Kinetically inert covalent adaptable networks enable holistic resource recovery from high-performance printed circuit boards

Abstract

The massive accumulation of electronic waste (e-waste) globally demands sustainable strategies for printed circuit boards (PCBs), a goal currently hindered by the pervasive use of irreversibly cured epoxy resins. Here, a kinetically inert covalent adaptable network (CAN) is designed for recyclable high-performance PCB based on gem-dimethyl dithioketal (gTK) epoxy resins. The gem-dimethyl effect significantly enhances the stability of gTK bonds through steric hindrance, kinetically blocking undesirable acidolysis and heat-induced exchange reactions. The fabricated gTKPCB satisfies the demanding commercial specifications, exhibiting robust thermomechanical stability (~ 220 °C), chemical resistance, and highly reliable electrical interconnection under long-term thermal conditions (105 °C, 90 d). Crucially, gTKPCBs undergo oxidation-specific degradation in H₂O₂ solution, enabling the direct, nondestructive recovery of high-purity copper (99.5%) and glass fibers, and the degraded resins as adhesives. Life cycle assessment confirms that this valorization approach significantly mitigates environmental impact compared to traditional metallurgical recovery processes. This work provides a powerful blueprint for sustainable transitions in the electronics industry and offers a holistic solution for e-waste recycling.

The accumulation of electronic waste demands sustainable strategies for printed circuit boards, a goal limited by the use of cured epoxy resins. Here the authors develop a kinetically inert covalent adaptable network designed for recyclable high-performance printed circuit boards based on gem-dimethyl dithioketal epoxy resins.

Introduction

With the rapid expansion of electronic devices, electronic waste (e-waste) has become the fastest-growing hazardous solid waste worldwide¹. Among the 62 million tons of global e-waste in 2022, only ~22.3% was recycled. Printed circuit boards (PCBs) form the core of e-waste and account for 38.4% of the carbon footprint². Epoxy thermosetting composites are specified for PCB manufacturing due to their favorable combination of electrical insulation, high mechanical strength, and robust thermal and chemical resistance^3,4. However, the irreversibly covalent crosslinking in epoxy thermosets leads to difficulties in the degradation and recycling of waste PCBs after the end-of-life^5,6. Life cycle assessment (LCA) of PCBs indicates that copper mining and production of epoxies are energy-intensive with resource consumption and environmental impacts, contributing to 30% abiotic depletion and 84% human toxic potential in the whole process of PCB production⁷. Thus, it is highly desirable to reuse the secondary metals and other materials. Industrial recycling for the waste PCB primarily focuses on high-value metal recovery with mechanical crushing, pyrometallurgy, and hydrometallurgy as the main approaches^8,9, often neglecting the reclamation of the resin matrix. Given that the dielectric matrix constitutes 70 wt% of the total board mass, current methods face challenges including complex waste management and high energy consumption^10–12. The integrated recycling of both copper and polymeric resins represents a major challenge to realizing sustainable reuse of waste PCBs.

Recently, covalent adaptable networks (CANs) relying on dynamic covalent bonds, which can undergo network reorganization or cleavage under external stimuli, have provided a promising tool to enhance the sustainability of epoxy thermosets^13–19. However, although the incorporation of dynamic covalent bonds (DCBs) provides reprocessability and degradability, the thermomechanical and environmental stability of polymer networks are inevitably compromised^20–23. Upon activation of dynamic bonds at high temperatures, the substantial exchange or dissociation of these linkages dramatically increases the plasticity of polymer networks, leading to material failure due to severe creep. Additionally, the susceptibility of certain dynamic bonds (e.g., acetal, imine, ester, or imide) to corrosive solutions (pH < 1 or pH > 10), which are required for PCB manufacturing, can trigger premature network degradation^{15–17,24–29}. These limitations severely impair the long-term reliability of PCBs, particularly in critical applications for CPU and GPU³⁰, where materials should withstand long-term exposure to elevated temperature as well (>85 °C).

