrene (ABS) were two of the most abundant plastics found in e-waste [3,4].
The abundance of these plastics can be taken advantage of by recycling. With recycling, the amount of plastic waste can be reduced while having the potential to create high-value products. The recycling process of plastic is not only an environmentally sustainable act but can also be viewed as an economically sustainable act [5][6][7]. One of the products that can be made by recycling plastics is a 3D-printing filament, which is essential in the additive manufacturing (AM) process and, more specifically, the fused deposition modeling (FDM) process. Additive manufacturing is a rapid prototyping technology that builds a structure layer by layer, following a computerized model [8]. The main advantages of the AM process are design flexibility and cost efficiency [9]. According to AM, FDM, better known as 3D printing, is frequently used to create prototypes or final products [10][11][12][13].

Raw Materials and Preparation
The raw materials (rHIPS-1) used in this study were obtained from Kitakaya Raharja Indonesia Inc., (Bekasi, Indonesia), in the form of chopped recycled HIPS (rHIPS) flakes in various sizes that are less than 2 cm, as shown in Figure 2. These flakes were washed

Raw Materials and Preparation
The raw materials (rHIPS-1) used in this study were obtained from Kitakaya Raharja Indonesia Inc., (Bekasi, Indonesia), in the form of chopped recycled HIPS (rHIPS) flakes in various sizes that are less than 2 cm, as shown in Figure 2. These flakes were washed with tap water and dish soap and dried in an oven at 60 • C for 1 h to obtain clean, recycled HIPS flakes. The clean, recycled HIPS flakes were considered first-cycle and would be extruded into second-cycle filament. After obtaining the second-cycle HIPS filament (rHIPS-2), the excess filament would be cut into pellets using a pelletizer at 50 rpm and cutting speed and at 15 rpm puller speed. The pellets were then used as the raw materials for extruding the third-cycle HIPS filament (rHIPS-3). The process was repeated and referred to as the reprocessing step until the fourth-cycle HIPS filament (rHIPS-4) was obtained. On the other hand, the commercially available 3D-printing filament chosen as the baseline in this study was from Shenzen Esun Industrial Co., Ltd., (Shenzhen, China). Table 1 summarizes the samples used throughout this study.

Raw Materials and Preparation
The raw materials (rHIPS-1) used in this study were obtained fro Indonesia Inc., (Bekasi, Indonesia), in the form of chopped recycled in various sizes that are less than 2 cm, as shown in Figure 2. These with tap water and dish soap and dried in an oven at 60 °C for 1 h to ob HIPS flakes. The clean, recycled HIPS flakes were considered first-c extruded into second-cycle filament. After obtaining the second-c (rHIPS-2), the excess filament would be cut into pellets using a pelle cutting speed and at 15 rpm puller speed. The pellets were then used for extruding the third-cycle HIPS filament (rHIPS-3). The process w ferred to as the reprocessing step until the fourth-cycle HIPS filamen tained. On the other hand, the commercially available 3D-printing fila baseline in this study was from Shenzen Esun Industrial Co., Ltd., (Sh ble 1 summarizes the samples used throughout this study.

Filament Extrusion
The extrusion machine used in this study is a simple single-screw extrusion machine made by the Mechanical Production Laboratory and owned by the Materials and Metallurgy Laboratory, Institut Teknologi Bandung (ITB), Indonesia. The extrusion process parameters were set as follows: extrusion temperature at 195 ± 5 • C; screw speed at 19.7 rpm; puller speed at 15 ± 0.5 rpm. The cooling element used in the extrusion process was air from fans. One of the resulting recycled HIPS filaments can be seen in Figure 3.

Filament Extrusion
The extrusion machine used in this study is a simple single-screw extrusion machine made by the Mechanical Production Laboratory and owned by the Materials and Metallurgy Laboratory, Institut Teknologi Bandung (ITB), Indonesia. The extrusion process parameters were set as follows: extrusion temperature at 195 ± 5 °C; screw speed at 19.7 rpm; puller speed at 15 ± 0.5 rpm. The cooling element used in the extrusion process was air from fans. One of the resulting recycled HIPS filaments can be seen in Figure 3.

