Ammonia-salt-assisted green hydrometallurgy for CuO nanosynthesis from waste PCBs

The escalating accumulation of electronic waste (e-waste), currently estimated at approximately 45 million tons annually worldwide, causes serious environmental and resource recovery challenges. Printed circuit boards (PCBs), key components of most electronic devices, are particularly valuable due to the presence of precious metals (Au, Ag, Pt, Pd, and Cu), the recovery of which is hindered by the combination of ceramics, polymers, and metals. Metal fraction makes up 25%–30% of total PCB mass, while among all metal elements, copper is present in the highest proportion—most often 60%–70% compared to the metal fraction mass. Copper has recently been designated as a critical raw material by the European Commission. Simultaneously, increasing global demand for copper, alongside limited mining capacity, has led to a growing supply gap. This research responds to both environmental and economic needs by recovering copper from discarded PCBs and converting it into copper oxide (CuO) nanoparticles, a material of high technological relevance—especially in industrial organic synthesis. Mechanical methods were applied to separate polymeric layers from metal-rich components, followed by chemical extraction using ammonium-based salts. Analysis of these extracts showed that copper extraction was more efficient if the polymer/ceramics fraction of PCB is removed by grinding. The present study can report that the PCB types (denoted PCB3) as the coarse fraction leached by (NH 4) 2 C 6 H 6 O 7 and obtaining 29.93 wt% Cu compared to the starting waste mass corresponds to ≈99.8% effective extraction when referenced to a Cu-rich 30 wt% feed. This result is the highest of the published data of ammoniacal-e-waste leaching. Purified extracts were used for CuO nanoparticle synthesis by means of microemulsion synthesis. Characterization by X-ray diffraction (XRD) analysis and scanning electron microscopy (SEM) confirmed the successful synthesis of CuO nanoparticles. These results point to a practical and eco-friendly hydrometallurgical strategy for copper recovery from electronic waste, followed by CuO nanoparticle production. The method addresses both e-waste pollution and the growing copper shortfall, supporting urban mining initiatives, promoting circular resource use, and complementing green transition goals.

Copper is a critical mineral for the global transition toward a low-carbon economy and the development of energy-efficient technologies (López et al., 2023). It is used in the manufacture of electrical components such as semiconductors and hard and flexible printed circuit boards, where it provides essential conductivity. The demand for copper is increasing with the expansion of digitalization in industry and daily life (Hossain et al., 2022a; Hossain et al., 2022b). Major global economies, including the United States, Canada, China, and Japan, have officially recognized it as a critical mineral, while the European Union’s Critical Raw Materials Act designates copper as a strategic resource. For decades, copper ore has been concentrated exclusively through flotation processes, which necessitate the use of large quantities of reagents—including collectors, modifiers, and frothers—that act as potential environmental pollutants, particularly within flotation tailings (Matveeva, 2022; Semushkina, 2022).

Given the non-renewable nature of mineral resources, this research represents a significant effort to maximize the utilization of technogenic raw materials, thereby preserving primary mineral resource reserves (Gümüşsoy et al., 2022; Matveeva, 2022). Current industrial practices for mining or even recovering this metal frequently also rely on pyrometallurgical methods that generate hazardous emissions and significant dust, prompting an urgent shift toward greener hydrometallurgical alternatives (Seif et al., 2023). As primary ore grades decline and mineralogy becomes more complex, hydrometallurgical recovery from secondary sources such as waste printed circuit boards has gained increasing attention. In particular, the selective recovery of copper from complex mixtures using eco-friendly reagents, such as ammonia-based solutions should be a sustainable alternative. Other alternatives are mineral acids, such as hydrochloric acid (HCl), sulfuric acid (H 2 SO 4), and nitric acid (HNO 3) or their combinations with other strong oxidants, such as H 2 O 2, O 2, Fe 3+, and Cl 2. These reagents are highly corrosive and demand highly specialized resistant equipment while requiring extreme attention to protect workers’ health from potential inhalation and dermal exposure risk (Hassan and Saleh, 2022). Another major hurdle in hydrometallurgy is the slow leaching kinetics caused by the passivity of certain minerals such as chalcopyrite (copper iron sulfide), which is the most abundant copper ore mineral (Dakkoune et al., 2023), and can also be the case in copper recovery from secondary raw materials. This problem is often solved by mechanical activation through high-energy milling, a powerful process-intensification strategy, since it modifies the structural and energetic state of the solid by producing lattice defects, crystallite refinement, and amorphization. These changes increase surface free energy and significantly enhance the reactivity of the material (Pereira, 2026). Recent studies show that specific milling durations (e.g., 20 min in a planetary mill) can drastically improve copper recovery by inducing mechanochemical changes that bypass traditional chemical barriers (Tešinský and Balá, 2025). While milling increases the chemical reactivity of particles, the resulting fine particles can cause operational issues forming “dead zones” where the extractant cannot reach the mineral (Golpayegani and Kiaei, 2023). Separation by particle size is therefore essential to maintaining optimal solution percolation through the ore or waste stack (Маринин et al., 2023).

