Recycle of printed circuit boards from waste electric and electronic equipment and their reusability as filler in 3D printed poly(lactic) acid composites

Abstract

Recycling raw materials (RMs) from waste electric and electronic equipment (WEEE) and reusing them in additive manufacturing applications, has tremendous benefits, including health risk reduction by landfill decongestion. Furthermore, final composite products are given enhanced properties, increasing their added-value commercialization. The non-conductive substrate of printed circuit boards, composed of glass fibre-reinforced epoxy resin was processed and tested as poly(lactic acid) (PLA) filament additive. Scanning electron microscopy confirmed carbon, oxygen, and silicon as the main elements of the composite. Differential scanning calorimetry of 5% and 10% composites showed that the addition of fillers did not result in a significant change of composite filaments’ thermal properties. 15% filler addition resulted in higher crystallinity and melting point. Rheological analysis showed that composite filaments’ viscosity increased compared to pure PLA (300–400 Pas compared to 100–200 Pas), maintaining their structural strength during printing. Mechanical performance analysis showed that Young’s modulus, flexural modulus, flexural strength, and elongation of the composites increased compared to pure PLA (up to 9.6%, 24.6%, and 28.2%, respectively), enhancing mechanical properties, structural integrity, and load-bearing capacity. The abundance and low cost of RM and the small number of processes, present the upscaling potential, increasing the profit margin in a swiftly growing market.

Keywords:

• Additive manufacturing
• circular economy
• waste recycle
• thermoplastics
• 3D printing
• glass fibres

1. Introduction

Material extrusion (MEX) three-dimensional (3D) printing, a subcategory of additive manufacturing (AM), is an advanced layer-by-layer manufacturing procedure based on a computer-aided design (CAD) model (Ligon et al., 2017). Currently, MEX (also referred to as Fused Filament Fabrication – FFF) is the most widely used AM process for fiber-reinforced thermoplastic polymer composite production, due to its low cost, simplicity, efficiency, and diversity (Sam-Daliri et al., 2023; Zhang et al., 2021). The production of customized shapes, geometries, and models, along with the plethora of polymeric matrices used, has inextricably linked MEX 3D printing technology with various, multidisciplinary applications such as medicine, electronics, sensors, and automation (Sam-Daliri et al., 2023; Wang et al., 2018).

Poly(lactic) acid (PLA), one of the most widely used thermoplastics, is recyclable, biodegradable, and biocompatible and its applications cover a wide spectrum (Bergström and Hayman, 2016; Schaedler and Carter, 2016). Its disadvantages include relatively high brittleness and low fracture toughness and impact strength (Bergström and Hayman, 2016). Its glass transition temperature (T_g) lies between 55 and 60 °C and is a function of molecular weight and stereochemistry, under which PLA is brittle (Žagar et al., 2020). PLA crystallizes above its crystallization temperature (T_c) in the range 80–150 °C (Gonçalves et al., 2017; Madeleine, 2021). The melting temperature (T_m) of PLA lies between 150 and 160 °C, above which, PLA behaves as a viscous fluid. These weaknesses limit its use in high performance applications. As a result, specific fillers, both of natural and synthetic origin, can often be used to improve materials’ processing and functional properties (enhancement of physical and mechanical properties or even addition of novel properties) (Sztorch et al., 2022).

The mechanical properties of lactic acid-based polymers can be largely varied, ranging from soft and elastic plastics to stiff and high strength materials. Semi-crystalline PLA is preferred over the amorphous polymer when higher mechanical properties are desired (Farah et al., 2016). Plasticizers can be added as fillers in order to substantially reduce T_c while maintaining high mechanical properties (Pillin et al., 2008). Uniform filler dispersion usually leads to enhanced properties, but poor dispersion may result in a drastic deterioration of them (Tjong et al., 2008). Finally, the filler content is a parameter, whose optimization is of great importance as the nozzle of the printer can easily be blocked, and it depends on the filler morphology and size (Pillin et al., 2008; Tjong et al., 2008).

The electrical and electronics industry is one of the most innovative and rapidly advanced, generating large amounts of waste. A United Nations report estimates that only in 2019, 53.6 million tons of waste electric and electronic equipment (WEEE) were discarded (the weight of WEEE discarded globally documented a 24% increase over 5 years), with 17.4% of them (9.3 million tons) being recycled. The target for 2023 is 30% (Iannace et al., 2014; Zhu et al., 2020).