3D Printing Process
The diameter of recycled HIPS 3D-printing filament was 1.785 ± 0.145 mm. In this paper, FlashForge Adventurer 3 was used as the 3D printer for printing the 3Dbenchy, calibration cube, tensile specimens (ASTM D638 type IV), bending specimens (ASTM D790), and Izod impact specimens (ASTM D256). The prepared specimen models were sliced in FlashPrint 5 software to set the printing parameters. The printing parameters used in this study for the 3Dbenchy, horizontal mechanical specimens, and vertical mechanical specimens are summarized in Table 2. In addition, all mechanical specimen models are further specified into two types in Figure 4.

3D Printing Process
The diameter of recycled HIPS 3D-printing filament was 1.785 ± 0.145 mm. In this paper, FlashForge Adventurer 3 was used as the 3D printer for printing the 3Dbenchy, calibration cube, tensile specimens (ASTM D638 type IV), bending specimens (ASTM D790), and Izod impact specimens (ASTM D256). The prepared specimen models were sliced in FlashPrint 5 software to set the printing parameters. The printing parameters used in this study for the 3Dbenchy, horizontal mechanical specimens, and vertical mechanical specimens are summarized in Table 2. In addition, all mechanical specimen models are further specified into two types in Figure 4.

Density Test
The density of the rHIPS samples was determined according to Test Method B in ASTM D792. This particular test used a 25 mL pycnometer and 96% ethanol as the apparatus and tester fluid.

Density Test
The density of the rHIPS samples was determined according to Test Method B in ASTM D792. This particular test used a 25 mL pycnometer and 96% ethanol as the apparatus and tester fluid.

Melt Flow Rate Test
The melt flow rate test was conducted using the Presto Melt Flow Index Tester at Kitakaya Raharja Indonesia Inc. The testing temperature was set at 200 • C, and the testing load was set to 5 kg, following ASTM D1238.

Gel Permeation Chromatography (GPC)
A gel permeation chromatography test was performed using a Shimadzu LC-20 system equipped with a refractive index (RI) detector at the Serpong Advance Characterization Laboratory to determine the molecular weight of each recycled HIPS filament. The solvent used for the test is tetrahydrofuran (THF), with polystyrene (PS) as the standard material.

Fourier Transform Infrared Spectroscopy (FTIR)
Fourier transform infrared spectroscopy (FTIR) was conducted on each rHIPS specimen using the Bruker FTIR ALPHA II spectrometer (ATR mode) at the Materials and Metallurgy Laboratory, ITB, Indonesia. Each specimen was first cleaned using wipes and technical-grade alcohol before placing them on top of the ATR crystal. The scanning range chosen for the test was between 3500-500 cm −1 with 50 average scans and a spectral resolution of 2 cm −1 .

Tensile Test
The tensile tests were conducted on all the tensile specimens, according to ASTM D638, using the Dynatech Tensilon RTF-1310 at the Materials and Metallurgy Laboratory, ITB, Indonesia. The test included two different types of specimens for measuring HIPS' horizontal and vertical infill tensile strengths, which both were printed in the specimentype IV shape. The 3D printing parameters for the mechanical specimens in both infill directions are stated in Table 2. The test speed was set at 5 mm/min, and the load cell rating was 10,000 N.

Bending Test
The bending tests were conducted on all bending specimens, following ASTM D790, using the Dynatech Tensilon RTF-1310 at the Materials and Metallurgy Laboratory, ITB, Indonesia. This test was conducted on the horizontal and vertical infill specimens, which followed the printing parameters in Table 2 to measure the overall HIPS bending strength. The bending test speed was set at 1.4 mm/min.

Izod Impact Test
The Izod impact tests were conducted on all the impact specimens owing to ASTM D256 using the Digital Presto IZC-222 Impact Tester at Kitakaya Raharja Indonesia Inc. The specimens were modeled after the ASTM D256 description. This test was also conducted with two types of specimens: one was the horizontal infill specimens, and the other was the vertical infill specimens. The horizontal and vertical 3D printing parameters are stated in Table 2.

Visual Inspection
Defect examinations were conducted on each 3D-printed sample to evaluate the printability of the 3D-printing filaments. The defect examinations were conducted by inspecting the HIPS 3Dbenchy specimens printed with the parameters in Table 2.

