Recovering palladium and gold by peroxydisulfate-based advanced oxidation process

Abstract

Palladium (Pd) and gold (Au) are the most often used precious metals (PMs) in industrial catalysis and electronics. Green recycling of Pd and Au is crucial and difficult. Here, we report a peroxydisulfate (PDS)–based advanced oxidation process (AOPs) for selectively recovering Pd and Au from spent catalysts. The PDS/NaCl photochemical system achieves complete dissolution of Pd and Au. By introducing Fe(II), the PDS/FeCl₂·4H₂O solution functioned as Fenton-like system, enhancing the leaching efficiency without xenon (Xe) lamp irradiation. Electron paramagnetic resonance (EPR), ¹⁸O isotope tracing experiments, and density functional theory calculations revealed that the reactive oxidation species of SO₄·⁻, ·OH, and Fe(IV)═O were responsible for the oxidative dissolution process. Lixiviant leaching and one-step electrodeposition recovered high-purity Pd and Au. Strong acids, poisonous cyanide, and volatile organic solvents were not used during the whole recovery, which enables an efficient and sustainable precious metal recovery approach and encourage AOP technology for secondary resource recycling.

A green and sustainable method of Pd and Au recovery uses peroxydisulfate-based AOPs.

INTRODUCTION

The precious metals (PMs) gold (Au) and silver (Ag) were treasured throughout ancient times for their beauty and industrial performance. Their importance increased the latter part of the 20th century, especially in the electrical industries, primarily due to their superior electrical conductivity and corrosion resistance, among other physicochemical properties (1–3). According to the World Gold Council, the United States produced 170 tons of Au products from its Au mining. The products find their primary application in jewelry, physical bars, central banks, electrical and electronic industry, etc. In addition, 90 tons of new and old scrap was recycled, which accounts for approximately 36% of reported consumption in 2022 (4). Recovering Au from the Au scrap would decrease the environmental and social impacts of mining (5–7). In addition, platinum group metals (PGMs), such as platinum (Pt), palladium (Pd), and rhodium (Rh), which are rare PMs with high economic values, are often applied in automobile three-way catalysts (8, 9). Because of the increasing applications, the global demand for PGMs has kept rising from 2011 to 2019 with an average annual demand of about 592 tons. In comparison to 2019, the usage of PGMs has decreased since 2020 because of the impact of COVID-19. PGM consumption has also begun to decline in recent years as a result of the notable rise in the market share of battery electric vehicles. Because of reduced availability, Pt may experience more severe deficits, while Pd is expected to continue facing a deficit (10, 11). Electronic waste (e-waste) and waste catalysts are the primary secondary resources of PMs, which account for the majority of the market (12, 13).

The recovery of PMs from ores, catalysts, and e-waste remains challenging. In general, the PMs need to be suitably enriched and separated using pyrometallurgy and hydrometallurgy. Pyrometallurgy is dominant in the recovery of low content PMs and is typically used as a means of enrichment. Our primary focus is on hydrometallurgy, a process that involves converting PMs into soluble compounds using chemical lixiviants for their subsequent separation and refining. The first step involves oxidative leaching, which is generally challenging because of the insoluble nature of PMs (14, 15). Cyanidation is extensively used industrial dissolution technique but is highly toxic in the environment (16–20). Because of the toxicity of cyanide, alternative nontoxic leaching chemicals including thiourea, thiosulfate, and iodine have been developed to dissolve Au. However, they are ineffective for PGMs, and the reaction mechanisms are often complex (21–23). To address this issue, Lin and colleagues (24) created “organic aqua regia” by combining thionyl chloride with some organic solvents like N,N-dimethylformamide, pyridine, or imidazole to dissolve PMs. Repo and colleagues (25) devised a method for dissolving Au using pyridine thiols and hydrogen peroxide (H₂O₂) in ethanol solutions. Recently, Repo et al. (26) dissolved Au quantitatively in ethanol using 2-mercaptobenzimidazole as a ligand in the presence of catalytic quantities of iodine; the process took approximately 13 hours. Love’s group (27) developed a new leaching solvent composed of triphenylphosphine dichloride, acetonitrile, and H₂O₂ for dissolving PMs quickly. Nevertheless, the conventional oxidative leaching method for PMs frequently uses potent polar organic solvents, halogens, or strong acids, which result in severe environmental damage due to the secondary pollution produced during the leaching process (as detailed in table S1, numerous hydrometallurgical processes on recovering PMs in laboratory-scale continue to generate quantities of environmentally hazardous waste). Presently, the high-pressure oxidation leaching process is widely used in industry. Despite the fact that it effectively reduces chemical costs, this method necessitates more advanced technology and entails certain security hazards. As a result, the search for an environmentally friendly, secure, and effective oxidation extraction method for PMs becomes increasingly critical.

Advanced oxidation processes (AOPs) represent a cutting-edge approach involving the production of radicals with high oxidation abilities for the degradation or transformation of pollutants (28–32). The strong oxidizing radicals generated during AOPs have recently been harnessed to achieve the oxidative dissolution of PMs. For example, under ultraviolet (UV) light, Li et al. (33) transformed Au nanoparticles into cyanide Au oligomers using a mixture of acetonitrile (CH₃CN) and benzaldehyde (C₇H₆O). Electron paramagnetic resonance (EPR) demonstrated that the active species ·CN and ·OH in CH₃CN and C₇H₆O, respectively, could dissolve Ag, Pd, and Pt. Similarly, Chen et al. (12) used a mixture of CH₃CN and dichloromethane (CH₂Cl₂) to oxidize and dissolve Ag, Au, Pd, Pt, Ru, Rh, and iridium (Ir) under light conditions and used TiO₂ photocatalysts in generating CH₂CN·, CHCl₂·, and O₂·⁻ radicals. However, C₇H₆O, CH₃CN, and CH₂Cl₂ are considered non-environmentally friendly chemicals. The use of organic solvents for PM recovery generates large volumes of liquid organic waste and some toxic volatile gases. By substituting strong inorganic acids and volatile organic solvents with environmentally friendly solvents and implementing process optimization techniques to reduce chemical consumption, PM recovery can be rendered more environmentally friendly, safe, and sustainable. However, the current lack of AOPs for PM extraction using inorganic solution systems represents a critical research gap.