Printed circuit boards (PCB) are abundant in practically all electronic equipment to mechanically support and electrically connect the components using conductive tracks (Forti et al., 2020). PCB are composed of three types of materials: a non-conductive substrate, printed conductive tracks, and components mounted on the substrate. The substrate typically contains glass fiber-reinforced epoxy resin, or paper reinforced with phenolic resin (primarily designed to insulate the copper (Cu) circuits on outer layers from oxidation from the environment), both with brominated flame retardants. The substrate constitutes more than 40 wt% of the PCB (Onwughara et al., 2010). Printed conducting tracks and components mounted on the substrate are comprised of precious metals and rare earth elements (REE). Four main types of materials can be retrieved from PCB: recyclable metals, recyclable polymeric materials, ceramic materials, and non‐recyclable polymers or other contaminants (Bizzo et al., 2014; Stratiotou Efstratiadis and Michailidis, 2022).

The reusability of the PCB substrate as raw material (RM), can lead to both environmental benefits, e.g. the decongestion of landfills from potentially hazardous materials, as well as significant economic benefits, deriving from recycling and reusing end-of-life (EOL) components in added value new materials, following the technological requirements.

Mechanical (multi‐crushing, grinding, electrostatic separation, gravity separation, fluid‐bed separation, density‐based separation, and magnetic separation) and hydrometallurgical (dissolution of metals in alkaline or acid media) methods of recycling EOL PCB offer a relatively low capital cost, reduced environmental impact, and high recoveries of metals compared to pyrometallurgical (heating of WEEE at high temperatures to recover valuable metals, producing hazardous flue gases that must be removed from the air) (Bizzo et al., 2014; Guo et al., 2009; Jadhav and Hocheng, 2015; Yang et al., 2012).

The primary objective of this work is the manufacturing of MEX 3D printed composite materials, using PLA as the polymeric matrix and the PCB substrate as the incorporated filler. The properties of the composites will then be tested against pure PLA MEX 3D printed materials, using Scanning Electron Microscopy (SEM) analysis, Differential Scanning Calorimetry (DSC) analysis, rheological analysis, and mechanical performance analysis (Li et al., 2018). This work has the potential to make a meaningful contribution to both the field of materials science and the broader goal of sustainable-by-design, zero-waste, and innovative manufacturing processes.

The novelty of this study lies in its innovative approach to recycling the non-conducting substrate of the abundance of the electronic waste RM, its focus on enhancing the properties of MEX 3D printed materials, and its potential for large-scale application due the small number of processes required.

2. Materials and methods

Thermoplastic PLA Ingeo™ Biopolymer 4043D for MEX 3D printing monofilament extrusion by NatureWorks LLC (Plymouth, MN, USA) was used as the matrix material in this study. It was purchased from 3devo B.V. (Utrecht, Netherlands) in the form of powder with a density of 1.24 g/cm³ (ASTM D792). Green FR4 fiberglass-reinforced epoxy resin sheets EPGC202 with a density of 1.9–2.1 g/cm³ were procured from Masterplatex e.K. (Berlin, Germany) and converted into powder mechanically through abrasion. Finally, the powdered thermosetting composite material was sieved into various size fractions. The finest powder obtained (<30 μm) was used for the composite filament production.

Filaments with different PCB precursor powder loadings (5, 10, and 15 wt.%) were manufactured, which are denoted as PLA_PCB_x, where x = 0.05, 0.1, 0.15, respectively. First, the PLA and PCB powders were dried at 80 °C for 4 h in a forced convection oven. The powder mixture homogenization was achieved by a 30-min mechanically induced rotation of their containers. A Next 1.0 Advanced (3devo B.V, Utrecht, Netherlands) desktop single-screw filament extruder, equipped with a self-regulating filament diameter control system, was employed for MEX. The filament’s diameter was set to 1.75 mm and the screw speed at 3.5 rpm. The four zones’ temperatures were set at 170, 185, 190, and 170 °C as the feedstock material passes through the extruder’s screw. These temperature values have been given by the extruder’s manufacturer for neat PLA filament production. During MEX, the filament is rapidly heated above its melting temperature, deposited, and then allowed to cool. During cooling, crystal formation occurs between the glass transition temperature and the melting temperature of the polymer (Mcniffe et al., 2023). Finally, the flow chart of the process is given in Figure 1.

Figure 1. Steps followed from RW to final MEX 3D printed products.

Pure PLA and the composite filaments were printed using in a Prusa i3 MK3S 3D-printer to print out specific samples for mechanical properties testing (Li et al., 2018). The printer and filament settings are given in Appendix A.