The Effects of Reprocessing Steps on Density, Melt Flow Rate, and Molecular Weight
The density test was conducted to discover changes in the rHIPS molecular weight after being reprocessed multiple times. Figure 5a shows a notable density decline from rHIPS-1 to rHIPS-2 and a slight decline for the following reprocessing steps. Meanwhile, Figure 5b shows an increasing trend for the MFR values after each reprocessing step, especially from rHIPS-3 to rHIPS-4. The MFR test was intended in order to discover changes in rHIPS viscosity, which can be further linked to the PDI value in the GPC test. These trends of decreasing density and increasing MFR values were due to the degradation of the polymer chains that caused a chain scission phenomenon, leading to a reduction in the HIPS molecular weight and higher PDI values. Furthermore, those statements were backed up by the GPC results that can be found in Table 3. stated in Table 2.

Visual Inspection
Defect examinations were conducted on each 3D-printed sample to evaluate the printability of the 3D-printing filaments. The defect examinations were conducted by inspecting the HIPS 3Dbenchy specimens printed with the parameters in Table 2.

The Effects of Reprocessing Steps on Density, Melt Flow Rate, and Molecular Weight
The density test was conducted to discover changes in the rHIPS molecular weight after being reprocessed multiple times. Figure 5a shows a notable density decline from rHIPS-1 to rHIPS-2 and a slight decline for the following reprocessing steps. Meanwhile, Figure 5b shows an increasing trend for the MFR values after each reprocessing step, especially from rHIPS-3 to rHIPS-4. The MFR test was intended in order to discover changes in rHIPS viscosity, which can be further linked to the PDI value in the GPC test. These trends of decreasing density and increasing MFR values were due to the degradation of the polymer chains that caused a chain scission phenomenon, leading to a reduction in the HIPS molecular weight and higher PDI values. Furthermore, those statements were backed up by the GPC results that can be found in Table 3.   Based on the MFR result in Figure 4b and PDI values in Table 3, it was found that the degradation phenomenon starts to heavily affect the HIPS filament viscosity in the fourth cycle, which is in line with the 35.18% increase in PDI values from the third-cycle to the fourth-cycle HIPS filament. Because of thermal reprocessing, the increasing PDI values represent the chain scission phenomenon in rHIPS chains. As a result, there were shorter chains than before, which increased the polymer's flowability of the polymer, thus lowering the viscosity.
The GPC results in Table 3 show a decreasing molecular weight trend as the reprocessing steps increase. For example, the rHIPS-4 sample shows a decline of 21.88%, from 1691 g/mol to 1321 g/mol in the number average molecular weight (Mn) compared to rHIPS-2, thus validating the decreasing trend in the density test results. However, there is a slight increase in molecular weight between the rHIPS-2 and rHIPS-3 samples, which can be assumed that the thermal oxidative degradation process is incomplete.
By definition, HIPS is a copolymer that comprises polystyrene that is grafted on butadiene-based rubber surfaces, such as polybutadiene (PB) [24]. Therefore, when HIPS is undergoing thermal oxidative degradation, it starts in the rubbery phase of polybutadiene (PB) before spreading to the polystyrene (PS) because the C-H bond of PS is more stable than the C-H bond of PB [25]. However, it was found that PS also underwent thermal degradation simultaneously, which will be further explained in the next subchapter. This incomplete thermal oxidative degradation process, as seen in Figure 6, is initiated by hydrogen abstraction, which leads to the making of hydroperoxide radicals that may have a higher molecular weight than the starting compound; hence, the slight increase in molecular weight [26,27]. The resulting degradation mechanism can lead to two outcomes, as seen in Figure 7: chain scission and crosslinking.
ing the viscosity.
The GPC results in Table 3 show a decreasing molecular weight trend as the reprocessing steps increase. For example, the rHIPS-4 sample shows a decline of 21.88%, from 1691 g/mol to 1321 g/mol in the number average molecular weight (Mn) compared to rHIPS-2, thus validating the decreasing trend in the density test results. However, there is a slight increase in molecular weight between the rHIPS-2 and rHIPS-3 samples, which can be assumed that the thermal oxidative degradation process is incomplete.
By definition, HIPS is a copolymer that comprises polystyrene that is grafted on butadiene-based rubber surfaces, such as polybutadiene (PB) [24]. Therefore, when HIPS is undergoing thermal oxidative degradation, it starts in the rubbery phase of polybutadiene (PB) before spreading to the polystyrene (PS) because the C-H bond of PS is more stable than the C-H bond of PB [25]. However, it was found that PS also underwent thermal degradation simultaneously, which will be further explained in the next subchapter. This incomplete thermal oxidative degradation process, as seen in Figure 6, is initiated by hydrogen abstraction, which leads to the making of hydroperoxide radicals that may have a higher molecular weight than the starting compound; hence, the slight increase in molecular weight [26,27]. The resulting degradation mechanism can lead to two outcomes, as seen in Figure 7: chain scission and crosslinking. Figure 6. Thermal oxidative degradation mechanism of polybutadiene. Adapted with permission from Ref. [26]. 2012, Elsevier. Figure 6. Thermal oxidative degradation mechanism of polybutadiene. Adapted with permission from Ref. [26]. 2012, Elsevier. In the rHIPS-4 samples, one could argue that the thermal oxidative degradation process had been completed, resulting in chain scission of the HIPS chain; therefore, the higher PDI value. An increase of 28.62% in the PDI value from rHIPS-2 to rHIPS-4 can also be linked to the MFR experimental findings that show an increase in the MFR value in rHIPS samples after experiencing more reprocessing steps, which meant the flowability of the chains had been improved due to having a more non-uniform chain length via the chain scission phenomenon. There are also several possibilities of crosslinking between the chains due to the produced free radicals [27]. These combinations of crosslinking possibilities are shown in Figure 7 and can affect the rHIPS molecular weight.

