Waste printed circuit boards secondary resource
Waste printed circuit boards are end-of-life conductive circuit layers on insulating substrates, often retaining mounted electronic components and solders. They are a heterogeneous secondary feed generated by equipment use and disposal, distinct from whole electronic waste, individual removed tantalum capacitors, elemental copper and a mineral species. The curated studies concern waste-computer boards, recycler-supplied mixed boards and ground mobile-phone board material; their compositions and outcomes do not define every board type. No criticality designation is inherited from constituent metals.
Microbial & Biochemical Mechanisms (3)
Culture-associated copper mobilization from waste-computer boards
bioleaching
Microbial
An Acidithiobacillus ferrooxidans inoculum was associated with greater seven-day filtered copper concentration than uninoculated acidic medium in a laboratory board-fragment assay. Ferric-mediated oxidation and differences in copper precipitation are plausible contributors. The experiment does not isolate their relative contributions, establish a board-wide extraction percentage or produce isolated copper metal.
Experimental Context & Analytical Limits:
An (2021), Methods 2.1–2.2, used A. ferrooxidans reported as KCTC4516, precultured for seven days in iron-containing 9K medium. Waste-computer boards were crushed to pieces below 1 cm2, washed with deionized water and oven-dried. Each treatment used 5 g fragments and a reported 50 mL liquid in a 50 mL conical tube: deionized water, iron-free 0K medium, or 0K with 1 mL culture concentrate. The estimated inoculated density was 1.4 × 10^6 cells/mL. Treatments ran in triplicate for seven days at 30 degrees C in a shaking incubator; 180 rpm is specified for preculture but not separately for board incubation. Section 3.2 describes 0K adjusted to pH 2, without a board-assay pH time course. Samples passed a 0.2 micrometre membrane before ICP-OES of Fe, Cu, Ni and Pb; retained solids underwent XRD and SEM-EDS. The reported filled volume leaves headspace and oxygen transfer unclear. Culture washing, residual preculture iron and an equivalent sterile spent-culture addition are not specified, so the board cannot be established as the sole iron input. The paper supplies no measured initial bulk copper inventory for an extraction denominator.
An (2021), Methods 2.1–2.2, used A. ferrooxidans reported as KCTC4516, precultured for seven days in iron-containing 9K medium. Waste-computer boards were crushed to pieces below 1 cm2, washed with deionized water and oven-dried. Each treatment used 5 g fragments and a reported 50 mL liquid in a 50 mL conical tube: deionized water, iron-free 0K medium, or 0K with 1 mL culture concentrate. The estimated inoculated density was 1.4 × 10^6 cells/mL. Treatments ran in triplicate for seven days at 30 degrees C in a shaking incubator; 180 rpm is specified for preculture but not separately for board incubation. Section 3.2 describes 0K adjusted to pH 2, without a board-assay pH time course. Samples passed a 0.2 micrometre membrane before ICP-OES of Fe, Cu, Ni and Pb; retained solids underwent XRD and SEM-EDS. The reported filled volume leaves headspace and oxygen transfer unclear. Culture washing, residual preculture iron and an equivalent sterile spent-culture addition are not specified, so the board cannot be established as the sole iron input. The paper supplies no measured initial bulk copper inventory for an extraction denominator.
Curated Evidence Points:
Support
Source:
an2021
Section 3.2 and Figure 2a report filtered Cu of 1047 mg/L in the inoculated arm and 413 mg/L in 0K alone at day seven; the day-one values were 18 and 41 mg/L respectively. These are source-reported liquid concentrations, not fitted optima or isolated-product yields. Figure 2 has standard deviations for three runs. Acidic medium also releases metals without inoculation.
Partial
Source:
an2021
Figure 1 demonstrates growth, Fe(II) loss and increased redox potential in the separate iron-rich preculture. These observations do not directly measure iron redox cycling in the PCB incubation. Section 3.2 describes an increase in filtrate Fe with inoculation, whereas Figure 2b generally places the inoculated series below the uninoculated 0K series; that disagreement prevents a quantitative Fe-enhancement claim. Figure 2c and the text show substantial Ni release in both acidic arms, with somewhat different plotted concentrations rather than proof of identical outcomes or exclusive microbial Ni leaching.
Partial
Source:
an2021
Figure 3 assigns libethenite peaks to the uninoculated 0K filter residue but does not assign that phase in the inoculated residue. This supports a precipitation-related qualification of filtered Cu, not absence of all Cu-containing solids. Figures 2d and 4 show lower late-stage filtrate Pb and Pb-bearing inoculated residue; SEM-EDS does not establish a unique Pb compound or iron-oxide phase. Section 3.3 misidentifies several Figure 4 panel letters, so the displayed panel labels and caption were checked directly. Neither bacterial Pb tolerance nor universal precipitation suppression is established by these observations.