PCB were subjected to SEM analysis. SEM uses a focused beam of high-energy electrons to generate signals at the surface of solid samples. The signals reveal information about the external morphology (texture), chemical composition, and internal crystalline structure of the sample. SEM analyses took place using Phenom ProX G6 from Thermo Fischer Scientific, using long lifetime thermionic source (CeB₆), with light optical magnification range: 20–134x and electron optical magnification range: 160–350,000x (“SEM,” XXXX; “Thermofischer (SEM),” XXXX). It is also equipped with an integrated energy-dispersive X-ray diffraction (EDX) detector, which makes elemental analysis of the sample possible.

PCB along with pure PLA filaments were subjected to DSC analysis. With DSC, thermal events such as T_g, T_m, T_c, cure reactions, onset of oxidation, and enthalpies can be measured (Swapp, (n. d.)). Its principle is based on the fact that when a sample undergoes some physical transformation, such as a phase change, more (or less) heat must flow through it than the reference material to keep both at the same temperature (Rajisha et al., 2011). The sample is encapsulated in an aluminium pan and, along with an empty reference pan, sits on a thermoelectric disk surrounded by the furnace. DSC analysis took place using a Discovery DSC 25 instrument from TA Instruments with refrigerated cooling system RSC90 (temperature range from –90 to 550 °C and heating rates from 0.01 to 100 °C/min) and sealed aluminium samplers (Swapp, (n. d.)).

All filaments were subjected to rheological analysis, which measures the amount of deformation material or liquid undergoes when a shear stress is applied. The analyses took place using a Discovery Hybrid Rheometer, HR 20 from TA Instruments, with a parallel plate arrangement and an Environmental Test Chamber (ETC), heating system with convection, temperature range from –160 to 600 °C and heating rate up to 60 °C/min) (TA Instruments/Waters Corporation, (n. d.)).

Finally, the mechanical performance analysis (tensile strength, 3-point-bending) of the pure and composite filaments took place using Electroforce 3550 from TA Instruments. It features a 15 kN dynamic force capability, using two linear motors in tandem. More specifically, it uses multiple force and frame configurations, delivering up to 15,000 N of force for reliably controlling force, displacement, or strain over a wide range of frequencies in tension or compression and dynamic characterization test of a wide variety of materials, components, and devices (TA Instruments – ElectroForce Systems Group (n. d.)).

3. Results and discussion

3.1. SEM analysis

SEM analyses of the PCB precursor powder before sieving, as well as the composite with the higher filler concentration can be seen in Figure 2. The element composition of four different points (two of the unsieved powder and two points of the PLA_PCB_0.15 composite) is given in Table 1. It can be seen that the powder before sieving, contains non-uniformly sized particles, and this would undoubtedly affect not only the composite filament’s output speed but also its MEX 3D printing and mechanical properties (Sam-Daliri et al., 2023).

Figure 2. SEM analysis at magnification levels (a) 1000× and FR4 point selection for element identification tool (b), 3300× and PLA_PCB_0.15 point selection for element identification tool.

Table 1. Elemental composition of selected points by EDS.

No.	Element number	Element symbol	Atomic concentration (%)
1	6	C	91.36
	8	O	8.64
2	6	C	64.02
	8	O	28.74
	14	Si	7.24
3	6	C	18.92
	14	Si	27.4
	8	O	53.68
4	6	C	92.57
	8	O	7.43

Points containing (C) confirm the polymeric nature of epoxy resin in the powder and the successful blend within the PLA matrix. Points containing silicon (Si) and oxygen (O) confirm the existence of glass fibres within the resin. As mentioned in the literature, glass fibres are expected to enhance mechanical properties, when aligned with the extrusion direction (Mcniffe et al., 2023; Sam-Daliri et al., 2022, 2023).

3.2. DSC analysis

The protocol used in the DSC analyses shown in Figure 3 included a heating ramp of 10 °C/min up to 230 °C, a cooling ramp of 10 °C/min down to 20 °C, and a 2nd heating ramp of 10 °C/min up to 230 °C to rule out the thermal history of the polymer during manufacturing. T_g of composite materials is slightly decreased compared to the one of Pure PLA and practically remains the same with higher filler concentration. Similarly, T_m of composites is lower than pure PLA, as the corresponding PLA concentration is lower than 100% and remains the same with higher filler concentration. Moreover, as PLA percentage decreases in composite filaments, the heat flow during melting is also decreased, indicating a reduction of PLA crystallinity.

Figure 3. DSC analysis of pure and composite filaments.

All samples, except for PLA_PCB_0.15, exhibit a broad exothermic crystallization peak during heating, starting at around 110 °C. The trend of the normalized heat flow graphs of pure PLA and the composite filaments are similar. The heat flow of composite filaments during the exothermic crystallization reaction is lower than that of pure PLA and is decreased with increasing filler percentage, as the PLA concentration decreases.