The Effects of Reprocessing Steps on FTIR Spectroscopy
The influence of the reprocessing steps on the rHIPS chains can be further discussed by referring to the FTIR spectroscopy results in Figure 8. In order to obtain a more comparable result of the structural changes in recycled HIPS caused by the reprocessing steps, the reference peak for absorbance normalization was set at 1492 cm −1 , according to the findings of Vilaplana. The peak at 1600 cm −1 was caused by the aromatic C=C stretching vibration, whereas the peak at 1449 cm −1 was related to the aromatic ring stretching vibrations [22]. From the FTIR spectra, the peak corresponding to the polybutadiene microstructure cis-1,4 from the polybutadiene phase can be seen at around 1650 cm −1 [29]. As was mentioned in the previous subchapter, the degradation of HIPS starts not only within the polybutadiene phase but also within the polystyrene phase at the same time. It was In the rHIPS-4 samples, one could argue that the thermal oxidative degradation process had been completed, resulting in chain scission of the HIPS chain; therefore, the higher PDI value. An increase of 28.62% in the PDI value from rHIPS-2 to rHIPS-4 can also be linked to the MFR experimental findings that show an increase in the MFR value in rHIPS samples after experiencing more reprocessing steps, which meant the flowability of the chains had been improved due to having a more non-uniform chain length via the chain scission phenomenon. There are also several possibilities of crosslinking between the chains due to the produced free radicals [27]. These combinations of crosslinking possibilities are shown in Figure 7 and can affect the rHIPS molecular weight.