The macroscopic stability of CANs is intrinsically governed by the chemical nature (thermodynamics and kinetics) of the DCBs^31,32. The thermodynamic lability that causes bond cleavage or rearrangement serves as the driving force for material recycling and degradation, yet simultaneously underlies structural instability. While thermodynamic principles suggest an inherent contradiction between network stability and dynamicity^33–35, kinetic controls, including catalyst design and tuning the availability of reactive groups, offer complementary approaches to regulate bond exchange rates and network relaxation behaviors^36–38. Nevertheless, these kinetically inhibited approaches exhibit limited improvements in high-temperature mechanical stability, because the Arrhenius-type temperature dependence of reaction kinetics causes exponential acceleration of bond exchange rates, resulting in a significant increase of network dynamics³⁹. Therefore, achieving an optimal trade-off between network stability (thermomechanical/chemical) and controllable degradation represents a fundamental design challenge for advanced CANs, where precise molecular engineering is required to reconcile these inherently conflicting material properties.

Herein, we report the kinetically inert CAN using gem-dimethyl-functionalized dithioketal epoxies (gTKEPs) to resolve the fundamental stability-degradability dilemma for sustainable high-performance PCBs (Fig. 1). The sterically hindered gem-dimethyl groups in dithioketal units serve as molecular locks, raising bond exchange energy barriers to kinetically block the undesirable bond exchange at elevated temperature and simultaneously enhance the stability in acidic/alkaline environments⁴⁰. Additionally, the gTK bonds exhibit selective cleavage of C-S bonds in response to H₂O₂, enabling specific degradation under eco-friendly conditions. The gTK printed circuit boards (gTKPCBs) manufactured with standard protocols could meet the industrial specifications and demonstrate high-temperature stability. Harnessing H₂O₂ as a green processing agent eliminates complex downstream separation for degraded products. The degraded resins are suitable for use as adhesives, and importantly, the copper can be recovered directly with high purity from the PCBs. The gTK crosslinked networks behave as quasi-static structures at elevated temperatures by decoupling thermal-mechanical robustness from oxidation-specific degradation, offering a holistic approach to resource recovery for sustainable high-performance PCBs.

Fig. 1. Schematic representation of the lifecycle of traditional PCBs and sustainable gTKPCBs from production to end-of-life.

a The conventional waste PCBs are recycled by hydrometallurgy and pyrometallurgy methods involving complex processes and high pollution, where the values of resins and glass fibers are neglected. b Manufacturing of gTKPCBs through general industrial procedures, including prepreg preparation, hot pressing to produce CCL, etching, drilling, coating, and welding. The gTKPCB synthesized by DGEBA and the gTK curing agent meets the specifications of commercial PCBs, notably demonstrating high thermal and acid-base stability. The high-reliability gTKPCB could be used for the high-performance processors (e.g., CPU and GPU) operating under prolonged high-temperature environments. The waste gTKPCB can be degraded in H₂O₂ solution for direct recovery of high-purity copper and glass fibers, and the degraded resins can be used as adhesives to achieve holistic resource recovery. gTK = gem-dimethyl dithioketal, DGEBA = diglycidyl ether of bisphenol A, EP = epoxy resin, CCL = copper clad laminate, CPU = central processing unit, GPU = graphics processing unit.