According to the literature, there are more than one crystalline structures in PLA and the filler and the phase transition from the disorder-to-order (δ to α crystalline structure), takes place during crystallization, where the chain packing of the crystal lattice becomes more compact, leading to a slight overlapping of the endotherm and the exothermic curves during the analysis (Wang et al., 2019; Zhang et al., 2021). DSC analysis of PCB powder is given in Appendix B and shows no other thermal transitions up to 300 °C (Delaizir and Calvez, 2012).

The increased filler particles concentration (epoxy resin and glass fibres) can act as nucleating sites, enhancing cold crystallization (Kiatiporntipthak et al., 2021; Lingesh et al., 2014). The act of powdered PCB filler as nucleating agent of the thermoplastic matrix has been also confirmed for polypropylene composites (Kumar et al., 2016).

Table 2 presents T_g, T_c, and T_m of pure and composite filaments. T_g and T_m remain practically the same, as PLA percentage decreases and are in accordance to the literature (Malinowski et al., 2015). However, T_c presents a significant decrease in the PLA_PCB_0.15 composite, due to the chain packing of the crystal lattice becoming more compact impeding its mobility and shifting its T_c to a lower temperature, along with the addition of nucleating agent, which initiates the crystallization at lower temperature (Malinowski et al., 2015; Shi et al., 2015; Zhang et al., 2021).

Table 2. T_g, T_c, and T_m of pure and composite filaments.

Sample	Tg (^oC)	Tc (^oC)	Tm (^oC)
PLA Pure	63.11	125.19	153.51
PLA PCB 0.05	62.58	124.78	150.88
PLA PCB 0.1	62.59	126.32	150.89
PLA PCB 0.15	62.82	120.26	151.61

The overlapping of exothermic and endotherm peaks in PLA_PCB_0.15 can be attributed to the existence of two types of crystal lamellae with different thicknesses. The melting of the thinner lamellae would be related to the lower temperature endotherm, and melting of the thicker to the higher (lammeral thickness model (Bassett et al., 1988)). Another theory is that the lower temperature endotherm is produced by the melting of the lamellae initially present. The partially melted amorphous material would recrystallize into a more perfect lamellae, which would melt at higher temperatures (melting-recrystallization model (Gracia-Fernández et al., 2012; Jonas et al., 1995).

The crystallinity (X_c) of the composite filaments is calculated using Equation (1)(1) $$x_{\mathrm{c}}\left(\mathrm{\%}\right)=\frac{{\mathrm{\Delta }H}_{\mathrm{m}}+{\mathrm{\Delta }H}_{\mathrm{\text{cc}}}}{{\mathrm{\Delta }H}_0\phi }\mathrm{*}100$$(1) (Gong et al., 2016), (1) $$x_{\mathrm{c}}\left(\mathrm{\%}\right)=\frac{{\mathrm{\Delta }H}_{\mathrm{m}}+{\mathrm{\Delta }H}_{\mathrm{\text{cc}}}}{{\mathrm{\Delta }H}_0\phi }\mathrm{*}100$$(1) where melting (ΔΗ_m) and crystallization (ΔΗ_cc) enthalpies are calculated using the integral of the area under the endothermic and exothermic transition of the DSC graph, respectively, φ is the percentage of the PLA matrix in the composite and ΔΗ₀ is the melting enthalpy of 100% perfectly crystalline PLA (93 J/g) (Huang et al., 2009). Table 3 presents X_c of pure PLA and the composite filaments.

Table 3. X_c of pure and composite filaments.

Sample	ΔHm (J/g)	ΔHcc (J/g)	φ	Xc (%)
PLA Pure	17.45	−14.9	1	2.74
PLA PCB 0.05	14.45	−12.24	0.95	2.5
PLA PCB 0.1	12.95	−9.79	0.9	3.77
PLA PCB 0.15	12.25	−3.59	0.85	10.95

As the filler percentage increases, PLA percentage decreases, leading to the decrease of both enthalpies, especially ΔΗ_cc. On the other hand, X_c presents a swift increase, as filler’s concentration, which acts as crystal nucleation agent, increases above 10% (Balani et al., 2015; Gracia-Fernández et al., 2012). The proportional trend between filler concentration and crystallinity has been observed in the literature by Tee et al. (2017), Pei et al. (2010), and Farid et al. (2018).