The present study took a radically different approach by separating the thoroughly ground fine particles from the macro pieces of E-waste containing copper. The idea was that if significant copper content is present in fine particles, mechanical activation should ensure, along with the high surface area generated during the milling, higher accessibility, and the efficiency of leaching reagents.

This study is novel in its systematic study on the balance of copper content into two extremely different fractions obtained by the milling (mechanical activation) of waste PCBs, which aims to provide a more efficient and sustainable pathway for copper recovery that minimizes reagent consumption and energy demand (León et al., 2025; Pereira, 2026). Recent studies addressing copper recovery from waste PCBs have relied predominantly on acidic/oxidative leach systems, such as electrochemical dismantling in H 2 SO 4 –CuSO 4 –NaCl, which carry significant corrosion and effluent-handling burdens. Ammonia alternatives remain comparatively rare: Abdelbasir et al. (2023b) reported 96.7% Cu recovery from WPCBs via an NH 4 Cl–NH 3 route at 35 °C and 2 mol L -1, El-Nasr et al. (2019) extended this approach to direct Cu nanoparticle synthesis, and Ravi et al. (2025) prepared ∼50 nm spherical CuO nanospheres from a whole-board pulp. However, none of these studies systematically examined how copper partitions between the coarse and fine fractions generated by mechanical pre-processing or compared different ammonium salt anions (chloride, citrate, and carbonate) against the same set of PCB types. Equally important, the impact of co-leached inorganic fillers calcium and magnesium species originating from epoxy-resin substrates on the morphological and phase purity of the downstream CuO product has not been quantified, and these are precisely the gaps the present study addresses.

We show that copper hydrometallurgy based on PCBs could be the most energy-efficient if the extracts obtained are used directly in the synthesis of CuO nanoparticles—a catalytic material in high demand. In this way, it is possible to bypass the need for intensive purification stages required in traditional metallurgical pathways. The presented, novel integrated approach aligns with the principles of green chemistry and resource efficiency emphasized by a recent review of similar methods (Oke and Potgieter, 2024).

Materials

Waste PCB samples of different origins (Figures 1a–c) were collected from Prizma Comerc Ltd. Banja Luka, a local company dealing with the purchase, processing, and placement of secondary raw materials. The samples were categorized into high, medium, and low grades based on their intrinsic metal value, often described by the expected gold content. PCB1 and PCB2 belong to the second group of printed circuit boards with less gold content, while PCB3 represents the first category of circuit boards with the highest gold content. PCB1 was a Canon Monterey Board (4L Controller) main board from a vintage commercial office photocopier/printer. PCB2 was a desktop PC motherboard manufactured by Fujitsu Siemens (or Foxconn as the OEM builder) for use in their Scenic or Scaleo desktop computer lines. PCB3 was an ECS (Elitegroup Computer Systems) P6VXA motherboard. All three types were deliberately chosen from large international producers as used broadly from the early-to-mid 2000s. Most of the electronic equipment in which they were used were either already retired, about to be expired, or irreparably broken. Their exact compositions are the protected intellectual property of the makers, but the average expected weight percent of all metals within the PCB is cited as 25%–30%, while the individual metallic content of copper within a total metal mass fraction ranged 60%–70% (Li et al., 2007; Tiwary et al., 2017). The present study relies on these approximations since the sample preparation by milling did not produce a fully homogeneous phase, and not even the XRF technique could be applied.