The development of peroxydisulfate (PDS)–based AOPs has gained attention from both academic and industrial communities. PDS acts as the precursor of sulfate (SO₄·⁻) or hydroxyl (·OH) radicals activated using heat, transition metals, ultrasonic stimulation, and light irradiations (34–38). Ammonium persulfate [(NH₄)₂S₂O₈] was used to extract nonleaching Au by partial substrate oxidation of Ni, Fe, and Cu from waste electrical and electronic equipment (39). With the oxidative leaching of other metals, however, the method for nonleaching of PMs has various restrictions. The purity of the reclaimed PMs will be diminished if the PM scraps contain non-oxidizable leach impurities. To the best of our knowledge, few or perhaps no studies have yet reported on recovering PMs using highly reactive radicals generated from PDS. Here, we used the highly reactive species generated from PDS-based AOPs to develop a simple leaching process that leverages a PDS/NaCl photochemical solution and enhanced Fenton-like system of PDS/FeCl₂·4H₂O solution to selectively recover Pd and Au without strong acids, volatile organic solvents, light irradiation, and photocatalysts. These two PMs were successfully recovered from their respective waste catalysts via PDS/FeCl₂·4H₂O dissolution and one-step electrodeposition. This study presents an efficient and sustainable approach that can inform the development of AOP technologies for efficient Pd and Au recovery.

RESULTS

Dissolution properties of PDS/NaCl photochemical system

The cleavage of the peroxide bond in the PDS molecule (Fig. 1A) occurs via energy and electron transfer processes and results in the production of highly reactive SO₄·⁻(40, 41). Reactive oxygen species (ROS) generated by the decomposition of PDS may be used for the oxidative extraction of PMs. However, PM leaching systems require both strong oxidation and coordination capabilities (42–44). Therefore, we selected PDS combined with NaCl, the most common salt, as a solution here for PM leaching because of the robust coordinating ability of chloride ions (Cl⁻) in an aqueous solution (45). The following equations describe the potential production and transformation of SO₄·⁻(41, 46)

$$\begin{array}{c}{\mathrm{S}}_2{\mathrm{O}}_8^{2-}+\text{energy} \text{transfer}\to 2\mathrm{S}{\mathrm{O}}_4^{-}· E^0({\text{SO}}_4·{}^{-}/{{\text{SO}}_4}^{2-})\hfill \\ =2.6 \text{to} 3.1 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (1)

$$\begin{array}{c}\mathrm{S}{\mathrm{O}}_4^{-}·+{\mathrm{H}}_2\mathrm{O}\to \text{HS}{\mathrm{O}}_4^{-}+·\text{OH} k[{\mathrm{H}}_2\mathrm{O}]\hfill \\ =460\pm 90 {\mathrm{s}}^{-1} E^0\text{OH}/{\text{OH}}^{-})=1.9 \text{to} 2.7 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (2)

$$\begin{array}{c}\mathrm{S}{\mathrm{O}}_4^{-}·+\mathrm{C}{\mathrm{l}}^{-}\leftrightarrow \mathrm{S}{\mathrm{O}}_4^{2-}+\text{Cl}· k=3.2\pm 0.2\times {10}^8 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1},\hfill \\ k_{-}=2.1\pm 0.1\times {10}^8 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1} E^0\text{Cl}/{\text{Cl}}^{-})=2.5 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (3)

$$\begin{array}{c}\text{Cl}·+\mathrm{C}{\mathrm{l}}^{-}\leftrightarrow \mathrm{C}{\mathrm{l}}_2^{-}· k=8.5\times {10}^9 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1},\hfill \\ k_{-}=6\times {10}^4 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1} E^0\text{Cl}/{\text{Cl}}^{-})=2.2 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (4)

Fig. 1. Photochemical leaching system.

(A) Schematic illustration of the PDS molecular structure. (B) Schematic diagram of the leaching of PMs in the mixed solution of PDS and NaCl under Xe lamp light. (C) Images of dissolved Pd from 0 to 120 min. (D) Dissolution ratios of five common PMs. Reaction conditions: Ag for powder (≤0.1 μm) and Au, Pd, Pt, and Rh for wires (0.1 mm), molar ratio of PDS to NaCl = 1:1, H₂O add% = 70%, 0.845 M for both PDS and NaCl, mass ratio of salt to metal = 100:1, reaction time = 300 min, temperature = 60°C, Xe light power of 50 W, photo flux into the system of 2.016 × 10¹⁹ photos s⁻¹ (see detailed calculation method in note S2).

PMs were mixed in a solution of PDS and NaCl under a xenon (Xe) lamp light. By using the reactive free radicals that generated in this solution, PMs can be leached in an environmentally friendly manner (Fig. 1B). The single PDS solution exhibited higher concentration of SO₄·⁻ than the system containing NaCl under the same condition (fig. S1), mainly due to the capture effect of Cl⁻ to SO₄·⁻ through one electron transfer (Eq. 3). The chloride radicals (·Cl) were also detected in PDS/NaCl solution (fig. S2). It indicated a notable decrease in the strength of SO₄·⁻ and ·OH when Xe light is not present. In addition, the SO₄·⁻ exhibited a modest reduction at a temperature of 20°C (fig. S1). The concentrations of SO₄·⁻ and ·OH were also measured, and fig. S3 shows the standard curve between the concentration of the free radical products and the peak area in the high-performance liquid chromatography analysis. These concentrations are listed in tables S2 and S3. The PDS concentration during the reaction process was measured by iodometric persulfate measurement (fig. S4). Figure S5 shows the decomposition rate of PDS in solution within 300 min. We found that Xe light irradiation and a relatively high temperature (60°C) accelerated the rate of activated decomposition of PDS. When the Pd wire was placed in a PDS/NaCl solution under Xe lamp irradiation, the color rapidly changed and became darker over time in this photochemical system (initial pH of PDS/NaCl photochemical system was 5.9; Fig. 1C). Pt, Rh, Au, and Ag were also leached in the PDS/NaCl to determine their solubilities (PMs were leached separately in the solution). Notably, Pd and Au dissolved rapidly in the PDS/NaCl solution. In addition, 97.18 ± 3.22% of Pd and 99.23 ± 2.18% of Au were dissolved in 300 min. However, Ag, Pt, and Rh did not dissolve in the PDS/NaCl (Fig. 1D). The insolubility of Ag could be attributed to the enormous quantities of Cl⁻ in PDS/NaCl solution (47). A silver chloride (AgCl) peak was identified in the insoluble Ag powder via powder x-ray diffraction (PXRD) analysis (fig. S6). This suggests that the AgCl layer played an important role to prevent Ag from dissolving in the PDS/NaCl solution. Although SO₄·⁻ may oxidize all the five PMs thermodynamically [E⁰(SO₄·⁻/SO₄²⁻) = 2.6 to 3.1 V_NHE, E⁰ (Rh³⁺/Rh) = 0.76 V_SHE < E⁰ (Ag⁺/Ag) = 0.799 V_SHE < E0(Pd²⁺/Pd) = 0.951 V_SHE < E0(Pt²⁺/Pt) = 1.18 V_SHE < E0(Au³⁺/Au) = 1.498 V_SHE, V_NHE = V_SHE + 0.24], how fast or easy a metal is dissolved is also determined by the kinetics of the system. It seems that Rh is the easiest to be oxidized thermodynamically among the five PMs. However, metal complexes must be formed in solution to dissolve. Rh ions are known to form kinetically inert complexes that form very slowly with chloride ions (48), which makes Rh an extremely insoluble PGM in PDS/NaCl solution. Pd and Au are the two PMs that can be leached most quickly in solutions containing chloride ions from literature work (49, 50). This finding is also in line with the outcomes of the PDS/NaCl photochemical system.