3.3. Rheological analysis

The logarithmic curves of the viscosity at 215 °C as a function of shear rate and the nominal curves of the shear stress as a function of shear rate are given for the pure PLA and PLA_PCB composite filaments in Figures 4 and 5, respectively.

Figure 4. Viscosity against shear rate for pure and composite filaments.

Figure 5. Stress against shear rate for pure and composite filaments.

The viscosity of the pure PLA composite remains relatively steady with increasing shear rate and velocity from 0.01 up to around 0.1 rad/s, exhibiting Newtonian behaviour. After such value, it exhibits a general decrease in viscosity with increasing shear stress, exhibiting pseudoplastic behaviour. The viscosity of the PLA_PCB composites slightly decreases with an increasing shear rate, exhibiting only pseudoplastic behaviour (Kaully et al., 2008; Zołek-Tryznowska, 2015).

Pure PLA presents lower viscosity in the whole range of the explored shear rate applied than composite PLA_PCB filaments. Higher hard filler concentration in composites leads to particles being more packed together favouring particle-particle interactions over matrix-particle interactions (Yin et al., 2005). The friction between particles results in energy dissipation, and thus, a greater viscosity (Kalb and Pennings, 1980; Malinowski et al., 2015; Thumsorn et al., 2022). Furthermore, the interactions and the relatively small specific area of the filler particles can lead to a potential increase of the of agglomeration probability, also resulting to a higher viscosity (Hausnerova et al., 2009; Rueda et al., 2017; Shi et al., 2015).

Higher viscosity can lead to improved mechanical properties as it can promote stronger layer adhesion as the material cools and solidifies (Beauson et al., 2022; Berretta et al., 2016). However, in other cases, it may pose challenges that need to be addressed through process optimization consideration and adjustment of printing parameters to ensure that the material flows properly and adheres effectively.

3.4. Mechanical performance analysis – Tension tests and 3-point-bending tests

It is noted that for every filament, three experiments took place to confirm the results and rule out the error. Specimens for each mechanical performance test were printed. For the tension tests, dimensions were based on ASTM D638-10 TYPE IV and for 3-point-bending tests on ISO 178 standards, respectively (dimensions given in Appendix C).

The tension tests resulted in the stress/strain graph shown in Figure 6. The elastic (or Young’s) modulus (E) is calculated from the slope of the linear part of the graph (Bassett et al., 1988; Beauson et al., 2022; Jonas et al., 1995). The mean value of E for all filaments is given in Figure 7, along with the maximum and minimum values of the three experiments. It can be seen that the insertion of fillers, increases stiffness and improves the fiber/matrix interface (Tazibt et al., 2023). The composite’s behaviour becomes more resistant to deformation in the elastic range, leading to a higher slope, hence the higher E, compared to pure PLA. Furthermore, higher percentage of filler, leads to increased E as expected. The addition of 15% filler leads to a swift increase of E, compared to the addition of 5 and 10%. In the case of PLA_PCB_0.15, DSC analysis showed that phenomena such as melting and cold crystallization of the filler occur, leading to the formation of crystals (increased crystallinity) and consequently to less space for layer movement, meaning that the sample needs more stress to achieve the same displacement (Beauson et al., 2022).

Figure 6. Stress/strain diagram of all four filaments – tension test.

Figure 7. Elastic modulus derived from tension tests of all four filaments.

Stress is transferred from the matrix to the filler by a shear transfer, as hard fillers tend to concentrate stresses. Tensile strength (σ) of composites is decreased, as the potential formation of filler aggregates can weaken the stress transfer between the PLA matrix and filler, disturbing the continuity of the matrix and limiting the capacity of the polymer chains to withstand stress (Tazibt et al., 2023). Composite filaments present lower σ than pure PLA, as the percentage of PLA decreases (Wang et al., 2019). However, it presents an increasing trend with filler increase, as seen in Figure 8. σ is the maximum strength achieved during testing, just before the snap of the sample and is measured using Equation (2)(2) $$\sigma =\frac{F}{A}$$(2) : (2) $$\sigma =\frac{F}{A}$$(2) where F is the tensile force (N) and A is the nominal cross-section of the specimen.

Figure 8. Tensile strength derived from tension tests of all four filaments.

It is worth mentioning that σ of composite filaments of 10 and 15% PCB concentration is higher than the tensile strength of 90 and 85% of pure PLA, respectively.

Finally, the elongation at break (ε) is calculated using formula (3), where the gage lengths are normalized (the neck dimensions of the samples, are given in Appendix C). (3) $$\epsilon =\frac{l-l_0}{l_0}\mathrm{*}100$$(3) where l is the final gage length (mm) and l₀ is the initial gage length.