Results

Stability and degradability of gTK bonds

The gem-dimethyl-TK-based aliphatic amine (gTK-1) and aniline (gTK-2) curing agents were scalably synthesized via the direct condensation reaction of cysteamine hydrochloride and 4-aminothiophenol with acetone, respectively (Supplementary Figs. 1–4). The total costs, including raw materials and energy for gTK-1 and gTK-2 curing agents production were estimated as $9.98/kg and $94.63/kg, respectively (Supplementary Tables 1, 2). For comparison, a dithioacetal-containing phenolic analog (TA) was also synthesized (Supplementary Fig. 5). To assess the structural stabilizing effect imparted by the gem-dimethyl motif, we systematically evaluated the chemical and thermal robustness of gTK bonds. Since the plasticity of dynamic polymer networks is primarily governed by the rearrangement of DCBs at high temperature, we first investigated the thermal exchange behavior of gTK bonds. Time-dependent variable-temperature NMR studies revealed that the gTK bonds exhibited remarkable thermal stability, showing no detectable exchange or dissociation after heating for 2 h at the temperature as high as 180 °C (Fig. 2a). By contrast, the TA bonds readily underwent thermal exchange, reaching equilibrium within 30 min at only 80 °C, with an activation energy (E_a) of 104.8 kJ/mol (Fig. 2b and Supplementary Fig. 6). Furthermore, the gTK bonds also demonstrated high acid and base stability. No degradation was observed under alkaline treatment, and less than 1% degradation occurred after treatment for 6 h in excess acid at 55 °C (Fig. 2c, d and Supplementary Figs. 7, 8). Conversely, the TA linkages underwent rapid hydrolysis to thermodynamic equilibrium under the same acidic conditions, and suffered complete hydrolysis after 150 h in an alkaline medium (Supplementary Figs. 9, 10). The degradation activation energy of the gTK bond (E_a,H⁺ = 76.8 kJ/mol) was remarkably higher than that of the TA bond (E_a,H⁺ = 28.9 kJ/mol) (Fig. 2e). Therefore, the steric effect of the gem-dimethyl group substantially increases the reaction energy barrier, kinetically blocking the undesirable exchange reactions and markedly inhibiting acidolysis.

Fig. 2. Dynamic exchange resistance, acid/base stability and oxidative degradation of the gTK.

a No dynamic exchange between gTK-2 and gTK-3 after incubation for 2 h at 180 °C as proved by ¹H NMR spectra (400 MHz, DMSO-d₆). b Dynamic exchange between TA and TA-1 generating TA-2, monitored by time-dependent ¹H NMR spectra (400 MHz, DMSO-d₆). The activation energy (E_a) was obtained by Arrhenius analysis. Acid and base hydrolysis stability for gTK-2 and TA in the presence of c 10 equiv. HCl and d 20 equiv. NaOH at 55 °C. e Arrhenius analysis for these hydrolysis reactions. f Scheme of H₂O₂-mediated oxidation of gTK-2: i) oxidation of the sulfide, ii) C-S bond scission, and iii) further oxidation and hydrolysis. g Time-dependent ¹H NMR spectra (400 MHz, CD₃CN), h oxidation kinetic profiles, and i Arrhenius analysis for the degradation reaction of gTK-2 in the presence of 10 equiv. H₂O₂ at 55 °C. The activation energy of TA oxidation was shown for comparison. DMSO = dimethyl sulfoxide.

Despite the thermal and chemical stability, the gTK bond permitted a selective oxidative cleavage by H₂O₂. The overall degradation reaction of the gTK moiety was found to be rather complex, because the oxidation of sulfides involves an intricate reaction network yielding multiple sulfur species in different oxidation states⁴¹. Initiated by the oxidation of the sulfide to a sulfoxide, subsequent C-S bond scission yielded acetone and a disulfide, the latter of which could be ultimately converted into a sulfonic acid via further oxidation and hydrolysis (Fig. 2f). As evidenced by the kinetic analysis, the oxidative degradation followed a sequential reaction pathway: the starting material, gTK, underwent exponential decay with an activation energy (E_a,H₂O₂) of 78.8 kJ/mol; the sulfoxide intermediate initially rose to a maximum value and then decreased to approach zero; the product acetone was observed to accumulate to a finite concentration in 18 h at 55 °C after an induction period (Figs. 2g, h and Supplementary Fig. 11). For the TA bond, oxidation rates and activation energies similar to the gTK bond were observed in the first oxidation stage (Fig. 2i and Supplementary Fig. 12). However, only trace quantities of sulfoxide intermediates (<3%) were detected, and the aldehyde was generated with a negligible lag time, suggesting a fast cleavage rate for the intermediate. These results imply that the gem-dimethyl group has a limited suppressive effect on the oxidative sensitivity of the gTK bond but rather increases the energy barriers for the hydrolytic cleavage of the C–S bond, which agrees with the results of the above hydrolysis experiments. The overall oxidative cleavage kinetics of gTK bonds were comparable to those of TA bonds, with both 96% degradation in 19 h and 20 h at 55 °C, respectively, suggesting that the initial oxidation might be the rate-limiting step (Fig. 2h and Supplementary Fig. 12h). Collectively, these findings demonstrate that the kinetically inert strategy effectively enhances the structural stabilization without affecting the oxidative degradation performance of gTK moieties, establishing a powerful tool for designing epoxy resins capable of on-demand, environmentally benign degradation.