Leaching experiments of PMs in PDS/NaCl system

We determined the optimal reaction conditions of leaching Pd from spent catalysts by assessing the operating variables, including PDS-to-NaCl molar ratio, stirring speed, and temperature. An analysis of the dissolution of Pd yielded insights into the kinetics of the leaching process, a heterogeneous fluid-solid reaction that is frequently characterized by the shrinking-nucleus model (51, 52)

$$k_{\mathrm{c}}t=1-{(1-x)}^{1/3}$$ (5)

where x represents the Pd dissolution ratio, k_c represents the rate constants (minutes⁻¹), and t (minutes) represents the reaction time.

At a PDS:NaCl molar ratio of 1:1 (0.845 M for both PDS and NaCl), almost 100% of Pd could be dissolved in 60 min. Following Eq. 5, the k_c could be determined, and we found that the k_c could be enhanced by 2.24 and 4.02 times for the ratios of 2:1 and 1:2, respectively (Fig. 2A). Nevertheless, a single PDS solution had no effect on PM extraction because the dissolution of PMs requires a coordinating ligand to maintain a favorable reaction. Excessive amounts of NaCl can enhance the capture effect of Cl⁻ on SO₄·⁻ (Eq. 3). So, the optimal PDS/NaCl molar ratio for the dissolving process is 1:1. The leaching curve of Pd under different stirring speed is shown in Fig. 2B. The findings indicate that increasing the stirring speed can notably enhance the dissolution rate. As seen in Fig. 2C, the dissolution ratios increased and subsequently decreased as the leaching temperature increased from 20° to 80°C. Increasing the temperature could improve the activity of the solution and thus enhance the dissolution ratios. However, excessively high temperatures can cause the PDS to decompose too quickly (the PDS decomposition rate constant was 0.0298 min⁻¹ under 80°C compared to that of 0.0233 min⁻¹ under 60°C), preventing the reactive radicals from reacting with Pd (fig. S7). Therefore, a comprehensive consideration led to the selection of 60°C as the optimal temperature.

Fig. 2. Dissolution properties of the photochemical system.

(A to D) Dissolution ratios of Pd in different conditions. Optimized dissolution conditions: Solution (5 g of salts with the molar ratio of the individual salt of 1:1, for example, 4.01 g of PDS, 0.99 g of NaCl, and 11.67 g of H₂O), 100 mg of spent Pd catalyst, 60°C, 500 rpm, leaching time of 60 min. (E) Solubility for Au with different solid-to-liquid ratios (60°C, 500 rpm, leaching time of 360 min; 10 g of PDS/NaCl solution; the masses of Au wires at different solid/liquid mass ratios of 0.1/50, 0.5/50, 0.8/50, 1/50, 1.5/50, 2/50, and 2.5/50 g/g are 20, 100, 160, 200, 300, and 500 mg, respectively).

The photochemical systems of PDS with various chloride salts were also tested in dissolution experiments. Aside from FeCl₂·4H₂O, the other chlorine salts including ZnCl₂, MgCl₂, CoCl₂, and AlCl₃·6H₂O with PDS had similar Pd dissolution ratios, compared with that of NaCl (Fig. 2D). The Fenton reaction entails H₂O₂ decomposition by Fe(II) to product ·OH as follows (53)

$${\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{F}{\mathrm{e}}^{2+}\to \mathrm{F}{\mathrm{e}}^{3+}+·\text{OH}+\mathrm{O}{\mathrm{H}}^{-} k=55.4 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$ (6)

The addition of Fe(II) to PDS transforms the solution into a Fenton-like system, promoting the generation of free radicals and increasing the leaching efficiency of Pd (54). Other chloride salts did not have a similar effect as FeCl₂·4H₂O and only served to provide Cl⁻. Consequently, Pd dissolution in the chloride salt systems did not differ much.

The dissolution ratio and the amount of dissolved Au were highly correlated with the solid/liquid ratio as shown in Fig. 2E and table S4. In general, the contents of dissolved Au increased with the increase of solid/liquid ratio, whereas the dissolution ratio decreased accordingly. The solubility of Au reached 69,367 parts per million (ppm) when at a solid/liquid ratio of 2.5:50.

Dissolution mechanism of PMs in PDS/NaCl system

Dissolution rates are affected by the PM/halide ion coordination mode difference. The dissolution ratio of Pd using the photochemical system was conducted when NaCl was replaced with other sodium halide salts like NaF, NaBr, and NaI. It was found that the dissolution rate of Pd in PDS/NaBr was faster than that in PDS/NaCl. Next came the PDS/NaI, while Pd was insoluble in PDS/NaF photochemical system (Fig. 3A). Pd was insoluble in PDS/NaF, probably due to the weak complexing ability of F⁻. While for NaI system, a large amount of purple smoke was generated when NaI powder was added to the aqueous solution of PDS (fig. S8). The process of rapid persulfate anions (S₂O₈²⁻) consumption by iodide ions (I⁻) occurs as follows (55)

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}+2{\mathrm{I}}^{-}\to 2\mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{I}}_2 k=4.11\times {10}^{-3} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$ (7)

Fig. 3. Dissolution mechanism of PDS/NaCl system.