It important to note that while other studies often prioritize high-grade boards because precious metals such as gold, silver, and palladium constitute approximately 90% of their total value (Oke and Potgieter, 2024), the present study focused exclusively on copper, which represents the most significant metal constituent by mass.

For the copper extraction process, three types of ammonium salts were used: ammonium chloride (NH 4 Cl, ≥99.8%, Lach:ner), ammonium citrate ((NH 4) 2 C 6 H 6 O 7), assay 98.0%–103.0%, CARLO ERBA Reagents), and ammonium carbonate ((NH 4) 2 CO 3), p.a. 30%–33% NH 3, Semikem). Hexadecyltrimethylammonium bromide (CTAB, assay ≥97%, MERCK) was utilized to form micelles, which served as a nano-reactor for CuO nanoparticle synthesis. Sodium hydroxide (NaOH, ≥98%, pearls, Lach:ner) was utilized to optimize the pH value for the precipitation of intermediate copper hydroxide forms which were finally calcined into CuO nanoparticles.

Mechanical preparation of e-waste PCBs

The mechanical processing of the PCBs involved the manual removal of the larger polymer components from the boards’ surfaces to prevent the polymer interfering in the later nanoparticle synthesis (Figure 1d).

Following this phase, crushing and grinding were performed utilizing a single-span plate jaw crusher (model: MORSE (2000) 4 × 6 JAW CRUSHER) with a capacity of 0.544 t/h and a lower size limit of 15 mm, and subsequently a laboratory ball mill. The capacity of the mill was 5 kg, the amount of grinding media in the mill was 40%, the number of revolutions per minute was 70, and the operating mode was dry grinding. The milling process lasted for 90 min to ensure a uniform structure of powdered fraction. The final stage of sample preparation was sieving through a 630 μm sieve. Thus, for all three types of the PCBs studied, two radically different fractions were obtained: fine −630 μm fraction and the coarse fraction +630 μm, which included macro pieces of the metal obtained through milling. These fractions will later in the text be depicted as coarse (Figure 1e) and fine (Figure 1f) fraction.

Hydrometallurgical copper ion extraction

The extraction process was performed in shaker bath (model: Witeg Labortechnik GmbH) over 24 h by immersing 5 g of each PCB sample (two fractions for each of three samples) in 150 mL of the following ammonia-based leaching solutions: 2 M NH 4 Cl, 0.5 M (NH 4) 2 C 6 H 6 O 7, and 0.5 M (NH 4) 2 CO 3 as per El-Nasr et al. (2019). The temperature was maintained at 30 °C. Most of the mixtures almost immediately appeared dark blue, indicating that copper was present and forming the tetraamminecopper(II) complex ion [Cu(NH 3) 4 ] 2+.

Following leaching, the resultant solutions were filtrated using Whatman slow flow filtration paper to separate the solid residue from the filtrate. A small portion of the filtrate was subject to standard closed digestion (Water Environment, 2012) in preparation for metal concentration measurement. The copper ion concentration in the filtrates was then determined using inductively coupled plasma optical emission spectroscopy (ICP-OES, model: Optima PerkinElmer) at a wavelength of 327,393 nm. The copper concentration in each filtrate was determined in triplicate injections on a PerkinElmer Optima ICP-OES at the Cu 327.393 nm emission line. The expanded combined measurement uncertainty of the method was estimated according to the EURACHEM/CITAC guide and equaled 3.1% RSD (k = 2, 95% confidence), which is consistent with values reported for PerkinElmer Optima-class instruments for copper determination (Martínez-Peñuñuri et al., 2025). Measurement uncertainty, rather than standard deviation of replicates, was chosen as the principal indicator of data dispersion, in line with the standard practice of analytical screening studies involving heterogeneous solid feedstocks.

Synthesis of copper oxide nanomaterials

The above-described filtrates containing tetraamminecopper(II) complex ion [Cu(NH 3) 4 ] 2+ were used for copper nanoparticle synthesis. The extract volume of 150 mL was diluted to 1,000 mL with distilled water to increase the mass transfer. CTAB served as the surfactant for micelle formation. At the reaction temperature of 37 °C, the critical micellar concentration of CTAB is 1.00 mmol/L (Pejić et al., 2012; DrissiDaoudi et al., 2021). To ensure a high density of micelles while avoiding the formation of undesirable rod-shaped agglomerates, twice the CMC concentration was used: 2 mmol/L. (Patel et al., 2014).