Synthesis and characterization of gTK-derived epoxy resins

To ensure complete network formation, we investigated the curing kinetics of gTK and TA curing agents via non-isothermal differential scanning calorimetry (DSC) at various heating rates (Fig. 3a and Supplementary Fig. 13). Each epoxy system exhibited a distinct exothermic peak, indicative of the reaction between curing agents and epoxies. As expected, gTK-1 initiated curing at room temperature, while gTK-2 and TA required thermal initiation with correspondingly higher curing activation energies. In addition, infrared (IR) spectra show the complete disappearance of the characteristic epoxy ring peak at 914 cm⁻¹ and the concomitant emergence of a broad hydroxyl peak at approximately 3400 cm⁻¹, corroborating full curing of the resin (Supplementary Fig. 14). All the cured resins showed high gel contents (> 99%), indicating the formation of densely crosslinked networks (Supplementary Fig. 15).

Fig. 3. Thermal and mechanical properties of gTKEP-1, gTKEP-2, TAEP, and HDAEP.

a DSC curves of the gTK-2/DGEBA mixtures at different heating rates. b TGA and c DSC thermal curves, d tensile and e flexural stress-train curves, and f a performance comparison of all epoxies. HDA = hexamethyldiamine.

Next, we evaluated the thermal and mechanical properties of gTK-based epoxy networks, benchmarking them against a conventional aliphatic amine-cured resin (HDAEP). Thermogravimetric analysis (TGA) confirmed that the gTK- and TA-derived epoxy resins possessed good thermal stability. Notably, the aniline-functionalized gTKEP-2 resin had a 5% weight loss temperature (T_d5%) as high as 318 °C, which is significantly higher than those of previously reported recyclable epoxy resins and slightly lower than that of permanently crosslinked HDAEP (Fig. 3b). Additionally, gTKEP-2 exhibited a much higher glass transition temperature (T_g = 126 °C) than gTKEP-1 and TAEP systems, owing to the combination of rigid structure and high crosslinking density (Fig. 3c and Supplementary Table 3). As shown in Fig. 3d, e and Supplementary Fig. 16, gTKEPs demonstrated high mechanical performances comparable to those of conventional EPs. Typically, the gTKEP-2 showed a flexural strength of 129.5 MPa with a flexural modulus of 2.0 GPa and a tensile strength of 79.0 MPa. The high-strength, high-modulus characteristics indicate that this material is well-suited for manufacturing laminates for PCBs. Following a 3-day immersion in eight solvents at room temperature, no apparent fragmentation was observed for gTKEPs, showing good solvent-resistance (Supplementary Fig.17). To make a comprehensive evaluation of gTKEPs, their key mechanical and thermal properties were compared with HDAEP in a radar plot, suggesting that the introduction of gTKs did not compromise the performance of epoxy resins (Fig. 3f and Supplementary Fig. 18).