(A) Dissolution ratios of Pd using PDS with different salts. (B) EPR signal intensity of PDS/NaCl and PDS/NaBr at different times. (C and D) Dissolution rates and rate constants of Pd dissolution with different quenching agents. (E) UV-vis spectra of PDS/NaCl leachate containing Pd(II). (F) The flow of the mechanism of the dissolution process. a.u., arbitrary units.

The expeditious volatilization of the iodine vapor produced throughout the reaction induced a prompt reduction in the oxidizing characteristics of the PDS/NaI photochemical solution system, thereby reducing the rate of Pd dissolution. We compared the intensities of SO₄·⁻ and ·OH between the PDS/NaCl and PDS/NaBr systems over time and found that the PDS/NaCl system had stronger 5,5-dimethyl-1-pyrrolidine-N-oxide (DMPO)–SO₄·⁻ and DMPO-·OH signal intensities (Fig. 3B), mainly because bromide ions (Br⁻) exhibit a higher SO₄·⁻ consumption rate compared to Cl⁻, resulting in the production of halogen radicals that are less reactive (see Eqs. 8 and 9) (56)

$$\begin{array}{c}\mathrm{S}{\mathrm{O}}_4^{-}·+\mathrm{B}{\mathrm{r}}^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\text{Br}· k=3.5\times {10}^9 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}\hfill \\ (E^0\text{Br}/{\text{Br}}^{-})=2.0 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (8)

$$\begin{array}{c}\text{Br}·+\mathrm{B}{\mathrm{r}}^{-}\leftrightarrow \mathrm{B}{\mathrm{r}}_2^{-}· k=1.2\times {10}^{10} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}\hfill \\ (E^0\text{Br}/{\text{Br}}^{-})=1.7 {\mathrm{V}}_{\text{NHE}}\hfill \end{array}$$ (9)

The stability of PM^x+ complexes diminishes with an increase in the electronegativity of the ligand donor atoms. Therefore, the stability order of halogen and Pd complexes is as follows: Br⁻ > Cl⁻ (48). This may elucidate the accelerated extraction of Pd in the PDS/NaBr photochemical system, which exhibits a reduced potency of reactive radicals.

The results of the free radical quenching experiments were shown in Fig. 3C. The methanol (MeOH) was used as SO₄·⁻ and ·OH scavenger because of the similar reactivity toward SO₄·⁻ (k_SO4·⁻ _+MeOH = 2.5 × 10⁷ M⁻¹ s⁻¹) and ·OH (k_·OH+MeOH = 9.7 × 10⁸ M⁻¹ s⁻¹), whereas tert-butanol (TBA) was used to capture the ·OH (k_·OH+TBA = 3.8 to 7.6 × 10⁸ M⁻¹ s⁻¹) instead of SO₄·⁻ (k_SO4·⁻_+TBA = 4.0 to 9.1 × 10⁵ M⁻¹ s⁻¹) (57). Chlorine radicals (·Cl) were excluded from consideration here because of the potent quenching effects of MeOH and TBA on them (k_·Cl+MeOH = 9 ± 0.5 × 10⁸ M⁻¹ s⁻¹ and k_·Cl+TBA = 3 × 10⁸ M⁻¹ s⁻¹) (58, 59). It appears challenging to identify a distinct quenching agent specifically for ·Cl. Furthermore, the formation of ·Cl occurs as a result of the reaction between SO₄·⁻ and Cl⁻ (Eq. 3). The potential oxidation of ·Cl can be ascribed to SO₄·⁻. The blank group without any quenching agent had a Pd leaching rate of 0.01343 min⁻¹. The Pd leaching rate after adding 1, 10, and 100 mM TBA decreased to 0.0092, 0.0075, and 0.0057 min⁻¹, respectively. The Pd leaching rate plummeted to 0.0000145 min⁻¹ after adding 100 mM MeOH, much lower than the group with only ·OH quenching. This suggested that ·OH and SO₄·⁻ were the principle active species responsible for the oxidative leaching of Pd in the PDS/NaCl system, with ·OH making up a higher proportional contribution of 57.9%.

Numerous investigations have been carried out regarding the chemical circumstance of PM^x+ ions in acid and aqueous solutions from both a science and a commercial point of view (60–62). In aqueous acidic solutions, previous research indicated that PM^x+ ions would seemingly form a square-planar framework with four Cl⁻ and an H₂O molecule as coordinating ligands. The ultraviolet-visible (UV-vis) spectroscopy of Pd leaching in the PDS/NaCl solution with varied concentrations was shown in Fig. 3E. The firm adsorption peaks at 207 and 237 nm mainly corresponded to the [PdCl₄]²⁻, which is dominant in the leachate (7, 63). On the basis of the experimental analyses, a possible mechanism is proposed to account for the photochemical dissolution of Pd in the PDS/NaCl solution. First, PDS was activated by Xe lamp, which contributed to the generation of SO₄·⁻ and ·OH in the PDS/NaCl solution (step 1). Second, Pd was oxidized and leached by these two reactive radicals to form Pd ions (step 2). In addition, a large amount of Cl⁻ in the solution continuously coordinated with Pd²⁺ to form complexes (step 3), promoting the positive progression of dissolution (64). The whole process roughly included free radical generation, oxidative dissolution, and coordination (Fig. 3F). The specific reaction equations are as follows

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}+\text{light activation}\to 2\mathrm{S}{\mathrm{O}}_4^{-}·$$ (Step 1A)

$$\mathrm{S}{\mathrm{O}}_4^{-}·+{\mathrm{H}}_2\mathrm{O}\to \text{HS}{\mathrm{O}}_4^{-}+·\text{OH}$$ (Step 1B)

$$\text{Pd}+\mathrm{S}{\mathrm{O}}_4^{-}·+·\text{OH}\to \mathrm{P}{\mathrm{d}}^{2+}+\mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{O}{\mathrm{H}}^{-}$$ (Step 2)

$$\mathrm{P}{\mathrm{d}}^{2+}+4\mathrm{C}{\mathrm{l}}^{-}\to {[\text{PdC}{\mathrm{l}}_4]}^{2-}$$ (Step 3)

Improved leaching system: The PDS/FeCl₂·4H₂O Fenton-like system

As seen in the PDS/NaCl system, the efficiency of Pd leaching was improved when NaCl was replaced by FeCl₂·4H₂O. S₂O₈²⁻ could be activated by Fe(II) to generate SO₄·⁻ and sulfate, as depicted by Eq. 10 (65)

