Spent potlining secondary resource
Exhausted aluminium-smelting cell lining containing variable proportions of carbonaceous cathode material, refractory cement and brick, and accumulated process salts. The accepted study used real mixed-cut waste supplied by Trimet Aluminium, Essen, Germany, with three visually selected grades from refractory-rich to mainly graphitic material. Exact cell and service histories were not resolved. This solid waste is distinct from aluminium dross, waste electrolyte, red mud, graphite as a mineral or commodity, and its constituent elements. Laboratory chemical extraction is documented without a microbial mechanism or an inherited criticality designation.
Microbial & Biochemical Mechanisms (2)
Abiotic mobilization from mixed-cut spent potlining
separation
Sequential alkaline-peroxide and sulfuric-acid treatment mobilized fluoride, aluminium and other constituents from heterogeneous potlining into leach liquors, leaving altered carbonaceous and refractory solids. This is chemical laboratory extraction from real industrial waste; changes in residual carbon percentage are not a carbon recovery yield, and solution release is not isolation of a marketable product.
Experimental Context & Analytical Limits:
Jaw-crushed waste grades A, B and C were sieved and the less-than-1.18 mm and 1.18-9.51 mm fractions tested. A was refractory-rich, B mixed and C mainly graphitic; these are subsamples of the same material study, not pure-phase or independent-study feeds. At approximately 20 degrees C, a typical treatment stirred 2.0 g material in 100 mL NaOH at pH 11, including 10 mL of 30% aqueous H2O2, for 3 h at 200 rpm; NaOH was then increased to 1 M and total volume to 250 mL for another 3 h. Solids were filtered, rinsed and dried at 50 degrees C for at least 24 h. Experimental describes an acid treatment of 2.0 g in 250 mL 0.5 M H2SO4 for 3 h at 200 rpm, whereas Figure 4 specifies 1 M acid and Table S2 gives a 2 h acid stage. These unresolved source differences prevent a single unqualified reproduction recipe. No cells, microbial products or biological treatment arm were used.
Jaw-crushed waste grades A, B and C were sieved and the less-than-1.18 mm and 1.18-9.51 mm fractions tested. A was refractory-rich, B mixed and C mainly graphitic; these are subsamples of the same material study, not pure-phase or independent-study feeds. At approximately 20 degrees C, a typical treatment stirred 2.0 g material in 100 mL NaOH at pH 11, including 10 mL of 30% aqueous H2O2, for 3 h at 200 rpm; NaOH was then increased to 1 M and total volume to 250 mL for another 3 h. Solids were filtered, rinsed and dried at 50 degrees C for at least 24 h. Experimental describes an acid treatment of 2.0 g in 250 mL 0.5 M H2SO4 for 3 h at 200 rpm, whereas Figure 4 specifies 1 M acid and Table S2 gives a 2 h acid stage. These unresolved source differences prevent a single unqualified reproduction recipe. No cells, microbial products or biological treatment arm were used.
Curated Evidence Points:
Support
Source:
robshaw2020
Experimental, Figures 2-4 and Tables 1-3 support heterogeneous feed characterization and chemical mobilization by PXRD, SEM/EDX, elemental analysis, ICP-MS, ion chromatography and fluoride-selective electrode measurements. Figure 4 shows rapid initial fluoride release with additional release during stronger caustic and acid stages. Table 3 values describe material leached during each treatment relative to as-received or already partially leached solids; they are not total feed composition or a closed whole-process balance. Treated refractory-rich A remains unlike the carbon-rich B/C residues.
Support
Source:
robshaw2020_si
Figure S3 shows the actual contrasting waste grades, Figures S5-S22 preserve all six before/caustic/full-treatment diffraction series, and Figures S23-S28 show solid morphology and selected elemental spectra. These support phase heterogeneity and alteration, without quantitative phase recovery. Table S1 is additional stage-specific leachate characterization; Table S3 instead summarizes previously published cell compositions and is not analysis of this feed.
Partial
Source:
robshaw2020
Figure 4 confidence limits come from three electrode measurements, not demonstrated independent leaching replicates. Before/after and sequential-stage comparisons do not isolate each reagent's causal contribution. Cyanate was not detected above baseline after tenfold dilution, and the authors report no available labile cyanide for testing that oxidation step. Cyanide destruction efficiency and complete detoxification are therefore unverified; post-treatment residual-solid leachability testing was future work. Table 2 point EDX percentages are not a bulk purity measurement.
Abiotic resin separation and coelution from real potlining leachates
recovery
A lanthanum-loaded chelating resin retained fluoride and aluminium from combined potlining leachates and released them together into alkaline eluate. The experiment demonstrates aqueous separation and enrichment, not isolated pure cryolite, complete recovery from the original waste, repeated resin-cycle durability or industrial deployment. Proposed aluminium-hydroxyfluoride binding and modeled aqueous speciation remain mechanistic interpretations rather than direct new structural proof.