Thermomechanical and environmental stability of gTKEP

To evaluate whether the kinetically inert gTK bond could improve the structural stability of polymers, the thermomechanical performance and environmental resistance of gTK-2 or TA-crosslinked EPs were comprehensively investigated. As revealed by dynamic mechanical analysis (DMA), gTKEP-2 exhibited a distinguished high-temperature mechanical stability, with the rubbery plateau storage modulus remaining constant up to 220 °C (Fig. 4a, Supplementary Fig. 19 and Supplementary Table 4). The topology freezing transition temperature (T_v) was obviously observed at 154.8 °C for TAEP because of substantial rearrangement of TA bonds at high temperature, whereas the gTKEP showed a similar strain-temperature dependency curve to that of permanently crosslinked EPs (Fig. 4b)^13,27. Stress relaxation experiments revealed that TAEP underwent rapid relaxation at 210 °C with a characteristic relaxation time (τ^∗) of only 30 s. In contrast, the τ^∗ for gTKEP-2 was as long as 2548 s at the same temperature (Fig. 4c and Supplementary Fig. 20), proving the high-temperature network stability of the gTKEP. Similarly, creep performance (Fig. 4d) showed that gTKEP had a substantially lower viscous deformation of 0.3% at 180 °C compared to TAEP (3.9%). Considering their comparable crosslinking densities (Fig. 4a), these results confirm that the kinetically inert effect of gTK hinders the thermal exchange of dynamic bonds, thereby enhancing the thermomechanical stability of the material.

Fig. 4. Thermomechanical and environmental stability of the gTKEP.

a Storage moduli of gTKEP-2 and TAEP along with temperature increase. b Temperature dependence of the thermal expansion of gTKEP-2 and TAEP. c Normalized stress relaxation curves of gTKEP-2 and TAEP at 210 °C. d Creep curves of gTKEP-2 and TAEP at 180 °C with a constant stress of 0.1 MPa. e DMA curves of gTKEP-2 after being treated with 3 M HCl or 3 M NaOH at 55 °C for 3 d. f The crosslink density (V_e) retention and g tensile strength retention of gTKEP-2 and TAEP after HCl, NaOH, and UV (60 °C, 7 d) treatments. The experiment was performed independently for five times. Data are presented as mean ± SD. h Weight retention of gTKEP-2 and TAEP after heating at 150 °C in different solvents for 1 d. i Comparison of storage modulus in plateau region at evaluated temperatures and stability in acid/base of epoxy resins containing the gTK bond and other degradable linkages reported in the literature (see Supplementary Table 5 for more details). NMP = N-methylpyrrolidone; DMF = N,N-dimethylformamide.

Environmental resistance is a critical property essential for epoxy resins used in PCB fabrication, because it provides robust protection against chemical corrosion throughout the manufacturing process and ensures long-term reliability. To evaluate chemical resistance, gTKEP and TAEP were immersed in harsh acidic (3 M HCl) and basic (3 M NaOH) solutions at 55 °C for 3 d. While both materials showed no significant change in mass (Supplementary Fig. 21), the TAEP networks suffered a substantial loss in crosslinking density (V_e). Conversely, the network structure of gTKEP was fully retained, highlighting its chemical resilience (Fig. 4e, f and Supplementary Fig. 22), which was further confirmed by the unchanged tensile strength of gTKEP following UV aging at 60 °C for 7 d (Fig. 4g). The high-temperature solvent resistance of CANs in good solvents is challenging, because the dilution effect shifts the dynamic equilibrium toward dissociation^32,42. Notably, the gTKEP networks displayed remarkable stability in aggressive organic solvents at 150 °C after 1 d, with a gel content as high as 99% in NMP compared to only 65% for TAEP, reaffirming its robust and stable crosslinked structure (Fig. 4h). Overall, the gTKEP exhibited highly competitive environmental and thermomechanical stability compared to other reported degradable epoxy resins (Fig. 4i and Supplementary Table 5)^{25,27,43–49}.

