Ultra-Efficient and Selective Gold Separation via Second-Sphere Coordination of Aurous Dihalide Using a Nonporous Amorphous Superadsorbent

Abstract

The escalating demand for gold, coupled with dwindling terrestrial reserves, underscores the urgent need for innovative separation strategies, including e-waste recycling and seawater extraction. However, the development of ultra-efficient, highly selective adsorbents capable of recovering trace amounts of gold from complex aquatic matrices remains a formidable challenge. Herein, a covalent organic superphane cage is reported as a nonporous amorphous superadsorbent (NAS) for selective and efficient gold recovery via intermolecular second-sphere coordination of AuBr₂⁻ (or AuCl₂⁻) ions, subsequently converted to metallic gold through disproportionation. NAS demonstrates outstanding performance, including an exceptional gold uptake capacity of 2750 mg g⁻¹, ultrafast adsorption kinetics (40 s), broad pH tolerance (1-11, up to 6 M acids), and remarkable gold uptake even in 36 wt.% HCl solution (821 mg g⁻¹). NAS achieves over 99% selective gold recovery, even amidst excess competing ions, retaining efficacy across 30 regeneration cycles. Its versatile and scalable design enables applications in gold separation from gold-bearing e-waste, catalytic residues, gold ores, and seawater. A large-scale trial recovered 23.8 Karat gold from printed circuit board leachates, positioning NAS as a sustainable and eco-friendly solution for industrial and environmental gold recovery.

Keywords: Covalent organic cage; Gold separation; Nonporous amorphous sorbent; Second‐sphere coordination; Superphane.

Figure 1

Working mechanisms for gold separation. a) Direct capture of Au(III) via first–sphere coordination as the driving force (known); b) direct capture of AuX₄ ⁻ via second–sphere coordination (known); c) direct AuX₂ ⁻ capture via second–sphere coordination and the gold recovery profile reported in this work. The energetically favorable AuX₂ ⁻ binding through second–sphere coordination offsets the endergonic nature of the decomposition reaction AuX₄ ⁻ ⇌ AuX₂ ⁻ + X₂, effectively propelling this reaction forward.

Figure 2

Single–crystal structure of Br⁻·2H₂O@NAS–HBA·2H⁺·AuBr₂ ⁻ complex. a) a snapshot of the complex with bromide dihydrate and AuBr₂ ⁻ shown in space–filling model; b) one selected Br⁻·2H₂O@NAS–HBA·2H⁺ complex surrounded by four AuBr₂ ⁻ ions; c) one selected AuBr₂ ^– interacting with four Br⁻·2H₂O@NAS–HBA·2H⁺ complexes through multiple hydrogen bonds as indicated by green dashed lines (short C/N─H···Br contacts of 2.72 – 3.46 Å); d) the AuBr₂ ⁻–mediated 3D supramolecular organic frameworks.

Figure 3

Direct AuBr₂ ⁻ adsorption by NAS–HBA. a) Time–resolved gold recovery efficacy. b) Post–adsorption NAS–HBA morphology via SEM and elemental composition via EDS, c) Temporal high–resolution XPS tracking of gold species evolution. d) Bromide ion concentration assay during adsorption. e) Comparative UV–vis spectral analysis of AuBr₄ ⁻ (0.01 mm), Br₃ ⁻ (0.1 mm) and post–NAS–HBA treatment of AuBr₄ ⁻ (4 mm) in water. f) Br 3d XPS characterization pre– and post–adsorption, g) Raman spectroscopic assessment of NaAuBr₄ and NAS–HBA after gold adsorption. h) The hypothesized mechanism and evidences for direct adsorption of AuBr₂ ⁻. Data represent mean ± SD from n = 3 independent experiments.

Figure 4

Gold adsorption performance of NAS–HBA. a) The adsorption kinetics of AuBr₄ ^– (in green) and AuCl₄ ^– (in black) solutions with the concentration of 600 ppm. b) The dependence of gold adsorption capacity on adsorption concentration, Ce is the concentration of Au(III) in aqueous solution before adsorption. c) The effect of pH value on the adsorption of AuBr₄ ^– over NAS–HBA. d) The dependence of gold adsorption capacity on concentration of hydrochloric acid. e) The adsorption rate of gold from complex liquids containing 20 ppm of NaAuBr₄ in the presence of 1 equivalent of equal–mass Mg²⁺, Al³⁺, Cr³⁺, Fe³⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Cd²⁺ and Pb²⁺ (as their chloride salts) using NAS–HBA solid. f) Removal efficiency (relative uptake) of ions from aqueous AuBr₄ ^– (20 ppm) solutions containing an excess (1 to 200 folds) of equal–mass competing cations, viz. Mg²⁺, Al³⁺, Cr³⁺, Fe³⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Cd²⁺ and Pb²⁺. g) Removal efficiency (relative uptake) of ions from aqueous AuBr₄ ^– (20 ppm) solutions containing an excess (1 to 200 folds) of equal–mass competing cations, viz. Cl⁻, Br⁻, NO₃ ⁻, SO₄ ²⁻ and PO₄ ³⁻. h) The gold removal efficiency with recycled adsorbent NAS–HBA during 30 cycles in the batch adsorption experiments. i) Partial ¹H NMR spectra of fresh (bottom) and recycled (top) NAS–HBA in DMSO–d₆ after undergoing 30 adsorption–desorption cycles. Error bars represent SD. n = 3 independent experiments.

Figure 5

The gold separation performance of NAS–HBA from real–world samples. a) Gold–containing catalytic wastewater; aqua regia leaching solutions obtained from b) the AMD CPUs, c) the mobile phone of PCBs, d) the gold ores; e) wastewater from Xiangjiang River (C₀ = 10 ppb); f) seawater from The Yellow Sea (C₀ = 10 ppb); g) the NBS/Py leaching solutions of mobile phone PCBs; h) the scaled–up selective gold adsorption from the NBS/Py leaching solutions of mobile phone PCBs; and i) the image of recycled pure metallic gold particles from h) after calcination. Error bars represent SD. n = 3 independent experiments.