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}+\mathrm{F}{\mathrm{e}}^{2+}\to \mathrm{F}{\mathrm{e}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{S}{\mathrm{O}}_4^{-}· k=15.33 {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$ (10)

We optimized the Pd leaching rate by modifying the composition of the leaching system, especially the H₂O added and Xe lamp. The Fenton-like system exhibited the highest Pd leaching rate when 75% H₂O (0.76 M for PDS or FeCl₂·4H₂O) was used. As shown in Fig. 4A and fig. S9, the dissolution of Pd in PDS/ FeCl₂·4H₂O solution fits the shrinking-nucleus model like that in PDS/NaCl solution. The decreased ion concentration reduced Pd leaching efficiency when the H₂O added reached 90%. When 75% H₂O was added, the Xe lamp did not irradiate, and the Pd leaching rate did not differ notably (dissolution rates decreased from 0.051 to 0.047 min⁻¹). The PDS/FeCl₂·4H₂O Fenton-like system does not require Xe lamp irradiation, which may reduce energy consumption and increase leaching rates, required only 15 min to achieve ~100% dissolution of Pd in the spent catalysts. The PDS decomposition rate was measured as 0.035 min⁻¹, higher than that in PDS/NaCl photochemical system. In the persulfate/Fe(II) system, the quick conversion of Fe(II) to Fe(III) leads to a swift decrease in the effectiveness of persulfate activation. Several researchers are endeavoring to improve the redox cycle of Fe(III)/Fe(II) by including reducing chemicals or developing tailored catalysts (66, 67). We investigated the addition of reducing agents, like citric acid (CA), ascorbic acid (AA), and H₂O₂ which may function as reducing or oxidizing agents, to examine how the concentration of iron species affects the dissolution rates of Pd (75% H₂O added, 0.076 M for CA, AA, and H₂O₂). The Pd leaching rates peaked when adding AA (Fig. 4B). Fe(II) was analyzed by the 1,10-phenanthroline method (see note S4). Fe_tot represents the total amount of iron ions minus the Fe(II). As shown in Fig. 4C, the molar ratio of Fe²⁺ achieved the highest when AA was added. This outcome implies that the use of AA expedited the transformation of Fe(III) to Fe(II), thereby enhancing the efficiency of Pd leaching. Figure 4D showed the reaction kinetics of Pd and Au in PDS/FeCl₂·4H₂O Fenton-like system versus PDS/NaCl photochemical systems. PDS/FeCl₂·4H₂O solution greatly improved PM leaching efficiency and did not require Xe lamp irradiation, thereby simplifying operation and may greatly reducing the energy consumption.

Fig. 4. Dissolution properties of the Fenton-like system.

(A and B) Dissolution ratios of Pd in PDS/FeCl₂·4H₂O under different conditions (1:1 molar ratio of PDS-to-FeCl₂·4H₂O in each solution; the pH values of solutions with 70, 75, 80, and 90% H₂O add% are 1.8, 2.2, 2.4, and 2.7, respectively, and the concentrations of PDS or FeCl₂·4H₂O of each solution are 0.98, 0.76, 0.57, and 0.25 M). (C) Concentrations of Fe²⁺ and Fe_tot for different reducing agents added. (D) Kinetic rates of Pd and Au in PDS/FeCl₂·4H₂O (without Xe lamp) and PDS/NaCl photochemical system.

Dissolution mechanism of PMs in PDS/FeCl₂·4H₂O Fenton-like system

The open circuit potential (OCP) was used to evaluate the electron transfer process to elucidate the reaction mechanism between PDS and FeCl₂·4H₂O (68). As shown in Fig. 5A, the OCP of the PDS solution remained stable within 70 s but rose immediately after FeCl₂·4H₂O was added. This indicated a fast transfer of electrons in the PDS/FeCl₂·4H₂O system. The OCP change in the PDS/ FeCl₂·4H₂O solution (~0.17 V) is larger than that observed after Xe lamp irradiation in the PDS/NaCl system (~0.1 V), indicating a more rapid activation of PDS by Fe(II) (fig. S10). Electron transfer helps to generate free radicals and other oxidatively active species (69), which can be identified using other exploration methods.

Fig. 5. Dissolution mechanism of PDS/FeCl₂·4H₂O system.

(A) OCP curves of the bare PDS solution and PDS after adding FeCl₂·4H₂O. (B) EPR spectra of DMPO-·OH, DMPO-SO₄·⁻, and TEMP-¹O₂. (C) Extracted ion chromatography (EIC) of PMS¹⁶O¹⁶O and PMS¹⁶O¹⁸O in PDS/FeCl₂·4H₂O solution with H₂¹⁸O. mAU, milliabsorbance units. (D) Free energy relative pathways of the PDS/FeCl₂·4H₂O solution to generate SO₄·⁻, ·OH, and Fe(IV). (E) EPR spectra for the detection of the DMPO-·OH adduct in PDS/FeCl₂·4H₂O in the addition of scavengers. (F) Dissolution of Pd in the presence of scavengers in the PDS/FeCl₂·4H₂O solution. (G) Schematic illustration of the oxidation dissolution of Pd and Au in PDS/FeCl₂·4H₂O. (H) Raman spectra of Pd(II) in the PDS/FeCl₂·4H₂O solution. (I) Diagram of the dissolution and coordination mechanism of the PMs in the PDS/FeCl₂·4H₂O solution.

Identification of the principle reactive species is greatly notable for comprehending the fundamental mechanism of PM leaching in the PDS/FeCl₂·4H₂O Fenton-like system. In light of the redox potential [normal hydrogen electrode (NHE) and standard hydrogen electrode (SHE)] of ROS [E⁰(SO₄·⁻/SO₄²⁻) = 2.6 to 3.1 V_NHE > E⁰(·OH/OH) = 1.9 to 2.7 V_NHE > E⁰(O₂·⁻,2H⁺/H₂O₂) = 0.89 V_NHE > E⁰(¹O₂,/ O₂·⁻) = 0.81 V_NHE] and PMs [E⁰ (Pd²⁺/Pd) = 0.951 V_SHE, E⁰ (Au³⁺/Au) = 1.498 V_SHE], superoxide radicals (O₂·⁻) is excluded from consideration due to their insufficient capacity to oxidize Pd and Au (V_NHE = V_SHE + 0.24). EPR techniques were used to distinguish the ROS. As seen in Fig. 5B, SO₄·⁻ and ·OH existed in the PDS/FeCl₂·4H₂O solution, while no signal oxygen (¹O₂) was found.