ε increases gradually with the increase of PCB filler insertion (5 and 10%), as the addition silica (or silicon dioxide – SiO₂) favours elongation up to an optimized concentration, after which leads to its decrease, as seen in Figure 9 (Tazibt et al., 2023; Wu et al., 2013). PLA_PCB_0.15 has the lowest ε as ε is more sensitive to crystallinity, and PLA_PCB_0.15 presented the highest crystallinity (Yang et al., 2021). The mean values are given in Table 4.

Figure 9. Elongation at break derived from tension tests of all four filaments.

Table 4. Mean E, σ, and ε of pure and composite filaments – tension tests.

Sample	E (GPa)	σ (MPa)	ε (%)
PLA Pure	1.91	53.75	5.88
PLA PCB 0.05	1.94	48.97	6.85
PLA PCB 0.1	2	49.4	8.19
PLA PCB 0.15	2.15	49.72	5.55

Flexural elastic modulus (E_f) is measured via 3-point-bending test. It is noted again, that for every filament (pure and composite), three experiments took place to confirm the results and rule out the error. The 3-point-bending tests result in the load/displacement graph shown in Figure 10. E_f is calculated using formula (4) (b and d of each sample are given in Appendix C). The flexural stress σ_f, is calculated using formula (5), the maximum of which is achieved right before the snapping of the sample. (4) $$E_{\mathrm{f}}=\frac{L^3\mathrm{*}m}{{4\mathit{*b*d}}^3}$$(4) (5) $${\sigma }_f=\frac{3\text{*}F\text{*}L}{2\text{*}b\text{*}{d}^2}$$(5) where L is the support span, 60 mm. m is the slope of the straight-line portion of the load/displacement graph (N/mm). b is the width of sample (mm). d is the thickness of sample (mm). F is the load at a given point on the load/displacement graph (N).

Figure 10. Load/displacement diagram of all four filaments – 3-point-bending test.

The mean value with the maximum and minimum values of the three experiments of flexural elastic modulus (E_f) for all filaments is given in Figure 11. It can be seen that the trend is similar to the tension test, as composites become more resistant to deformation in the elastic range, hence higher E_f, compared to pure PLA. The insertion of fillers interacts with the PLA matrix, increasing the adhesion forces between the matrix layers. E_f of PLA_PCB_0.15 is again swiftly increased, compared to the other composites.

Figure 11. Flexural elastic modulus derived from 3-point-bending tests of all four filaments.

Flexural stress (σ_f) of composites is increased significantly as filler concentration increases, following the trend of E_f, as seen in Figure 12 (Wang et al., 2019). Furthermore, the rectilinear 45° angle of printing of the samples and the potential parallel orientation of the filler’s glass fibres to the tension force could affect σ_f. The mean results are given in Table 5.

Figure 12. Flexural stress derived from 3-point-bending tests of all four filaments.

Table 5. Mean E_f and σ_f of pure and composite filaments – 3-point-bending tests.

Sample	Ef (GPa)	σf (MPa)
PLA Pure	3.1	91.73
PLA PCB 0.05	3.38	92.18
PLA PCB 0.1	3.54	94.59
PLA PCB 0.15	4.11	104.68

4. Conclusions

PCB is one of the key components of WEEE and its recycling can have significant results in the production of added-value materials with enhanced mechanical properties that serve circular economy. This study tested the influence of 5, 10, and 15% PCB precursor (glass fiber-enhanced epoxy resin) filler addition to PLA filaments on the morphological, thermal, rheological, and mechanical properties of final composite MEX 3D printed samples.

SEM analysis confirmed the abundance of C in the polymeric epoxy resin network and the glass fibres within the resin. DSC analysis showed that T_g and T_m were not greatly influenced by the addition of PCB, heat flow was decreased with decreased PLA composition and finally, the addition of 15% filler, led to the more compact chain packing of the crystal lattice and a decreased T_c. Rheological analysis showed that pure PLA presented lower viscosity than PLA_PCB composite filaments (100–200 Pas compared to 300–400 Pas), due to filler particle interaction leading to energy dissipation due to the friction between particles. Furthermore, the interactions and the relatively small specific area of the filler particles can lead to a potential increase of the agglomeration potential, also resulting to a higher viscosity, which in this study proved beneficial to the increase of mechanical properties.

Finally, mechanical performance analysis showed that E, E_f, σ_f, and ε were increased compared to PLA (up to 11.2, 24.6, 9.6, and 28.2%, respectively), due to stiffness and crystallinity increase and the improvement of the interactions and adhesion forces between epoxy resin and PLA. σ of PCB composite filaments decreased due to the decreased PLA concentration but presented an increasing trend as the filler percentage increased, confirming the incorporation and the uniform dispersion of the filler in the matrix.