Samples were placed in Erlenmeyer flasks on a magnetic stirrer at 37 °C and 200 rpm. Upon reaching the target temperature, approximately 1 L of 2 M NaOH was added in sequences to optimize the pH for copper hydroxide intermediate formation, which was at pH 12.5. Magnetic stirring continued over 3 days, representing the “ageing” period. The mixture was left to settle 2.5 h for easier decantation.

Subsequently, the samples were filtered (Whatman filter, 450 nm), and the sediment was transferred to Petri dishes and placed in a dryer (model: VIMS ELEKTRIK LSW-53) at 110 °C for 5 h. Following cooling in a desiccator, the samples were crushed with a mortar and pestle, transferred to porcelain crucibles, and annealed in a muffle furnace (model: Nabertherm) in an air atmosphere. The annealing program involved heating at 10 °C/min to 800 °C and being there for 2 h at 800 °C. After cooling, the samples were dehydrated in a desiccator.

Characterization techniques

Fourier transform infrared spectroscopy (FT-IR) analysis was applied to identify the type of resin that served as the carrier for metal components in all three types of PCBs. Synthetic polymers in general are inert to the salts used in the present study. Thus, the identification of the polymer by FTIR will serve as the supplementary material for future studies on polymer recycling after metal extraction. This information is not crucial to the present research but is shown as side information for future studies on the recycling potential of this component of the PCB. Spectra were collected in the 4,000–400 cm -1 range using Shimadzu IRSpirit FT-IR spectrophotometer with a single-reflection ATR accessory (QATR-S).

X-ray diffraction (XRD) was performed to confirm the form in which the copper nanoparticles were synthesized. XRD measurements were performed using a Philips X’Pert PRO diffractometer equipped with a Cu-anode lamp and a filter. XRD data were analyzed by Match! version 4.1 build 311 with COD-Inorg database version 2024.12.08. CuKα was the primary radiation source. From the diffraction pattern obtained in the measurement, Kα2 was subtracted in the program at the beginning of analysis, so the selected wavelength was Kα1 = 1.540598 Å. Diffraction patterns were treated with background subtraction and basic smoothing. Qualitative analysis was performed with help of an automatic search–match option. Quantitative measurements for weight composition were performed with the RIR method. Crystalline size estimations were performed with the Scherrer formula with instrumental broadening correction.

Scanning electron microscopy (SEM) was utilized to examine the surface morphology of synthesized nanomaterials using a Thermo Scientific Apreo 2 model. Voltage was 15 kV, an ETD detector was used, and the pixel dwell time was 3.00 μs.

Energy-dispersive X-ray spectroscopy (EDS) was measured at Ametek’s EDAX. The voltage was 15 kV using an Octane Elect Super Silicon Drift Detector, allowing for the precise elemental mapping and quantification of the elements contained in the synthesized nanomaterials.

All measurements were performed after proper calibration of the instruments.

Optimization of copper extraction

Before reaching out for the ammonia-salt based extraction of the copper contents from the fine and coarse fraction, the fine fraction was analyzed by FTIR. It was expected that this fraction would represent mostly pulverized polymer. This would thus be important backup information for any future reproducibility studies of the present research by this group or other authors.

The results are shown in Figure 2; all three analyzed samples show a match with the FT-IR spectrum of the epoxy resin. This identification confirms that the fine fraction is primarily composed of the non-conductive polymeric matrix inherent in printed circuit boards, thereby validating the mechanical separation process as an efficient precursor for hydrometallurgical treatment. Key characteristic absorption bands were identified, typically corresponding to C–H stretching vibrations (approximately 2,900 cm -1), aromatic ring vibrations (approximately 1,500–1,600 cm -1), and C–O–C ether linkages (approximately 1,100–1,250 cm -1) (Pancoko et al., 2024). In summary, spectral features confirm the presence of the epoxy resin matrix inherent in the most commonly used printed circuit board substrates.

Copper complex formation was successfully confirmed after extraction by observation of the characteristic blue color. ICP-measured concentrations were recalculated to absolute mass of copper present in the whole volume of the extract and compared to the starting mass of ground PCB. This way, the mass percentage of copper in particular PCB fractions were calculated. Data is shown at Figure 3, including the designation of the type of the salt, type of the PCB, and the fraction identification.