Experimental Context & Analytical Limits:
Acid and caustic leachates were cautiously combined and adjusted to pH 3. Separate loading experiments used the two size fractions of sample A, chosen for contrasting fluoride-to-aluminium ratios. The column held 5.50 mL wet-settled La-MTS9501 resin, 1.79 g dry mass, with reverse-flow pumping near 20 degrees C and fractions collected every 0.5 bed volume. Experimental states a 1.00 bed-volume-per-hour flow rate, whereas Figures 5-6 and S33 state 0.50; this is unresolved. Loaded resin was washed with deionized water then eluted with 1 M NaOH. The displayed elution experiment is for A less than 1.18 mm, not proof of the same elution performance for every waste grade or loading arm.
Acid and caustic leachates were cautiously combined and adjusted to pH 3. Separate loading experiments used the two size fractions of sample A, chosen for contrasting fluoride-to-aluminium ratios. The column held 5.50 mL wet-settled La-MTS9501 resin, 1.79 g dry mass, with reverse-flow pumping near 20 degrees C and fractions collected every 0.5 bed volume. Experimental states a 1.00 bed-volume-per-hour flow rate, whereas Figures 5-6 and S33 state 0.50; this is unresolved. Loaded resin was washed with deionized water then eluted with 1 M NaOH. The displayed elution experiment is for A less than 1.18 mm, not proof of the same elution performance for every waste grade or loading arm.
Curated Evidence Points:
Support
Source:
robshaw2020
Results sections on column loading and elution, Figures 5-6 and Table 5 show two-stage fluoride breakthrough followed by coincident aluminium and fluoride elution into a liquor still containing sodium and other constituents. The reported 59.0% recovery is obtained from the main fluoride elution-peak area divided by theoretical column uptake from dynamic model fitting, as defined in Experimental. It is not a measured original-feed-to-product recovery percentage; early water-wash losses and proposed recycle remain separate.
Support
Source:
robshaw2020_si
Figure S4 documents the miniature column; Figures S33-S35 and Table S5 show the second sample-A loading arm and model fits. The first and second breakthrough regions are fitted separately. Dose-response, Thomas and Yoon-Nelson parameters characterize these curves but do not independently establish a molecular pathway or industrial design capacity. Figure S2 is a conceptual circuit, not evidence of operating all its proposed downstream stages.
Partial
Source:
robshaw2020_si
Table S4 and Figures S29-S32 show precipitation while mixing leachates despite pH control. Source-assigned cryolite occurs with amorphous material; it is a loss before the resin and not the pure final commodity proposed for future precipitation work. Figure S1 describes probable resin binding using prior studies and is not a new structural measurement on this heterogeneous feed.
Partial
Source:
robshaw2020
Table 4 presents Aqion predictions, not measured aqueous ligand assignments. Figure 5 uncertainty reflects three electrode readings and Figure 6 duplicate measurements, without clearly established independent column campaigns. Comparisons with earlier simulant work do not substitute for matched controls or a new replication. Pure-product precipitation, full multielement mass closure, residual waste suitability and process-scale economics were not established.
Sources & References
-
robshaw2020: Robshaw, Bonser, Coxhill, Dawson and Ogden (2020). Development of a Combined Leaching and Ion-Exchange System for Valorisation of Spent Potlining Waste. Waste and Biomass Valorization 11:5467-5481. DOI 10.1007/s12649-020-00954-1. Complete original article, Experimental and Results and Discussion, Figures 1-6 and Tables 1-5; institutional PDF includes one repository cover before the 15 article pages.
https://doi.org/10.1007/s12649-020-00954-1 (accessed 2026-10-14) -
robshaw2020_si: Robshaw et al. (2020), Supporting Information, institutional version T J Robshaw Waste Biomass Val SI v2.pdf. Eighteen numbered pages plus repository cover; Figures S1-S35 and Tables S1-S5. Same experimental family as the original article, with some explicitly cited prior-study diagrams and comparison data rather than independent replication.
https://eprints.whiterose.ac.uk/id/eprint/157796/7/T%20J%20Robshaw%20Waste%20Biomass%20Val%20SI%20v2.pdf (accessed 2026-10-14)
Curation History
| Timestamp | Curator | Action | Summary |
|---|---|---|---|
2026-10-14T22:54:44.238258+00:00 |
Codex (GPT-6; Codex) | CREATE | Created mixed-cut spent potlining with qualified abiotic chemical leaching and resin separation after complete primary and supplement inspection, retaining recipe conflicts, model-defined recovery and scale limits. |
Record source: data/records/spent-potlining.yaml