Future work will include the production of composite filaments with higher filler concentration, and the potential addition of critical metals and REE, to further enhance composites’ thermal, rheological, and mechanical properties. Finally, different polymer matrices (acrylonitrile butadiene styrene – ABS, polyethylene terephthalate modified with glycol – PETG, etc.), with the aforementioned fillers will be compared to PLA composites.

The enhanced thermal properties can be advantageous in applications where exposure to high temperatures is a concern, such as electronic devices, automotive, and aerospace applications (Halim et al., 2021). Similarly, enhanced rheological and mechanical properties can make the material suitable for precision MEX 3D printing applications, including intricate prototypes and functional parts and for applications where structural integrity and load-bearing capacity are essential, such as in automotive components, aerospace parts, and structural engineering, respectively, where more durable, high-performance products with extended lifecycle and less frequent replacements are needed (Halim et al., 2021; Mishra and Jagadesh, 2022). Finally, the adoption of a wider range of polymeric matrices in composite production will signify the creation of additional recycling pathways for polymers that are currently being disposed of or landfilled. The plethora of polymeric matrices and filler combinations can lead to a broader range of customized applications, optimized properties, and even to the addition of novel ones. The aforementioned findings of this study can resonate with a wider audience concerned about environmental impact, including policymakers, industry leaders, and consumers who prioritize eco-friendly and sustainable products and practices, while offering practical solutions to pressing environmental challenges.

Abbreviations
MEX	=	material extrusion
3D	=	three dimensional
AM	=	additive manufacturing
CAD	=	computer-aided design
FFF	=	fused filament fabrication
PLA	=	poly(lactic) acid
WEEE	=	waste electric and electronic equipment
PCB	=	printed circuit boards
RM	=	raw material
EOL	=	end of life
SEM	=	scanning electron microscopy
DSC	=	differential scanning calorimetry
EDX	=	energy-dispersive X-ray diffraction
ETC	=	environmental test chamber
ABS	=	acrylonitrile butadiene styrene
PETG	=	polyethylene terephthalate modified with glycol.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was implemented and funded in the framework of the project ‘ΚΜΡ6-0293124 – Development of a Smart System for the Construction of Portable Houses Using the Prosthetic Construction Methodology’, co-financed by the Greek State, the European Union and the European Regional Development Fund (ERDF), in the framework of the Operational Program ‘Central Macedonia’ of the NSRF 2014-2020, Investment Innovation Plans.

Notes on contributors

Vasileios Stratiotou Efstratiadis

Vasileios Stratiotou Efstratiadis is a PhD Candidate in the field of advanced materials in the Physical Metallurgy Laboratory (PML), Department of Mechanical Engineering, School of Engineering, Aristotle University of Thessaloniki (AUTH), Greece. He is also a member of Centre for Research & Development οn Advanced Materials – a Joint initiative between Texas A&M Engineering Experiment Station (TEES) and AUTH. He is an MSc holder in Petroleum Engineering from Imperial College London (First Class Honors and individual project Thesis on ‘Pore Compressibility of Carbonate Rocks’ awarded from BP as best project of the year). Diploma holder in Chemical Engineering from AUTH (First Class Honors), with main interests in sustainability and reduction of CO2 emissions. Co-founder of eggxcel, a venture with a mission to produce highly efficient, thermally stable, sustainable CO2-capturing sorbents from waste eggshells so target industries can become sustainable and reduce CO2 emissions cost.

Apostolos Argyros

Apostolos Argyros is a PhD candidate researching on the development of additive manufacturing techniques and materials for large scale applications in Physical Metallurgy Laboratory (PML), Department of Mechanical Engineering, School of Engineering, Aristotle University of Thessaloniki (AUTH), Greece. He is also a member of the Centre for Research & Development οn Advanced Materials which is a Joint initiative between TEES and AUTH. He is author of 16 journal papers and has contributed to the research work presented in nine international conferences. Diploma holder in Electrical Engineering from Democritus University of Thrace.

Pavlos Efthymiopoulos

Pavlos Efthymiopoulos is a Materials Scientist and Engineer. He holds a PhD in Chemistry from the University of Ioannina and is currently a Postdoctoral Researcher at the Department of Chemistry of the International Hellenic University, Greece. He has participated as a researcher in seven National and European research projects and has taught at the University of Ioannina, University of the Aegean, International Hellenic University, and Democritus University of Thrace. His research interests focus on polymers and composite materials and especially on their applications in Additive Manufacturing and Chemical Technology.