Bioreactor regeneration of ferric lixiviant for a separate PCB column
bioleaching
Microbial
An immobilized living acidophilic community regenerated oxidizing leach liquor in a flooded packed-bed bioreactor connected to a column of waste boards. The combined laboratory system maintained high redox potential after startup and mobilized copper. Its higher extraction than a separate unconnected-column run is confounded by different ferric concentration, feed and liquid inventory, so the difference cannot be assigned solely to microbial regeneration.
Experimental Context & Analytical Limits:
Iglesias-Gonzalez et al. (2022), Methods 2.1–2.4, used recycler-supplied boards from RECILEC in Seville, cut to a mean area of 4.4 cm2 without pulverization or coating-removal pretreatment. A consortium originally from Rio Tinto mine drainage, described as mainly Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans with Acidiphilium-related heterotrophs, was immobilized on 6–8 mm siliceous support particles. Individual strains and member-specific contributions are unresolved. The connected run used 100 g boards, 736 mL total liquid, initial 20 g/L Fe(III), approximately 0.8 L/h circulation and 750 mL/h air supplied to the bioreactor. Whole-system loading was 13.6% solid/liquid; the reported 65.4% refers to the packed column, not the whole liquid inventory. The unconnected run specifies room temperature; no numerical connected-system temperature is supplied. Medium and replacement solutions were adjusted to pH 1.25 and sulfuric acid was added during operation. Section 3.2 says pH was kept below 2, but Figure 5 includes values above 2; a constant below-2 condition is not assumed. Four liquor purges/replacements and acidified evaporation replacement occurred. Dissolved metals were measured by atomic absorption, Fe(II) by dichromate titration, and final solid Cu by aqua-regia digestion after acid washing. The initial Cu inventory was reconstructed from dissolved, wash and residual fractions rather than an independent starting-feed measurement. Replicated column runs and uncertainty estimates are not specified.
Iglesias-Gonzalez et al. (2022), Methods 2.1–2.4, used recycler-supplied boards from RECILEC in Seville, cut to a mean area of 4.4 cm2 without pulverization or coating-removal pretreatment. A consortium originally from Rio Tinto mine drainage, described as mainly Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans with Acidiphilium-related heterotrophs, was immobilized on 6–8 mm siliceous support particles. Individual strains and member-specific contributions are unresolved. The connected run used 100 g boards, 736 mL total liquid, initial 20 g/L Fe(III), approximately 0.8 L/h circulation and 750 mL/h air supplied to the bioreactor. Whole-system loading was 13.6% solid/liquid; the reported 65.4% refers to the packed column, not the whole liquid inventory. The unconnected run specifies room temperature; no numerical connected-system temperature is supplied. Medium and replacement solutions were adjusted to pH 1.25 and sulfuric acid was added during operation. Section 3.2 says pH was kept below 2, but Figure 5 includes values above 2; a constant below-2 condition is not assumed. Four liquor purges/replacements and acidified evaporation replacement occurred. Dissolved metals were measured by atomic absorption, Fe(II) by dichromate titration, and final solid Cu by aqua-regia digestion after acid washing. The initial Cu inventory was reconstructed from dissolved, wash and residual fractions rather than an independent starting-feed measurement. Replicated column runs and uncertainty estimates are not specified.
Curated Evidence Points:
Support
Source:
iglesias2022
Sections 3.2–4 and Figures 5–6 report 90% copper dissolution after 25 days in the connected 100 g run, using a reconstructed initial Cu grade of approximately 30%. Figure 5 shows the bioreactor outlet approaching about 700 mV versus Ag/AgCl after an initial drop and rise, not constant potential from time zero. The result concerns solubilization into circulating and purged liquors, not a recovered copper-metal product. Fraction accounting and the changing solution inventory matter to its denominator.
Partial
Source:
iglesias2022
The unconnected comparison used 124 g boards, 2 L liquor, three successive stored biogenic-ferric batches initially 12.7, 13.1 and 13.8 g/L, and approximately 25% reconstructed Cu grade; it reached 62.2% Cu dissolution after 25 days. It is not a matched sterile control for the connected 100 g, 736 mL, 20 g/L Fe(III) run. Both systems used biologically produced lixiviant and sterilization of the stored liquor is not demonstrated. Sections 2.2 and 3.1 describe no observed metal solubilization in preliminary acid-only tests at pH 1.25, but supply no quantitative detection limits or equivalent long-term column control.
Partial
Source:
iglesias2022
Figures 5–6 support sustained operation under this particular feed and culture regime; they do not prove universal resistance to PCB toxicity, complete separation of all cells from boards, or which community member regenerated iron. Constant outlet potential is an operational proxy rather than a species-specific activity assay. The slower Cu-release interval coincided with higher dissolved Cu, and changing liquor also changed chemistry. Claims of zero reagent consumption are not adopted because acid, fresh ferric liquor and liquid replacements were used. Solvent extraction and electrowinning are proposed downstream operations, not demonstrated isolated-product recovery or industrial deployment in this study.