The leaching efficiency varied significantly between the fine and coarse fractions, with the former mostly demonstrating a higher recovery rate. The only exception was PCB2 extracted by chloride and citrate, and PCB3 extracted by chloride, where more copper was extracted from the fine fraction. Since some studies had contrasting results, where more copper was extracted from the finer fractions, it could be noted that the PCB type plays an important role. Similar studies claim that the overall leaching yield depends on variations in the internal board architecture and the accessibility of copper pathways to the ammonia-based leaching reagents, as the degree of liberation significantly dictates the accessibility of metallic copper surfaces (Abdelbasir et al., 2023a). In the present study, PCB3, the coarse fraction, leached by (NH 4) 2 C 6 H 6 O 7 and obtaining 29.93 wt% Cu comparing to the starting waste mass, corresponds to ≈99.8% effective extraction when referenced to a Cu-rich 30 wt% feed. This result sits at the top of the published data of ammonia-e-waste leaching. Nevertheless, based on the average extraction results in the present study, the most effective ammonium salt was identified as ammonium chloride. Despite some variations, we focused on extracts obtained by this salt, bearing in mind that data could serve for the project of a single-line configuration of a potential industrial pilot plant. Thus, extracts from the three PCB samples obtained using this salt were subsequently used to synthesize CuO nanoparticles by microemulsion. This selection was based on the fact that citrate is itself a strong chelating ligand that competes with NH 3 for Cu 2+ coordination; it stabilizes soluble Cu–citrate complexes against the hydrolysis step required for CuO nucleation and would leave organic residue in the final oxide product. By contrast, the chloride ion is a drive force in the [Cu(NH 3) 4 ]Cl 2 leach complex, and the well-established NH 4 Cl–NH 3 route to CuO from waste PCBs as demonstrated by Abdelbasir et al. (2023b) is the most appropriate synthesis platform. The carbonate system, although capable of selective Cu recovery giving 99.9% Cu purity at 3.5 V electrolysis voltage with ammonia–carbonate (Rudnik et al., 2014), has been documented as producing basic copper carbonates instead of CuO on standing and is therefore better suited to electro-winning than to oxide precipitation. Finally, the chloride leachate gave the most reproducible feed-Cu values across six PCB/fraction combinations, whereas the citrate leachate spanned 0.09–29.93 wt%—an irreproducible feed that would compromise the potential future industrial pilot study.

Nanoparticle characterization

Scanning electron microscopy (Figure 4) of the synthesized products revealed slightly different structures, although dominated by particle sizes of approximately 30–900 nm.

It was only in the case of the samples synthesized from PCB1 copper extract that the particles were highly uniform, having a somewhat cylindrical shape with particles which appear interconnected or agglomerated into the porous structure.

In the case of particles obtained from PCB2 copper extract, the shape was predominantly spherical with a flat surface but with a wide variety of particles sizes of 30–500 nm.

Samples derived from PCB3 copper extracts displayed more heterogeneous polyhedral morphology with visible surface defects or “furrowed grooves”. In terms of the potential use of these particles as catalytic materials, the observed morphological diversity, especially in the case of PCB1- and PCB3-based products, offered options for further study and defining the methodology for tailoring their surface-to-volume ratios, improving active site availability for heterogeneous catalytic applications. Heterogeneous catalysis demand exactly defined surface architectures and high accessibility of active sites; such properties are directly influenced by the morphological consistency and porosity of the synthesized copper oxides (Abdelbasir et al., 2023b; Dörner et al., 2019; Lignier et al., 2011; Sarkar et al., 2023; Zedan et al., 2022). However, pronounced differences in the morphology of the samples may be attributed to the unique composition of the residual organic matrices and minor metal impurities inherent in the various PCB substrates. The formation of these distinct structures likely reflects the influence of intermolecular forces acting during the precipitation phase, which can be further modulated by the presence of residual hydrophobic components from the original PCB epoxy resins (Alonso et al., 2015).

Furthermore, XRD patterns (Figure 5) were analyzed in order to confirm the crystalline phase and purity of the synthesized products. Diffractograms revealed characteristic reflections corresponding to the monoclinic CuO structure in all three samples. However, the mass percentage differed between them. In particular, the following phases were identified with corresponding percentages.