gTKEP-based printed circuit boards

In addition to the chemical resistance, an optimal epoxy resin for PCBs must possess a robust combination of mechanical, thermal, and electrical properties to meet the rigorous demands of modern electronics manufacturing and long-term reliability. The homogeneous resin solution of the gTKEP-2 formulation was hot pressed to produce a copper clad laminate (CCL), featuring copper foil on both sides (Fig. 5a and Supplementary Fig. 23). This highly crosslinked structure endowed the gTKCCL with substantial solvent resistance (Supplementary Fig. 24). Subsequent processing steps, including circuit patterning, etching, and drilling, were performed to yield the final gTKPCB. The peel strength of the gTKCCL was determined to be 1.58 N/mm using the ICP-TM-650 method, showing robust copper bonding (Fig. 5b)¹⁶. The interlaminar shear strength (ILSS) of the composites was 57.7 MPa, indicating high interfacial bonding between the resin matrix and the glass fiber reinforcement to prevent delamination during service (Supplementary Fig. 25). Besides, the glass fiber reinforced gTKEP (GFREP) showed high flexural strength (545.6 MPa) and tensile strength (299.8 MPa), suggesting its good protection for electronic components (Supplementary Fig. 26a, b). The coefficient of thermal expansion (CTE) is a crucial parameter for PCBs to avoid failure caused by thermal stress. The CTE values of our materials were 72.8 (below T_g) and 262.2 ppm/^oC (above T_g) (Supplementary Fig. 26c), which meet the demands for commercial PCB applications. Remarkably, the GFREP exhibited a dielectric constant (Dk) of 3.89 and a loss tangent (Df) of 0.0129 at 1 MHz (Fig. 5c and Supplementary Fig. 26d), both lower than those of the commercial PCB, which are essential for minimizing signal delay and maintaining signal integrity¹⁶.

Fig. 5. Characterization of gTKEP-based printed circuit boards.

a Manufacturing process of the gTKPCB. The E51, gTK-2, and 2-MI were mixed to prepare prepreg, which was then formulated into a CCL through hot pressing. After thermal transfer of a mask, the unprotected copper foil was etched out by FeCl₃/HCl solution, followed by drilling to obtain the final PCB. Finally, the corresponding electronic components were welded to obtain the ZJU flash lamp. b Copper peel strength, c dielectric constant (Dk) and loss tangent (Df), and d water absorption of the GFREP. The experiment was performed independently for 5 times. Data are presented as mean ± SD. e A radar chart showcasing the key performance indicators of the gTKPCB with previously reported recyclable PCBs (Oxime-ester-PCB, PLA-PCB, and Ester-PCB) and commercial PCB specification (SPEC-PCB). f Thermal imaging photos and functional verification of microcontroller-based signaling model after daily cycles for 90 d. 2-MI = 2-methylimidazole.

Furthermore, the volume and surface resistivities of GFREP were measured to be 2.699 × 10⁹ MΩ·cm and 1.059 ×  10⁹ MΩ (Supplementary Fig. 26e), respectively, validating reliable insulation properties to resist electrical current leakage. The hygroscopicity tests, which included a 24 h immersion in water at room temperature and a subsequent 4 h boiling period, yielded a low water absorption rate of just ~0.33%, underscoring the material’s water-resistance and its capacity to maintain performance during long-term usage (Fig. 5d). The volume resistivity and surface resistivity of gTKPCB could still meet the specification of commercial PCB after the hygrothermal treatment, indicating the good electrical stability of PCBs in humidity environments (Supplementary Fig. 26f). In addition, the thermal conductivity of the GFREP was 0.362 W/mK, a value similar to the commonly used PCB (Supplementary Fig. 26g). A radar plot analysis of gTKCCL conclusively validated that its performance profile not only exceeded the benchmarks established in the literature, particularly in T_g, hydrolysis resistance, volume resistivity, and flexural strength, but also comfortably satisfied the demanding performance criteria for commercial PCB applications (Fig. 5e, Supplementary Fig. 26h and Supplementary Table 6). To accelerate the simulation of the long-term thermal operating conditions of CPU and GPU, the fabricated microcontroller-based signaling model with sequential flashing in the word ZJU was placed in a high-temperature environment for reliability assessment. After daily cycles (8 h heating at 105 °C followed by 30 min operation) for 90 d, the device would run reliably, indicating that the epoxy matrix of the gTKPCB had satisfactory thermal durability (Fig. 5f).