Ferric regeneration with mobile-phone PCB leach liquor in separate laboratory stages
bioleaching
Microbial
A living acidophilic community regenerated ferric lixiviant from liquor produced by chemical ferric leaching of ground end-of-life mobile-phone board material. Leaching, biooxidation and solvent-extraction stages were tested, but downstream electrowinning used a synthetic solution. The work supports stage-specific laboratory feasibility rather than a demonstrated continuously closed loop or copper-metal recovery from the actual boards.
Experimental Context & Analytical Limits:
Iglesias-González et al. (2024), Methods 2.1–2.2, used end-of-life mobile phones after removing casing and battery, milled to a stated target of 800–2000 micrometres and subsampled with a rotary divider. The paired dataset README instead says 2–7 mm, leaving particle-size provenance unresolved. Nine starting-feed Cu assays averaged 35.71 wt% with standard deviation 4.16 percentage points. Continuous chemical leaching ran for 96 hours at 60 degrees C and 800 rpm, with reported six-hour hydraulic retention time, approximately 135 mL/h liquid feed, 40 g/L biogenic Fe(III), pH 1 and 5 g/L Cu added as commercial copper sulfate to simulate recycle. Effluent was collected for subsequent filtration and biooxidation. The flooded packed-bed bioreactor contained a living Riotinto-mine community reported mainly as Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans; strains and member-specific roles were not resolved. Reported inoculum density was 9 × 10^7 cells/mL; 7 mm siliceous support, 0.55 L working volume, 31 degrees C, 100 mL/h liquid and 750 mL/min air were used, with liquid flow reduced to 37 mL/h at the last reported endpoint. Metals were analyzed by atomic absorption; the dataset README describes dichromate titration for Fe(II) in batch leaching. Matched sterile biooxidation columns and replicated process runs are not reported. No community CURIE, industrial operation or isolated actual-board copper product is asserted.
Iglesias-González et al. (2024), Methods 2.1–2.2, used end-of-life mobile phones after removing casing and battery, milled to a stated target of 800–2000 micrometres and subsampled with a rotary divider. The paired dataset README instead says 2–7 mm, leaving particle-size provenance unresolved. Nine starting-feed Cu assays averaged 35.71 wt% with standard deviation 4.16 percentage points. Continuous chemical leaching ran for 96 hours at 60 degrees C and 800 rpm, with reported six-hour hydraulic retention time, approximately 135 mL/h liquid feed, 40 g/L biogenic Fe(III), pH 1 and 5 g/L Cu added as commercial copper sulfate to simulate recycle. Effluent was collected for subsequent filtration and biooxidation. The flooded packed-bed bioreactor contained a living Riotinto-mine community reported mainly as Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans; strains and member-specific roles were not resolved. Reported inoculum density was 9 × 10^7 cells/mL; 7 mm siliceous support, 0.55 L working volume, 31 degrees C, 100 mL/h liquid and 750 mL/min air were used, with liquid flow reduced to 37 mL/h at the last reported endpoint. Metals were analyzed by atomic absorption; the dataset README describes dichromate titration for Fe(II) in batch leaching. Matched sterile biooxidation columns and replicated process runs are not reported. No community CURIE, industrial operation or isolated actual-board copper product is asserted.
Curated Evidence Points:
Support
Source:
iglesias2024
Results 3.3 and Table 6 report inlet Fe(II) near 4.6–4.8 g/L with 6.8 g/L Cu and outlet Fe(II) of 0, 0.04 and 0.05 g/L at 24, 45 and 67 hours. At 95 hours, inlet Fe(II) 9.29 g/L and Cu 10.45 g/L accompanied outlet Fe(II) 2.37 g/L; outlet Fe(II) was reported as zero at 168 hours only after reducing liquid flow to 37 mL/h. These source-reported assay values support conditional operation of the living column, not absence of inhibition at every loading or exclusive causal activity of either named species.
Partial
Source:
iglesias2024
Results 3.2, Figure 4 and Table 5 report high Cu extraction during ferric leaching and an overall estimate of 99% based on feed and residual-solid assays. Table 5 gives 135.01 g feed at 35.71% Cu, 21.93 g filtered solid at 0.66% Cu and 53 g retained solid at 0.31% Cu. The dataset identifies acid washing of the retained solid before analysis, so the reported estimate is not adopted as an independently closed reactor-only Cu mass balance. Added inlet copper, feed heterogeneity, sampled fractions and possible wash losses matter. Ferric dissolution in the 60-degree reactor is chemical; microbial involvement is regeneration in the separate 31-degree column.