The Effects of Reprocessing Steps on FTIR Spectroscopy
The influence of the reprocessing steps on the rHIPS chains can be further discussed by referring to the FTIR spectroscopy results in Figure 8. In order to obtain a more comparable result of the structural changes in recycled HIPS caused by the reprocessing steps, the reference peak for absorbance normalization was set at 1492 cm −1 , according to the findings of Vilaplana. The peak at 1600 cm −1 was caused by the aromatic C=C stretching vibration, whereas the peak at 1449 cm −1 was related to the aromatic ring stretching vibrations [22]. From the FTIR spectra, the peak corresponding to the polybutadiene microstructure cis-1,4 from the polybutadiene phase can be seen at around 1650 cm −1 [29]. As was mentioned in the previous subchapter, the degradation of HIPS starts not only within the polybutadiene phase but also within the polystyrene phase at the same time. It was proven by the quite significant increase in C-H bond intensity of the rHIPS samples at around 3000-2840 cm −1 after experiencing more reprocessing steps. That would not have been possible if only the polybutadiene was experiencing thermal oxidative degradation and, thus, hinted at the possibility of polystyrene degrading simultaneously. As a result of the degrading polybutadiene, the C=C bonds of polybutadiene experienced a decrease in intensity with each reprocessing step at around 966 cm −1 .      All horizontal infill specimens have higher tensile strengths than the vertical ones due to the difference in infill to tensile load directions. In the horizontal specimens, the tensile load direction is flat to the infill directions, making it more resilient to failure. On the other hand, the vertical specimens' infill directions are perpendicular to the tensile load directions, making them rely heavily on the interlayer bonding strength.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Tensile Properties
Based on horizontal tensile strength in Figure 9a, it was found that the rHIPS tensile strength increased with each addition of the reprocessing step. It is caused by the  All horizontal infill specimens have higher tensile strengths than the vertical ones due to the difference in infill to tensile load directions. In the horizontal specimens, the tensile load direction is flat to the infill directions, making it more resilient to failure. On the other hand, the vertical specimens' infill directions are perpendicular to the tensile load directions, making them rely heavily on the interlayer bonding strength.
Based on horizontal tensile strength in Figure 9a, it was found that the rHIPS tensile strength increased with each addition of the reprocessing step. It is caused by the All horizontal infill specimens have higher tensile strengths than the vertical ones due to the difference in infill to tensile load directions. In the horizontal specimens, the tensile load direction is flat to the infill directions, making it more resilient to failure. On the other hand, the vertical specimens' infill directions are perpendicular to the tensile load directions, making them rely heavily on the interlayer bonding strength.
Based on horizontal tensile strength in Figure 9a, it was found that the rHIPS tensile strength increased with each addition of the reprocessing step. It is caused by the decreasing total infill gap in the 3D-printed specimens as the rHIPS viscosity and melt strength decrease. This phenomenon has been schematized in Figure 11, highlighting the differences between the second-and fourth-cycle 3D-printed samples. Having a lesser infill gap equals more contact area between the infill and, thus, a higher tensile strength. Another explanation is the temporary crosslinking possibilities resulting from the degradation mechanism, as seen in Figure 7, which also helped increase the tensile strength. However, the same cannot be said for the vertical tensile strength in Figure 9b when the reprocessing steps increased. There is only a slight difference in value between the second, third, and fourth cycles when the standard deviation is taken into account. Therefore, their strength was not that affected by the infill gap. In the vertical infill direction, the specimen's tensile strength relied heavily on the interlayer bonding strength, thus making their strengths similar.
strength was not that affected by the infill gap. In the vertical infill direction, the specimen's tensile strength relied heavily on the interlayer bonding strength, thus making their strengths similar.
The elastic modulus of the horizontal HIPS specimens can be found in Figure 9c, in which the highest horizontal elastic modulus of 2558.27 MPa was in the fourth cycle, and the lowest horizontal elastic modulus, 1514.17 MPa, was in the second-cycle HIPS filament. When compared to the elastic modulus of 3D-printed virgin HIPS, which holds for 1210 MPa, the recycled HIPS filaments show a much better performance for every reprocessing step [30]. Next, the elastic modulus of the vertical HIPS specimens can also be found in Figure 9d, where the highest vertical elastic modulus of 2557.13 MPa was in the second cycle, and the lowest vertical elastic modulus, 1298.7 MPa, was in the commercial HIPS filament. The elastic modulus experimental results have a similar trend to the tensile strength experimental result for the same reason explained in the previous paragraph.
Overall, the horizontal tensile test increased the tensile strength and elastic modulus as the reprocessing steps increased. Although the fourth-cycle HIPS filament boasted the highest value, further testing will be required to find out the limit of reprocessing on recycled HIPS filament tensile strength. Meanwhile, the vertical flexural test demonstrated that a lower tensile strength and elastic modulus are expected with the repetition of the reprocessing steps. Considering the lower viscosity of the fourth-cycle HIPS filament, the lower melt strength was also attributed to weaker interlayer bonding, thus lowering the tensile strength and elastic modulus altogether.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Flexural Properties