Georgios Maliaris

Georgios Maliaris graduated from the Department of Mechanical Engineering of the Aristotle University of Thessaloniki (AUTh) in 2000. He is currently a Professor in the Department of Chemistry of International Hellenic University, Vice-President of the Department and Director of the Post-Graduate Program ‘Nanotechnology’. Participated in more than 43 research projects originating from European, national or direct industrial funding, coordinating four of them. Member of organizing and scientific committees of 12 national and international conferences. He has published 34 papers on scientific journals and 49 papers on peer reviewed conferences with more than 550 citations according to Scopus (h-index = 13). His research interests involve, among others, additive manufacturing methods and materials, modelling of stochastic lattice structures, CAD/CAM/CAE systems, evaluation of the mechanical response of materials and structures, and the design of experimental devices utilizing state of the art technologies for material testing and characterization.

Nikolaos Michailidis

Nikolaos Michailidis is Professor and Director of the Physical Metallurgy Laboratory (PML), Department of Mechanical Engineering, School of Engineering, Aristotle University of Thessaloniki (AUTH), Greece. He is also Research Professor at Texas A&M Engineering Experiment Station (TEES), President of the Hellenic Metallurgical Society (HMS), Chair of the Centre for Research & Development οn Advanced Materials – a Joint initiative between TEES and AUTH and co-founder of PLiN Nanotechnology S.A. (AUTH’s spin-off). He is Fellow of the International Academy for Production Engineering (CIRP), member of various Scientific Societies and Boards, Chair of the Scientific Committee of EUROMAT 2019 and Member of the Executive Committee of the Federation of European Materials Societies (FEMS). He served as Director of the Interdepartmental Post-Graduate Studies Program: ‘Processes & Technology of Advanced Materials’ – AUTH and as Chair of the Design & Construction Division (School of Mechanical Engineering-AUTH), while he was Visiting Professor at Fraunhofer Institute for Production Technology (IPT), Aachen-Germany.

Appendix A:

Printer and filament settings

Table A.1. Printer and filament settings.

Parameter Printer settings	Value	Parameter Filament settings	Value
Layer height (mm)	0.2	Diameter (mm)	1.75
No. of perimeters	2	Extrusion multiplier	1.1
Seam position	Rear	Bed temperature (^oC)	60
Speed first layer (mm/s)	20	Nozzle first layer temperature (^oC)	215
Speed others (mm/s)	40	Nozzle other layer temperature (^oC)	210
Extruder width (mm)	0.45	Slow down if layer time is below	10 s

Appendix B:

EPGC202 DSC analysis

The protocol used in the DSC analysis shown in Figure B.1 included a heating ramp of 10 °C/min up to 300 °C, and a cooling ramp of 15 °C/min down to 25 °C (two cycles to rule out the thermal history) to inspect the thermal behaviour up to 300 °C.

Figure B.1. EPGC202 powder DSC analysis.

Appendix C:

Sample dimensions

Figure C.1. D638-10 Type IV ASTM dimensions.

Figure C.2. ISO 178 parallelepiped dimensions.

Table C.1. Width of samples (b) and thickness of samples (d).

D638-10 Type IV ASTM (D) Samples (neck)	b (mm)	d (mm)	Parallelepiped (P) Samples	b (mm)	d (mm)
1D Pure PLA	6.02	3.56	1P Pure PLA	10.35	4.4
2D Pure PLA	6.17	3.61	2P Pure PLA	10.2	4.28
3D Pure PLA	5.98	3.39	3P Pure PLA	10.3	4.34
1D PLA_PCB_0.05	6.2	3.65	1P PLA_PCB_0.05	10.14	4.18
2D PLA_PCB_0.05	6.08	3.58	2P PLA_PCB_0.05	10.13	4.18
3D PLA_PCB_0.05	6.07	3.51	3P PLA_PCB_0.05	10.11	4.09
4D PLA_PCB_0.1	6.33	3.59	4P PLA_PCB_0.1	10.23	4.25
5D PLA_PCB_0.1	6.35	3.64	5D PLA_PCB_0.1	10.21	4.23
6D PLA_PCB_0.1	6.16	3.58	6D PLA_PCB_0.1	10.12	4.12
7D PLA_PCB_0.15	6.29	3.63	7P PLA_PCB_0.15	10.27	4.04
8D PLA_PCB_0.15	6.27	3.57	8P PLA_PCB_0.15	10.12	4.13
9D PLA_PCB_0.15	6.27	3.56	9P PLA_PCB_0.15	10.07	4.08