Partial
Source:
iglesias2024dataset
Data.xlsx, composition B21 and D23:D24, records 1.5% HCl washing of retained reactor solid and a mass change from 62.03 to 53 g. The README particle-size range differs from Methods 2.1. Batch C2 labels 1.2 L whereas the paper describes a 1 L batch; model settings are not treated as observed operating conditions. The biooxidation sheet labels liquid flow in g/h while the paper uses mL/h, and its 2023 dates include a 16/02/24 row despite a README collection date in 2022. These unresolved discrepancies preclude density conversion, invented analysis dates and automatic native concentration imports.
Partial
Source:
iglesias2024
Results 3.4 and Figure 5 describe actual biooxidized liquor contacted with 15 vol% LIX 973 NS-LV in kerosene at 20 degrees C; the reported 91.4% Cu transfer at equal phase volumes refers to solvent extraction, with organic concentrations calculated from aqueous analyses. Methods 2.2 and Results 3.5 explicitly identify electrowinning feed as synthetic 15 g/L Cu plus 155 g/L sulfuric acid, despite Table 4 labeling its origin AP SX and listing 150 g/L acid. Figures 7–8 therefore concern a separate synthetic-feed deposit. Reported 99.2% current efficiency is not product purity or whole-board yield; local SEM-EDS Cu peaks do not establish certified bulk purity. Simulated stripping electrolytes and the synthetic electrowinning test do not demonstrate the proposed closed circuit or zero reagent and effluent claims.
Partial
Source:
iglesias2024dataset
The six-sheet workbook belongs to this same experiment, not an independent replication. EW C16 contains 1.02 under CE (%) whereas the paper reports 99.2%; conversion or reconciliation is not assumed. EW contains an additional 50-degree series whose correspondence to the selected manuscript experiment is unresolved. Batch normalization and fitted trajectories remain distinct from independent feed-denominator measurements. Only contextual prose with locators is curated; no workbook-wide import or reproduced process mass balance is claimed.
Sources & References
-
an2021: An, J. (2021). Characteristics of Metals Leached from Waste Printed Circuit Boards Using Acidithiobacillus ferrooxidans. Minerals 11(2):224. DOI 10.3390/min11020224; published 22 February 2021. Full eight-page article, Methods 2.1–2.2, Results 3.1–3.3 and Figures 1–4 inspected.
https://doi.org/10.3390/min11020224 (accessed 2026-10-11) -
iglesias2022: Iglesias-Gonzalez, Ramirez, Lorenzo-Tallafigo, Romero-García, Mazuelos, Romero and Carranza (2022). Copper recovery from unground printed circuit board by biogenic ferric at high solid/liquid ratio. Minerals Engineering 180:107471. DOI 10.1016/j.mineng.2022.107471; April issue, article states available online 4 March 2022. Full eight-page primary article, Methods 2.1–2.4, Results 3.1–3.2, discussion and Figures 1–6 inspected.
https://doi.org/10.1016/j.mineng.2022.107471 (accessed 2026-10-11) -
iglesias2024: Iglesias-González, N., Dorado, A.D., Ramírez, P. and Mazuelos, A. (2024). A high productivity bioprocess for obtaining metallic copper from printed circuit boards (PCBs). Minerals Engineering 205:108459. DOI 10.1016/j.mineng.2023.108459. Accepted manuscript, Methods 2.1–2.2, Results 3.1–3.5, Tables 2–6 and Figures 2–8 inspected; repository cover year 2003 conflicts with verified 2024 journal metadata.
https://doi.org/10.1016/j.mineng.2023.108459 (accessed 2026-10-11) -
iglesias2024dataset: Iglesias-González, N., Dorado, A.D., Ramírez, P. and Mazuelos, A. (2024). Experimental dataset of a high productivity bioprocess for obtaining metallic copper from printed circuit boards. DOI 10.12795/11441/161670; deposited 25 July 2024. README and six-sheet Data.xlsx acquired; XML cells and formulas parsed, with focused checks of claim-related cells. Paired data for the same experiment.
https://doi.org/10.12795/11441/161670 (accessed 2026-10-11)
Curation History
| Timestamp | Curator | Action | Summary |
|---|---|---|---|
2026-10-11T12:40:45.202303+00:00 |
codex (GPT-6; codex) | CREATE | Created process-defined waste-board identity and two separately scoped primary laboratory leaching mechanisms, preserving contrasting controls, source discrepancies, analytical denominators and scale limitations. |
2026-10-11T13:16:41.074639+00:00 |
codex (GPT-6; codex) | EDIT | Added 2024 stage-specific ferric-regeneration evidence and paired dataset, preserving synthetic electrowinning scope, acid-wash accounting and source discrepancies; earlier mechanisms and provenance unchanged. |
Record source: data/records/waste-printed-circuit-boards.yaml