The experimental results of the HIPS horizontal flexural strength and flexural modulus are shown in Figure 12a  The elastic modulus of the horizontal HIPS specimens can be found in Figure 9c, in which the highest horizontal elastic modulus of 2558.27 MPa was in the fourth cycle, and the lowest horizontal elastic modulus, 1514.17 MPa, was in the second-cycle HIPS filament. When compared to the elastic modulus of 3D-printed virgin HIPS, which holds for 1210 MPa, the recycled HIPS filaments show a much better performance for every reprocessing step [30]. Next, the elastic modulus of the vertical HIPS specimens can also be found in Figure 9d, where the highest vertical elastic modulus of 2557.13 MPa was in the second cycle, and the lowest vertical elastic modulus, 1298.7 MPa, was in the commercial HIPS filament. The elastic modulus experimental results have a similar trend to the tensile strength experimental result for the same reason explained in the previous paragraph.
Overall, the horizontal tensile test increased the tensile strength and elastic modulus as the reprocessing steps increased. Although the fourth-cycle HIPS filament boasted the highest value, further testing will be required to find out the limit of reprocessing on recycled HIPS filament tensile strength. Meanwhile, the vertical flexural test demonstrated that a lower tensile strength and elastic modulus are expected with the repetition of the reprocessing steps. Considering the lower viscosity of the fourth-cycle HIPS filament, the lower melt strength was also attributed to weaker interlayer bonding, thus lowering the tensile strength and elastic modulus altogether.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Flexural Properties
The experimental results of the HIPS horizontal flexural strength and flexural modulus are shown in Figure 12a strength from the second to the third cycle was caused by degradation in the rHIPS chains, as discussed in the earlier subchapter regarding the MFR and GPC tests. However, as the degradation mechanism continues, a temporary crosslink between the rHIPS chains could form, resulting in the sudden increase of flexural strength and modulus from the third to the fourth cycle. The crosslinking limited the rHIPS chain's movement; hence, stiffer chains with higher bending strength were established. Moreover, the flexural strength of rHIPS can be further improved by implementing the optimal 3D print parameters, as observed in a recent experiment by Tanoto, which utilized different factors, including the orientation position, fill pattern, fill density, and layer thickness [31]. degradation mechanism continues, a temporary crosslink between the rHIPS chains could form, resulting in the sudden increase of flexural strength and modulus from the third to the fourth cycle. The crosslinking limited the rHIPS chain's movement; hence, stiffer chains with higher bending strength were established. Moreover, the flexural strength of rHIPS can be further improved by implementing the optimal 3D print parameters, as observed in a recent experiment by Tanoto, which utilized different factors, including the orientation position, fill pattern, fill density, and layer thickness [31]. Although the same trend of decreasing flexural strength and modulus from the second to the third cycle and increasing flexural strength and modulus from the third to the fourth cycle matches with the horizontal flexural test result, the value changes were not as significant because the vertical bending specimens relied more on the interlayer bonding strength. Then, the increase in the flexural strength and modulus of the fourthcycle HIPS specimens can also occur thanks to the lower viscosity, hence the larger contact area between the infills. In general, the horizontal flexural strength is higher than the vertical ones because more horizontal infills withstood the bending load than the vertical infills, as seen in Figure 13. Figure 14 shows the 3D-printed bending specimens in two different infill courses. Although the same trend of decreasing flexural strength and modulus from the second to the third cycle and increasing flexural strength and modulus from the third to the fourth cycle matches with the horizontal flexural test result, the value changes were not as significant because the vertical bending specimens relied more on the interlayer bonding strength. Then, the increase in the flexural strength and modulus of the fourthcycle HIPS specimens can also occur thanks to the lower viscosity, hence the larger contact area between the infills. In general, the horizontal flexural strength is higher than the vertical ones because more horizontal infills withstood the bending load than the vertical infills, as seen in Figure 13. Figure 14 shows the 3D-printed bending specimens in two different infill courses.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Impact Properties
The experimental results of the impact testing on the HIPS specimens are summarized in Figure 15, and the 3D-printed impact specimens can be seen in Figure 16. The second-cycle HIPS filament had the highest impact strength at 30.34 kJ/m 2 (horizontal infill) and 6.37 kJ/m 2 (vertical infill). Meanwhile, the fourth-cycle HIPS filament held the lowest impact strength with means of 8.33 kJ/m 2 (horizontal infill) and 4.76 kJ/m 2 (vertical infill). The experiment result of rHIPS-2 shows that with increasing the reprocessing steps, the overall horizontal impact strength of the HIPS filament decreased while the vertical impact strength only marginally decreased. Continuing from the MFR test result in Figure  5b, which had a trend of increasing value after each reprocessing step, the lower viscosity gave way to the rHIPS filament to fill the gap between the infill. With fewer infill gaps, the overall porosity in the impact specimens decreased; therefore, there was little to no slip motion that could help absorb the impact load.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Impact Properties
The experimental results of the impact testing on the HIPS specimens are summarized in Figure 15, and the 3D-printed impact specimens can be seen in Figure 16. The second-cycle HIPS filament had the highest impact strength at 30.34 kJ/m 2 (horizontal infill) and 6.37 kJ/m 2 (vertical infill). Meanwhile, the fourth-cycle HIPS filament held the lowest impact strength with means of 8.33 kJ/m 2 (horizontal infill) and 4.76 kJ/m 2 (vertical infill). The experiment result of rHIPS-2 shows that with increasing the reprocessing steps, the overall horizontal impact strength of the HIPS filament decreased while the vertical impact strength only marginally decreased. Continuing from the MFR test result in Figure  5b, which had a trend of increasing value after each reprocessing step, the lower viscosity gave way to the rHIPS filament to fill the gap between the infill. With fewer infill gaps, the overall porosity in the impact specimens decreased; therefore, there was little to no slip motion that could help absorb the impact load.