Nanomaterials from PCB1 (Figure 5a): tenorite (CuO, entry number: 96–901-6327), 98.1% by mass; copper (Cu, entry number: 96–901-3015), 1.4% by mass; and cuprite (Cu 2 O, entry number: 96–900-5770), 0.5% by mass. There are also 14 uncorrelated peaks.

Nanomaterials from PCB2 (Figure 5b): tenorite (CuO, entry number: 96–901-6327), 45.5% by mass; calcite (CaCO 3, entry number: 96–900-9668), 34.6% by mass; portlandite (Ca(OH) 2, entry number: 96–900-0114), 19.2% by mass; and cuprite (Cu 2 O, entry number: 96–100-0064), 0.7% by mass. There are also 12 uncorrelated peaks.

Nanomaterials from PCB3 (Figure 5c): tenorite (CuO, entry number: 96–901-6058), 49.5% by mass; portlandite (Ca(OH) 2, entry number: 96–900-0114), 36.8% by mass; periclase (MgO, entry number: 96–900-6793), 13.5% by mass; and cuprite (Cu 2 O, entry number: 96–100-0064), 0.2% by mass. There are also seven uncorrelated peaks.

There are uncorrelated peaks on all the diffractograms, so the mass percentages mentioned before are subject to a relatively high degree of uncertainty. The origin of the uncertainty lies in the principles of the RIR method, which is often described as quasi-quantitative. As this method has no internal standard, it solely relies on the integral intensities of all the phases. Having uncorrelated peaks in the diffractogram causes similar overestimations as with the samples containing amorphous phases. These arguments indicate that calculated values are systematically higher.

The best results were obtained for sample PCB1 with 98.1% CuO, which is confirmed by the well-developed structure observed under SEM. This confirms this type of waste PCB as potential raw material—a high-purity precursor for the synthesis of high-performance catalytic nanomaterials, although the NaOH concentrations also influence the purity of the product as suggested by earlier studies on the direct synthesis of nano-CuO from p.a. quality chemicals (Sudha et al., 2021). Conversely, the presence of significant calcium and magnesium compounds in PCB2 and PCB3 samples suggests that inorganic flame retardants within the resin matrices impede the formation of high-purity crystalline phases during the precipitation process. These mineral impurities most probably originate from the composite substrates and compete for nucleation sites during the nanomaterial microemulsion synthesis. This process limits the yield of the desired tenorite phase as co-precipitation is occurring (Cocchiara et al., 2019; Ravi et al., 2025). These considerations are based on observations of accessory phases such as calcite and portlandite, which likely stem from inorganic fillers and glass fiber reinforcements inherent in the original epoxy-resin substrates, which co-leach during the extraction process (Ren et al., 2025). To mitigate these secondary phase formations, future optimization of the leaching protocols may require selective filtration or chelating agent adjustments to reduce the concentration of these leached mineral contaminants before the precipitation phase.

Finally, the chemical and elemental composition obtained by the EDS technique shows clear identification of all elements indicated in the XRD. Results are shown in Table 1. In PCB1, the sum of the oxygen and copper give around 79%, probably mostly associated with CuO.

The remaining 21% of the mass in PCB1 corresponds to Ca, Mg, and C, but since some measurement uncertainty is expected, the most thorough analysis would also be expected to reveal trace elements such as Si, Al, and Sn, consistent with the typical metallic composition found in pulverized PCB substrates studied by other researchers working with more aggressive leaching agents (Masavetas et al., 2013). However, in this study, EDS is used strictly for bulk elemental composition and has no sensitivity to long-range atomic order; crystallographic phase identification is therefore the sole domain of XRD. Re-examining the diffractograms of Figure 5 through this lens, gypsum (CaSO 4 ·2H 2 O) and quartz (SiO 2) are not observed as crystalline phases in any of the three samples, although their constituent elements would have been detectable by EDS if present in sufficient concentration. The crystalline Ca- and Mg-bearing phases that XRD actually identifies are calcite (CaCO 3) and portlandite (Ca 2) in PCB2, and portlandite together with periclase in PCB3 operating with RIR weight fractions of 34.6%, 19.2%, and 13.5%, respectively. Si, Al, and trace Sn, which others have reported in PCB-derived residues leached with more aggressive agents, were below the EDS detection limit in our study and are also not observed as crystalline phases by XRD. This indicates that they remain either as finely dispersed amorphous precipitates or are below the practical ≈1 wt% detection limit of powder XRD (Barragán-Mantilla et al., 2023; Stutzman et al., 2016). Si and Sn entrapment within the leached material is mentioned in other studies as a phenomenon that persists through dissolution and precipitation (Vardanyan et al., 2022; Huyen et al., 2016). Consequently, further optimization of the dissolution parameters is necessary to minimize the co-leaching of these inert inorganic components and selectively suppress Ca and Mg, which otherwise nucleate as calcite, portlandite, and periclase and compete with CuO for precipitation sites, thereby enhancing the phase purity of the final CuO product (Kaffash et al., 2024; Vardanyan et al., 2022; Huyen et al., 2016).