Degradation and recyclability of the gTKPCB

The gTK bond reliably provides the network stability of epoxy resins while maintaining the oxidation-specific degradability under mild conditions. After immersion of the gTKPCB in 20 mL of H₂O₂/THF solution at 55 °C, surface-oxidized copper foil was directly separated from the system at day 1 (Fig. 6a). Following a simple washing step with 0.1 M HCl, high-purity copper (99.5%) was recovered with the same structure as the original one in a good recycling efficiency of 72.3% (Fig. 6b, c and Supplementary Fig. 27). This straightforward copper recovery benefits from the specific, mild degradation of gTK bonds by H₂O₂, which minimizes damage to the copper circuitry. Subsequent full degradation of the gTKEP composite within 3 d yielded recycled glass fibers and a solution of degraded fragments. The recycled fibers were confirmed to be free of residual resins, exhibiting an identical structure to that of the original fibers (Supplementary Fig. 28). Besides, the degraded products of gTKEP were identified as oligomers containing polar groups, such as hydroxyls and sulfonic acids, which were directly repurposed as an adhesive, demonstrating good bonding strength on various substrates (Fig. 6d and Supplementary Fig. 29). Collectively, these results demonstrate that full resource recovery of the PCB has been successfully achieved through the gTK-based oxidation-specific degradation strategy.

Fig. 6. Degradation and recyclability of the gTKPCB.

a Scheme of the gTKPCB degradation at 55 °C in a 20 mL 3 M H₂O₂/THF (7:3, v/v) solution, where the recycled copper was separated at day 1, and the recycled glass fibers and resins were harvested at day 3. b Elemental analysis and c X-ray diffraction analysis (XRD) patterns of the original and recycled copper. d Lap shear strength of recycled resins for gluing different substrates. The experiment was performed independently for at least four times. Data are presented as mean ± SD. e Environmental impact of different sections in the process of gTKPCB degradation. f LCA of this work compared with traditional pyrometallurgical and hydrometallurgical processes. THF = tetrahydrofuran; GFs = glass fibers.

Life cycle assessment (LCA) of the gTKPCB was performed in two stages: gTKPCB production and gTKPCB degradation to recycle copper. Most of the environmental impact in gTKPCB production came from electricity energy consumption and copper foil production, indicating that the environmental impact of gTK-2 production was small (Supplementary Figs. 30, 31 and Supplementary Tables 7, 8). Compared with the industrial PCB production, the gTKPCB showed a comparable environmental impact due to their similar compositions and manufacturing methods (Supplementary Fig. 32 and Supplementary Tables 9, 10). In the recycling process of the gTKPCB, the three types of materials were reused, and the resulting environmental benefits, particularly those attributed to end-of-life allocation and crediting of recycled products, significantly surpassed the environmental impact associated with the reagents and energy consumed (Fig. 6e, Supplementary Fig. 33 and Supplementary Table 11). Finally, we compared the impacts of our recovery method with pyrometallurgy and hydrometallurgy with respect to ecosystem security, human health, and resource depletion (Supplementary Fig. 33 and Supplementary Table 12). All the categories of environmental impact showed negative values, suggesting the direct copper recovery approach could generate environmental profit (Fig. 6f).

Discussion

This study introduces a sustainable and highly reliable PCB manufactured from gTK-cured epoxy resin composites that demonstrates sufficient thermal stability, acid/base chemical resistance, mechanical properties, and long-term durability to meet commercial PCB specifications, making it well-suited for rigorous applications with high-performance processors (e.g., CPU and GPU). The gTK synthesis features high scalability owing to its commercially available raw materials, mild room-temperature conditions, and chromatography-free industrial purification. Additionally, the subsequent laminate fabrication utilizes standard industrial processes under moderate curing conditions, ensuring seamless technology transfer for large-scale manufacturing. Through degradation in a mild, eco-friendly H₂O₂ solution, the process directly yields high-purity copper foil (99.5%) with a recovery efficiency of 72.3% and intact glass fibers, while the degraded resin can be repurposed as an adhesive, thereby achieving holistic resource recovery of PCBs. In contrast to traditional pyrometallurgy and hydrometallurgy, this approach simplifies the recycling protocols and enables the reclamation of both resins and glass fibers, resulting in a minimized environmental footprint as substantiated by life cycle assessment. Ultimately, this work provides a cleaner, sustainable alternative for addressing pivotal environmental challenges in e-waste management.