Moreover, Fe(IV) had been found in a similar Fenton-like system: Fe(II)/KIO₄ solution (32). One of the most defining aspects of high-valence metal-oxo species is their ability to engage in oxygen atom exchange with water. Therefore, incorporating ¹⁸O isotope into products to confirm the presence of Fe(IV) is essential. Consequently, by using an ultrahigh-performance liquid chromatography quadrupole time-of-flight premier mass spectrometer (UPLC-Q-TOF MS), the oxidation byproduct of methyl phenyl sulfoxide (PMSO) in the PDS/FeCl₂·4H₂O system in H₂¹⁸O was identified. Figure 5C showed that extracted ion chromatography (EIC) found two signal peaks at mass/charge ratio (m/z) = 176.0625 and m/z = 174.0583, which agreed with the estimated m/z of ¹⁸O isotope–labeled methyl phenyl sulfone (PMSO₂; noted as PMS¹⁶O¹⁸O, m/z = 176.0625 [M-NH₄]) and normalized PMSO₂ (noted as PMS¹⁶O¹⁶O, m/z = 174.0583 [M-NH₄]), respectively. Thus, the results demonstrated that ¹⁸O was effectively integrated into PMSO₂ during PMSO oxidation in the PDS/FeCl₂·4H₂O system, confirming that Fe(IV) was generated as a reactive intermediate (70).

Furthermore, the chemical pathways and thermodynamics of the PDS/FeCl₂·4H₂O process were thoroughly investigated using density functional theory simulations. Subsequently, three possible reaction routes were proposed where SO₄·⁻, ·OH, and Fe(IV) were generated as follows (fig. S11 and table S5 contain the model drawings and the summary of the total Gibbs free energy)

$$\begin{array}{c}{\mathrm{S}}_2{\mathrm{O}}_8^{2-}+\text{Fe}{({\mathrm{H}}_2\mathrm{O})}_6^{2+}+6{\mathrm{H}}_2\mathrm{O}\to \mathrm{S}{\mathrm{O}}_4^{-}·+\hfill \\ \mathrm{S}{\mathrm{O}}_4{({\mathrm{H}}_2\mathrm{O})}_6^{2-}+\text{Fe}{({\mathrm{H}}_2\mathrm{O})}_6^{3+}\hfill \end{array}$$ (11)

$$\mathrm{S}{\mathrm{O}}_4^{-}·+8{\mathrm{H}}_2\mathrm{O}\to \mathrm{S}{\mathrm{O}}_4{({\mathrm{H}}_2\mathrm{O})}_6^{2-}+·\text{OH}+{\mathrm{H}}_3{\mathrm{O}}^{+}$$ (12)

$$\begin{array}{c}\text{Fe}{({\mathrm{H}}_2\mathrm{O})}_6^{2+}+{\mathrm{S}}_2{\mathrm{O}}_8^{2-}+14{\mathrm{H}}_2\mathrm{O}\to \text{Fe}{({\mathrm{H}}_2\mathrm{O})}_5{(\mathrm{O})}^{2+}+\hfill \\ 2\mathrm{S}{\mathrm{O}}_4{({\mathrm{H}}_2\mathrm{O})}_6^{2-}+2{\mathrm{H}}_3{\mathrm{O}}^{+}\hfill \end{array}$$ (13)

As shown in Fig. 5D, the total changes in Gibbs free energy for SO₄·⁻ and ·OH formation were −1.49 and −1.45 eV, respectively and that for Fe(IV) formation was a bit higher and reached −0.018 eV.

Considering the slow interaction between Fe(IV) and TBA [k_Fe(IV)+TBA = 6.0 × 10¹ M⁻¹ s⁻¹], TBA at 200 mM was used to test for the presence of competing reactive species, specifically SO₄·⁻ and ·OH (k_SO4·⁻+TBA = 4.0 to 9.1 × 10⁵ M⁻¹ s⁻¹, k_·OH+TBA = 3.8 to 7.6 × 10⁸ M⁻¹ s⁻¹) (32). EPR spectra of Fig. 5E demonstrated the detection of intense peaks consistent with a DMPO-·OH adduct, which were attributed to the contribution of ·OH and/or Fe(IV) interference by the widely accepted Forrester-Hepburn mechanism (71). A much lower dimethyl sulfoxide (DMSO) concentration (5.0 mM) as the scavenger for SO₄·⁻,·OH, and Fe(IV) [k_{SO4·⁻+DMSO} = 2.6 to 3.4 × 10⁹ M⁻¹ s⁻¹, k_·OH+DMSO = 4.5 to 7.1 × 10⁹ M⁻¹ s⁻¹, k_Fe(IV)+DMSO = 1.26 × 10⁵ M⁻¹ s⁻¹] (72) notably weakened the observed peaks of the DMPO-·OH adduct when compared to the TBA with a concentration of 200 mM, showing that the EPR signal was predominately associated with Fe(IV) and that ·OH may function as a secondary intermediate of reactivity for SO₄·⁻ and Fe(IV). Scavenging experiments were conducted to determine the proportional contribution of SO₄·⁻, ·OH, and Fe(IV) in the oxidizing leaching process. TBA reduced the dissolution rates of Pd from 0.0464 to 0.0292 min⁻¹, which contributed to the proportion of ·OH in the leaching of Pd. Pd dissolution rates reduced by the addition of MeOH was less than that of DMSO because MeOH quenches only SO₄·⁻ and ·OH instead of Fe(IV) [k_{SO4·⁻+MeOH} = 2.5 × 10⁷ M⁻¹ s⁻¹, k_·OH+MeOH = 9.7 × 10⁸ M⁻¹ s⁻¹, k_Fe(IV)+MeOH = 2.5 × 10³ M⁻¹ s⁻¹] (32). By calculating the difference between the Pd dissolution rates, it is possible to derive the contribution of SO₄·⁻, ·OH, and Fe(IV) as 57.8, 37.1, and 3.6%, respectively. The schematic illustration of the oxidation dissolution of Pd and Au in PDS/FeCl₂·4H₂O by the reactive species SO₄·⁻, ·OH, and Fe(IV) was shown in Fig. 5G.