This study set out to develop a sustainable, closed-loop methodology for recovering high-purity copper from waste printed circuit boards (PCBs) using non-aggressive ammonia salts and to convert the recovered copper into copper oxide (CuO) nanoparticles via a microemulsion-based synthesis. Copper was chelated selectively as the tetraamminecopper complex ion, so that the dissolution of refractory substrate components was limited during the leaching stage. Three PCB grades were processed through coarse/fine mechanical separation and sequential leaching with three ammonium salts (NH 4 Cl, (NH 4) 2 C 6 H 6 O 7, and (NH 4) 2 CO 3); the most promising extracts were then taken forward to precipitation at pH 12.5, 3-day ageing, and calcination at 800 °C.

The combination of coarse-fraction processing with (NH 4) 2 C 6 H 6 O 7 on PCB3 yielded 29.93 wt% Cu relative to the starting waste mass, corresponding to ≈99.8% effective extraction when referenced to a typical 30 wt% Cu-bearing feed. Across all three PCB grades, ammonium chloride produced the most reproducible feed-Cu values across the six PCB/fraction combinations and was therefore selected as the precursor stream for nanoparticle synthesis.

Quantitative XRD with the RIR method showed that nanoparticles synthesized from the PCB1 extract contained 98.1 wt% monoclinic tenorite, 1.4 wt% metallic Cu, and 0.5 wt% cuprite (Cu 2 O). Nanoparticles from PCB2 contained 45.5 wt% tenorite together with 34.6 wt% calcite, 19.2 wt% portlandite, and 0.7 wt% cuprite. Nanoparticles from PCB3 contained 49.5 wt% tenorite together with 36.8 wt% portlandite, 13.5 wt% periclase, and 0.2 wt% cuprite. CuO purity is therefore strongly dependent on the starting PCB grade: PCB1 yielded a single-phase tenorite product, while PCB2 and PCB3 yielded multi-phase products in which Ca- and Mg-bearing minerals act as co-precipitates.

SEM imaging showed that PCB1-derived particles were uniformly cylindrical and formed interconnected, porous aggregates. PCB2-derived particles were predominantly spherical with sizes spanning ≈30 nm to over 500 nm. PCB3-derived particles exhibited heterogeneous, polyhedral morphology with visible surface grooves. Across all three products, the dominant particle size was around 30 nm, with the upper bound of the size distribution reaching ≈900 nm.

EDS confirmed the presence of Cu, O, Ca, Mg, and C in all three products, with Cu loading of 60.59 wt% in the PCB1-derived material, compared with 32.96 wt% in PCB2 and 34.14 wt% in PCB3. This is consistent with the phase composition inferred from XRD.

Among the three grades studied, PCB1 is the most promising raw material for producing a high-purity, catalytic-grade CuO precursor via the green ammonia route described here. The procedure effectively minimizes energy-intensive purification steps while achieving substantial copper recovery from coarse-ground electronic waste. For PCB2 and PCB3, additional optimization of the leaching protocols including selective chelation or pH-window control of Ca 2+ and Mg 2+ is required to suppress the formation of calcite, portlandite, and periclase and to enhance tenorite purity.

Future work will quantify the gravimetric CuO synthesis yield and establish the full mass balance across the precipitation–ageing–calcination chain in our laboratory and investigate selective suppression of Ca and Mg co-leaching through chelation or pH-window control. Finally, we plan to perform catalytic benchmarks to validate the practical applicability of the final product in industrial organic synthesis.