The Effects of Reprocessing Steps on Horizontal and Vertical rHIPS Impact Properties
The experimental results of the impact testing on the HIPS specimens are summarized in Figure 15, and the 3D-printed impact specimens can be seen in Figure 16. The secondcycle HIPS filament had the highest impact strength at 30.34 kJ/m 2 (horizontal infill) and 6.37 kJ/m 2 (vertical infill). Meanwhile, the fourth-cycle HIPS filament held the lowest impact strength with means of 8.33 kJ/m 2 (horizontal infill) and 4.76 kJ/m 2 (vertical infill). The experiment result of rHIPS-2 shows that with increasing the reprocessing steps, the overall horizontal impact strength of the HIPS filament decreased while the vertical impact strength only marginally decreased. Continuing from the MFR test result in Figure 5b, which had a trend of increasing value after each reprocessing step, the lower viscosity gave way to the rHIPS filament to fill the gap between the infill. With fewer infill gaps, the overall porosity in the impact specimens decreased; therefore, there was little to no slip motion that could help absorb the impact load.
Consequently, the rHIPS-4 horizontal impact specimens absorbed the least impact load among the others. This statement also applies to the vertical impact results, however, with a lesser magnitude, as their strength relies more on interlayer bonding. Nonetheless, the rHIPS-4 specimens were still comparable to the commercial HIPS impact strength. Furthermore, research conducted by Patterson provides the 3D-printed virgin HIPS impact strength data of 14.1 kJ/m 2 , which are lower than rHIPS-2 and rHIPS-3 but higher than rHIPS-4 [32]. This comparison consolidates the potential use of recycled HIPS filament in products that require high-impact absorbance. Figure 17 was added to help visualize the difference between the horizontal and vertical infill impact specimens when subjected to an impact load.  Consequently, the rHIPS-4 horizontal impact specimens absorbed the least impact load among the others. This statement also applies to the vertical impact results, however, with a lesser magnitude, as their strength relies more on interlayer bonding. Nonetheless, the rHIPS-4 specimens were still comparable to the commercial HIPS impact strength. Furthermore, research conducted by Patterson provides the 3D-printed virgin HIPS impact strength data of 14.1 kJ/m 2 , which are lower than rHIPS-2 and rHIPS-3 but higher than rHIPS-4 [32]. This comparison consolidates the potential use of recycled HIPS filament in products that require high-impact absorbance. Figure 17 was added to help visualize the difference between the horizontal and vertical infill impact specimens when subjected to an impact load.  Consequently, the rHIPS-4 horizontal impact specimens absorbed the least impact load among the others. This statement also applies to the vertical impact results, however, with a lesser magnitude, as their strength relies more on interlayer bonding. Nonetheless, the rHIPS-4 specimens were still comparable to the commercial HIPS impact strength. Furthermore, research conducted by Patterson provides the 3D-printed virgin HIPS impact strength data of 14.1 kJ/m 2 , which are lower than rHIPS-2 and rHIPS-3 but higher than rHIPS-4 [32]. This comparison consolidates the potential use of recycled HIPS filament in products that require high-impact absorbance. Figure 17 was added to help visualize the difference between the horizontal and vertical infill impact specimens when subjected to an impact load.