Methods

Scalable synthesis of 4,4’-(propane-2,2-diylbis(sulfanediyl))dianiline (gTK-2)

4-Aminothiophenol (1000 g), acetone (590 mL, 2 equiv.), and conc. hydrochloric acid (800 mL, 1.2 equiv.) were fully mixed and reacted for 4 h at room temperature. Then the product was poured into NaOH solution (3.0 M, 1600 mL), and washed with water to obtain a yellow solid powder (94.5% yield, m.p. = 134.9 °C). ¹H NMR (400 MHz, CDCl₃) δ 7.38 (d, J = 8.5 Hz, 4H), 6.62 (d, J = 8.5 Hz, 4H), 3.79 (s, 4H), 1.42 (s, 6H). ¹³C NMR (100 MHz, CDCl₃) δ 147.49, 138.72, 120.38, 114.89, 59.11, 30.33.

Preparation of glass fiber-reinforced epoxy resin composite (GFREP)

The composite was composed of glass fiber cloth and gTK-2 epoxy resin, and was fabricated through the hand lay-up technique. After E51 (20 g), gTK-2 (7.40 g), 2-methylimidazole (0.2 g) and acetone (20 mL) were fully mixed, the glue was poured on the glass fiber cloth, and dried at 80 °C to obtain the prepreg with a fiber-to-resin ratio of 45:55. The thickness of the composite depended on the number of layers of laminated glass fiber prepreg. Five layers of prepregs were placed in a hot press at 160 °C for 2 h at a pressure of 5 MPa to obtain the GFREP with a thickness of 1 mm.

Preparation of gTK copper clad laminate (gTKCCL)

The prepregs were produced according to the GFREP method. Two 70 μm thick copper foils were placed on both sides of prepregs of four layers. Subsequently, the laminates were placed in a hot press at 120 °C with a pressure of 3 MPa for 1 h, and 160 °C with a pressure of 3 MPa for 2 h.

Preparation of gTK printing circuit board (gTKPCB)

The circuit was fabricated by a chemical etching process. The copper on the laminate was coated with a mask by a heat transfer printing method, and then the lamination was immersed in 40% ferric chloride with HCl at room temperature for 20 min. After all the unprotected copper was etched, the mask was removed, and the circuit board was washed with distilled water. Finally, the laminate was drilled for the installation of electronic components.

Degradation experiments

After gTKPCB (5.50 g) was immersed in 20 mL 3 M H₂O₂/THF solution at 55 °C for 1 d, the copper foil was isolated from the solution and washed with 5 mL 0.1 M HCl solution to remove surface oxides. The degradation was continued by incubating in fresh 20 mL 3 M H₂O₂/THF. The reclaimed glass fibers were subsequently separated after 2 d, washed with ethanol and dried at 80 °C for 1 h. Finally, THF was removed from the degradation solution (40 mL in total) by rotary evaporation, and the degraded resins were isolated by filtration and washed with water. The copper content was tested by inductively coupled plasma-mass spectrometry. The temperature and pH of solution should not be too high to avoid rapid decomposition of H₂O₂, which would affect the reaction efficiency.

Author contributions

H.H., W.L., and C.G. conceived the idea. H.H. designed and carried out the experiments and analyzed the data. T.S., X.C., Y.K., L.H., and Z.J. assisted with discussion of experimental data. H.H., W.L., and C.G. wrote the manuscript. W.L. and C.G. supervised the project. All authors discussed the results and commented on the manuscript.

Peer review

Peer review information

Nature Communications thanks Xinxing Zhang and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Funding

This study is financially supported by the Joint Fund of National Natural Science Foundation of China (U22A20155), the National Natural Science Foundation of China (52573142, 52303210), the State Key Laboratory of Transvascular Implantation Devices (012024004), the 111 Project (B16042), and the Overseas Academician Joint Lab for Supramolecular Polymeric Biomaterials, Zhejiang University.

Data availability

Detailed characterization and all proton and carbon NMR spectra are available in the Supplementary Information. Source data for the experimental results are provided. All data are available from the corresponding author upon request. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77223-7.