Raman spectroscopy investigation was performed on the Pd(II) containing PDS/FeCl₂·4H₂O solution. As shown in Fig. 5H, the band at 275 cm⁻¹ was observed in both the fitted curves, corresponding to [PdCl₄]²⁻ (63). UV-Vis spectroscopy was not used there because of the interference of Fe³⁺ on the UV signal (73). Combined with the UV-vis spectra of Pd(II) in the PDS/NaCl filtrate in Fig. 3E, it could be inferred that Pd complexes formed the [PdCl₄]²⁻ after being oxidized in the PDS/FeCl₂·4H₂O system. The dissolution and coordination mechanism of PMs in the PDS/FeCl₂·4H₂O solution was shown in Fig. 5I. Radicals with high redox potentials (2.60 to 3.10 V_NHE for SO₄·⁻/SO₄²⁻ and 1.90 to 2.70 V_NHE for ·OH/OH⁻) and Fe(IV) all contributed to the oxidative leaching process. Also, Fe(IV) may transform to Fe(II) after reducing by PM^x+, which enhanced the cycle of Fe(II) (74).

Simulated recovery process of Pd and Au

Because of the large difference in reduction potential between PM and other metal ions, electrodeposition is frequently used to reduce PM ions in solution for recovery (75–77). Figure 6A showed the cyclic voltammetry curves at 333 K for the PDS/FeCl₂·4H₂O solution containing Pd(II). Two pairs of oxidation-reduction peaks were observed on the curve, which corresponded to Fe³⁺/Fe²⁺ and Pd²⁺/Pd, respectively. The overall simulated recovery process was conducted on PM wires and their simulated recovery solutions. The Pd wire was first leached in PDS/FeCl₂·4H₂O solution, followed by one-step constant current deposition of 0.02 A (cathodic) using a three-electrode system (0.75 cm² for the working electrode and 1.50 cm² for the counter electrode). The electric potential curve with electrolysis time (Fig. 6B) illustrated that the Pd(II) in leachate decreased from 862.5 to 37.5 ppm after electrodeposition of 3600 s, indicating a high electrodeposition recovery efficiency of 95.6%. Equation S7 (Supplementary Materials) also calculated the current efficiency, which attained 84.3%.

Fig. 6. Recovery process of PMs.

(A) CV curves of Pd(II) in PDS/FeCl₂·4H₂O. (B) Potential-time curve of Pd(II) in PDS/FeCl₂·4H₂O. (C) PXRD pattern of Pd product. (D) Dissolution results of recycled solution. (E) Comparisons of chemical input and energy consumption in our work with other similar work. (F) Schematic drawing of the PM recovery system for industrial applications. (G) Thermogravimetric analysis coupled with mass spectrometry (TG-MS) curve of PDS/FeCl₂·4H₂O solution.

The energy-dispersive x-ray spectroscopy (EDS) results (figs. S12 to S14 and table S7) indicated that the Pd product was of high purity (>98.8%), and the PXRD pattern confirmed the product’s structure (Fig. 6C). For the PDS/FeCl₂·4H₂O solution recycling, 4.5 g of PDS was added to the solution as a supplement. After 5 cycles of operation, the dissolution ratios and electrodeposition recovery were nearly identical to the initial solution (Fig. 6D). Table S6 listed the dissolution ratios, constant current electrolytic recovery, current efficiency, and PDS consumption ratios during the five simulated cycles of Pd. No formation of iron sludge was seen during the whole process. Through the assessment of iron concentration and pH levels following each iteration, it was seen that the prevalence of Fe²⁺ escalated in tandem with the number of cycles, while the pH exhibited a corresponding decline (fig. S15). This may be due to an oxygen evolution reaction at the counter electrode: 2H₂O-4e⁻ → 4H⁺ + O₂. This phenomenon effectively ensured that Fe³⁺ remained within a secure pH range, hence preventing Fe(OH)₃ precipitation. It is possible that the Fe³⁺ undergoes partial reduction to Fe²⁺ on the working electrode alongside with the Pd²⁺, resulting in a partial decrease in the current efficiency during cycling. The recovery method of Au closely resembled that of Pd in a simulated way. A total of 97.3 ± 3.1% of Au was dissolved, with 94.9 ± 1.7% recovered within 1 hour at 0.02 A. The current efficiency maintained a high level and reached 80.1 ± 2.3%. Both the PXRD pattern and EDS results (figs. S16 to S19) demonstrated the high purity of the Au product (the chemical composition of elements was listed in table S10).

Recovery Pd and Au from spent catalysts

The mass composition of two catalysts is listed in tables S11 and S12. Spent catalysts containing PMs were also recycled. PDS/FeCl₂·4H₂O solution leaching and electrodeposition were used to recover PMs from spent catalysts. (Figures S20 and S21 depict scanning electron microscopy images taken prior leaching). The recovery efficiency and purity of Pd and Au were also calculated. More than 98% of the targeted elements in both spent catalysts were dissolved, and the Pd and Au recovered after electrodeposition showed a high purity (99.8% for Pd and 99.6% for Au; see detailed mass composition of PM products in tables S11 and S12). The PDS/FeCl₂·4H₂O solution maintained high dissolution stability of spent Pd catalyst and can be reused for more than 25 times (fig. S22). In addition, the solution can be recycled 20 times when leaching spent Au catalyst using the same method. Solution recycling diminishes the energy and material consumption of conventional recycling techniques. On average, the chemical inputs using PDS/FeCl₂·4H₂O as a lixiviant are about ¹/₄ of the traditional hydrometallurgy process, and the energy consumption is approximately ¹/₅ of other similar work when leaching the same quality of PMs (Fig. 6E and table S1). Because of its rapid Pd and Au leaching capabilities and environmentally friendly setup, our approach shows potential for industrial applications, which includes three steps: (i) PDS and FeCl₂·4H₂O are added to water to form a Fenton-like system. (ii) The spent catalysts are mixed with PDS/FeCl₂·4H₂O solution for the leaching. (iii) The PM^x+ are deposition using an electronic device, while the solution is recycled with quantitative addition of PDS into the first step (Fig. 6F). Thermogravimetric analysis coupled with mass spectrometry (TG-MS) spectrum of PDS/FeCl₂·4H₂O indicates that no potentially environmentally unfriendly gases (HCl, Cl₂, and SO₂) are evaporated during use of PDS/FeCl₂·4H₂O even heated it to exceed 120°C (Fig. 6G).