The Effects of Reprocessing Steps on the Visual of 3D-Printed HIPS Product
At the time of writing, the 3Dbenchy model was anonymously accepted as one of the standards in the 3D printing community, with increasing research authors mentioning it in scientific articles; for example, Rousseau and Novak [33,34]. The visual examinations were conducted by analyzing the visual defects on 3D-printed 3Dbenchy using commercial and rHIPS filaments. The results were then summarized in Table 4. All 3Dbenchy specimens had some defects, such as ringing, blobs, and zits, and overhang. Meanwhile, Figure 18 shows that the rHIPS-4 3Dbenchy had an additional stringing defect near the back when compared to the rest. This defect was caused by the rHIPS-4 filament having a higher MFR value, leading to some of the filament at the hot end of the 3D printer oozing out while the extruder was moving around. Overall, the difference in the visual defects of the 3D-printed rHIPS-2 and rHIPS-3 specimens is negligible compared to the 3D-printed commercial HIPS specimens. However, the rHIPS-4 filament shows a downgrade by having a stringing defect.

The Effects of Reprocessing Steps on the Visual of 3D-Printed HIPS Product
At the time of writing, the 3Dbenchy model was anonymously accepted as one of standards in the 3D printing community, with increasing research authors mentioning in scientific articles; for example, Rousseau and Novak [33,34]. The visual examinatio were conducted by analyzing the visual defects on 3D-printed 3Dbenchy using comm cial and rHIPS filaments. The results were then summarized in Table 4. All 3Dbenc specimens had some defects, such as ringing, blobs, and zits, and overhang. Meanwh Figure 18 shows that the rHIPS-4 3Dbenchy had an additional stringing defect near back when compared to the rest. This defect was caused by the rHIPS-4 filament havin higher MFR value, leading to some of the filament at the hot end of the 3D printer oozi out while the extruder was moving around. Overall, the difference in the visual defects the 3D-printed rHIPS-2 and rHIPS-3 specimens is negligible compared to the 3D-prin commercial HIPS specimens. However, the rHIPS-4 filament shows a downgrade by h ing a stringing defect.

The Effects of Reprocessing Steps on the Visual of 3D-Printed HIPS Product
At the time of writing, the 3Dbenchy model was anonymously accepted as one of the standards in the 3D printing community, with increasing research authors mentioning it in scientific articles; for example, Rousseau and Novak [33,34]. The visual examinations were conducted by analyzing the visual defects on 3D-printed 3Dbenchy using commercial and rHIPS filaments. The results were then summarized in Table 4. All 3Dbenchy specimens had some defects, such as ringing, blobs, and zits, and overhang. Meanwhile, Figure 18 shows that the rHIPS-4 3Dbenchy had an additional stringing defect near the back when compared to the rest. This defect was caused by the rHIPS-4 filament having a higher MFR value, leading to some of the filament at the hot end of the 3D printer oozing out while the extruder was moving around. Overall, the difference in the visual defects of the 3D-printed rHIPS-2 and rHIPS-3 specimens is negligible compared to the 3D-printed commercial HIPS specimens. However, the rHIPS-4 filament shows a downgrade by having a stringing defect.

Conclusions
This study successfully produced low-cost recycled HIPS filaments from electronic waste, which is non-biodegradable and abundant. All the recycled filaments possessed good mechanical properties and printability except for the fourth cycle, which had slightly worse printability than the others due to an increased melt flow rate. Furthermore, the melt flow rate and density of recycled HIPS filaments were somewhat affected by the thermal-oxidative degradation due to reprocessing. The experimental results show that the rHIPS-4 filament has the highest horizontal tensile strength of 21.48 MPa, while rHIPS-3 has the highest vertical tensile strength of 15.69 MPa. For the highest flexural strength, rHIPS-4 boasted the highest in both infill directions, 34.35 MPa for horizontal and 22.79 MPa for vertical. Lastly, regarding the highest impact strength, the rHIPS-2 filament holds the record of 30.34 kJ/m 2 in the horizontal and 6.37 kJ/m 2 in the vertical infill directions. Considering recycled HIPS filaments' mechanical and printability properties, the total reprocessing steps can still be pushed to a greater extent. In this case, this is until the recyclability limitation of rHIPS as a 3D-printed filament is reached. In conclusion, HIPS from electronic waste can be recycled into 3D-printed filament until the fourth cycle, with comparable or even better mechanical properties when compared to commercial ones.