DISCUSSION

This study presents a green and sustainable approach using PDS-based AOPs to develop a simple leaching process for recovering Pd and Au from spent catalysts. Our method leverages the PDS/NaCl photochemical system, which effectively and completely dissolved Pd and Au after 60 min. The enhanced PDS/FeCl₂·4H₂O Fenton-like system required no Xe lamp light activation and greatly improved the leaching efficiency. Through radical identification experiments, ¹⁸O isotope tracing experiments, and theoretical calculations, we established that SO₄·⁻, ·OH, and Fe(IV) act as the main reactive oxidation species. High-purity Pd and Au were recovered integrating lixiviant leaching and one-step electrodeposition. The selective leaching of Pd and Au in the solution provides an admired benefit in the recovery of these metals. The solution retains its excellent dissolution properties even after 25 cycles with no addition of extra FeCl₂·4H₂O. We performed a techno-economic study of our recovery method, based on the costs of materials and the energy consumption of the whole process (see note S8). In comparison to literature work, the PDS/FeCl₂·4H₂O solution has an exceptionally low material cost and an admired low energy consumption. When leaching PMs of equivalent quality, the chemical inputs using PDS/FeCl₂·4H₂O as a lixiviant are, on average, ¹/₄ of the energy consumption observed in conventional hydrometallurgy processes, and the energy inputs are approximately ¹/₅ of those of comparable work. Our system also provides various advantages such as safe operation, easy storage and transportation, and low impact on the environment. Both PDS and FeCl₂·4H₂O are highly reactive, relatively cheap, and environmentally friendly. The whole recovery process does not involve any strong acids, toxic cyanide, volatile organic solvents, light activation, and any photocatalysts. No toxic gases were detected when even the solution was heated above 120°C. The strongly oxidizing free radicals and high-valent iron that are generated in the dissolution provide a method to achieve green dissolution. This study presents a major step toward advancing efficient Pd and Au recovery using AOPs that are environmentally friendly and sustainable thus meeting green climate demands. Future studies will focus on the use of PDS-based AOP technology for the recovery of more insoluble PMs such as Rh, osmium, and Ir.

MATERIALS AND METHODS

Radical identification experiments

EPR (Bruker EMX PLUS, German) spectra were collected to identify the intermediate oxidants or radicals generated from the interaction between PDS and FeCl₂·4H₂O, NaCl, or NaBr with DMPO and 2,2,6,6-tetramethylpiperidine (TEMP) or N-tert-butyl-α-phenylnitrone (PBN) as spin-trapping reagent (DMPO for ·OH and SO₄·⁻; TEMP for ¹O₂; PBN for ·Cl; the concentrations for the three spin-trapping reagents are all 100 mM). The EPR spectra were collected under the following conditions: a center field of 3510 Gs, a sweep width of 100 Gs, a static field of 3460 G, a microwave power of 6.325 mV, a microwave frequency of 9.85 GHz, a modulation frequency of 100 kHz, a modulation amplitude of 1.0 Gs, and a sweep time of 30 s.

Radical quenching experiments

Radical quenching measurements were conducted to identify the dominant reactive species in PDS/FeCl₂·4H₂O or PDS/NaCl systems with MeOH, TBA, and DMSO, which were added before the reaction.

Solution analysis characterization

The concentrations of PMs were analyzed by an inductively coupled plasma-optical emission spectrometer (ICP-OES; ICAP PRO, Thermo Fisher Scientific Inc., USA). Detailed operation procedures were listed in note S2 in the Supplementary Materials. The solution analysis characterization of Pd(II)-loaded PDS/NaCl and PDS/FeCl₂·4H₂O solutions was characterized using a UV–vis–near infrared spectrophotometer (UH5300, Hitchi High-Tech Corporation, Japan) and a Raman spectrometer (WITec Alpha, WITec, German) using Zeiss LD EC Epiplan-Neofluar HD Dic 50×/0.55 as the objective lens, respectively.

¹⁸O isotope tracing experiments

One milliliter of H₂¹⁸O solution containing 3 mm of PMSO was prepared. A total of 0.163 g of PDS and 0.137 g of FeCl₂·4H₂O were added to form the sample after a predetermined time of 15 min, during which the transformation of PMSO was completed. ¹⁶O/¹⁸O isotope–labeled PMSO₂ was detected by a UPLC-Q-TOF MS (Waters Co., USA). For the precise determination of PMSO₂, ¹⁸O isotope PMSO₂ and/or other oxidation products of PMSO in PDS/FeCl₂·4H₂O system, a BEH C18 column (100 mm by 2.1 mm in inner diameter, 1.7 μm; Waters, Milford, USA) was used. The gradient mobile phase ratio of A/B was set as: the ratio kept at 95/5 for the first 5 min, then changed linearly from 95/5 to 5/95 in the next 15 min, and held for 10 min, followed by a sharp decline to 95/5 in 0.1 min, and kept for 10 min for reequilibration, where A represents ultrapure water and B represents acetonitrile with a flow rate of 0.15 ml/min. Accurate MS and tandem mass spectrometry spectra of PMSO₂ and/or other products were analyzed in a molecular ion scanning mode (m/z, 50 to 600) in negative ESI mode.

Theoretical calculations

All calculations were performed using the software Gaussian 16 (78). All calculations involved the implementation of the PBE0 hybrid functional in conjunction with the D3BJ dispersion correction. For geometry optimization, the mixed basis set (BS1) of def2-SVP for the Fe element and 6-31+G(d) for all other atoms (79), along with the IEFPCM (perfect coverage model using the intergral equation formalism model) solvent model for water, was used. Without any structural constraints, the geometries have been completely optimized. Using the SMD (a continuum solvation model based on the quantum mechanical charge density of a solute molecule interacting with a continuum description of the solvent) continuum solvation model with the larger mixed basis set (BS2) of def2-TZVP for Fe and 6-311+G(d,p) for all other atoms, the final and solvation energies of the fully optimized structures in water were calculated.

Recovering Pd and Au from spent catalysts

We opted for waste Pd catalyst for combustion exhaust gas treatment process and waste Au catalyst for methacrolein esterification. The spent Pd catalyst was extracted from a chemical facility, whereas the Au catalyst was collected from a laboratory. To assess the metal contents in Pd and Au catalysts, microwave digestion of the waste Pd and Au catalysts was first performed, followed by ICP-OES to test the metal composition in the two catalysts. After electrodeposition, the Pt sheets of the working electrodes and the PM product obtained on the electrodes were subjected to microwave digested together and then tested by ICP-OES to calculate the recovery efficiency and purity of the recovered PMs.

Supplementary Materials

This PDF file includes:

Supplementary Text

Notes S1 to S9

Figs. S1 to S22

Tables S1 to S13

References