A novel green approach for Au/Pt recovery from lean solutions through photodeposition and alkaline leaching on WO₃ nanoparticles

Abstract

The efficient extraction and recycling of noble metal resources have become an inevitable trend for sustainable socio-economic development. Traditional methods suffer from lengthy processes, high energy consumption, and serious environmental hazards, necessitating the development of new, environmentally friendly, cost-effective, and efficient methods for noble metal extraction. This study primarily employs a photochemical method to reduce and enrich low-concentration Au/Pt noble metal solutions on the surface of the WO₃ semiconductor, followed by alkali leaching and acidification steps to achieve Au/Pt extraction and WO₃ recycling. It can complete the enrichment of noble metals within 20 min, with a deposition efficiency of over 99% and a total recovery efficiency of over 93%, suitable for noble metal concentrations ranging from 0.01 to 10 mM. The extracted Pt particles have a diameter of approximately 5 nm, and the Au particles have a diameter of about 60 nm, both with a purity of 99.9%. The acid-regenerated WO₃ semiconductor can be reused cyclically, maintaining stable performance. This study establishes a novel technology for the recycling and recovery of noble metals from lean solutions based on semiconductor photochemistry, holding significant scientific and practical value for noble metal resource recycling.

---

1. Introduction

Noble metal gold (Au) possesses excellent electrical and thermal conductivity,^1,2 chemical stability,³ and biocompatibility,⁴ and is widely used in the electronics industry,⁵ nuclear industry⁶ and healthcare sector.^7–9 Resources of noble metals like Au are scarce, with very low annual production, relying mainly on secondary resource recovery.^10–12 Traditional methods for recovering noble metals include displacement methods^13,14 and extraction methods.¹⁵ In recent years, new methods such as carbon adsorption^16,17 and hydrogel adsorption¹⁸ have emerged, but these methods face issues of low extraction efficiency, low extraction purity, and long experimental cycles. Currently, in the waste liquid from copper smelting, the concentration of Au ions is about 20 mg L⁻¹ (approximately 0.1 mM), which is very low, making efficient recovery difficult with existing methods.¹⁹

In traditional catalytic industries, noble metals are frequently deposited onto semiconductor surfaces through photochemical methods to enhance photocatalytic performances.^20–27 Naphaphan et al.²⁸ prepared Au/TiO₂ composite photocatalysts by reducing Au ions with TiO₂ under UV irradiation, achieving a reduction time of less than 10 min and an Au ion concentration in the precursor of just 50 mg L⁻¹. This study demonstrated that photocatalysts can reduce low concentrations of Au ions in solution through a process that is both environmentally friendly and efficient. Joseph et al.²⁹ utilized ZnO as a substrate and sequentially deposited minute quantities of Au and Pt onto its surface. Their experiment indicated that semiconductor catalysts can recover various noble metals such as Au and Pt, highlighting the environmental benefits of photocatalysis. Zheng et al.³⁰ reduced Au ions in a 0.1 mM HAuCl₄ solution using UV light, depositing them onto ZnO, and prepared ZnO/Au materials within 10 min, effectively reducing Au ions at very low concentrations. According to the aforementioned studies, photodeposition methods can efficiently deposit noble metals onto semiconductor surfaces under low concentration conditions, and this technology has matured in the preparation of noble metal-modified semiconductor catalysts.³¹ However, to achieve efficient separation of photogened electrons and holes, the loading amount typically remains low. When applied to noble metal extraction, this technology's extraction efficiency and maximum loading capacity remain unknown.

Posachayanan et al.³² used TiO₂-AC as a photocatalyst to recover noble metal Ag from electroplating wastewater under UV conditions, recovering 94% of the Ag in the wastewater within 45 min. Muscetta et al.³³ used ZnO to deposit Pd from wastewater under UV conditions, completing the deposition of 0.1–1 mM Pd ions within 90 min. The loading of Pd can further enhance photocatalytic performance, and subsequent acidification was used to separate Pd–ZnO. It is worth noting that the reduction potentials of TiO₂ and ZnO are relatively negative, approximately −0.3 V (vs. NHE).³⁴ Theoretically, if non-noble metals are present in the solution system, they will also be reduced, ultimately leading to the extracted noble metals' purity being compromised. Moreover, ZnO is unstable under acidic conditions, while waste liquids in actual industrial production are mostly acidic. Although the aforementioned methods offer valuable insights for our research, catalysts like TiO₂ and ZnO are not suitable for recovering noble metals from acidic waste solutions. The reduction potential of WO₃ is approximately 0.7 V (vs. NHE),³⁵ situating it between noble metals and non-noble metals. Theoretically, it can reduce noble metals such as Au without affecting non-noble metals like Fe, Co, and Ni. WO₃ is relatively stable under acidic conditions (pH < 3),³⁶ facilitating effective Au deposition. It readily dissolves under alkaline conditions (pH > 8),³⁷ enabling the separation and purification of noble metals from WO₃, while also allowing for its recycling and regeneration.

In this study, WO₃ was selected as the catalyst, and nanoscale WO₃ was prepared via the calcination method. The photoreduction method was employed to extract Au ions from acidic solutions, investigating the applicable concentration range and the kinetics of Au ion photodeposition. Through conditional experiments, the effects of varying WO₃ content, pH levels, and ethanol concentrations on the photodeposition effect were studied, and the optimal process parameters for Au recovery were determined. Additionally, the effectiveness of recovering and extracting Au/Pt mixed ions and the repeated recovery performance of WO₃ for Au/Pt were evaluated. Finally, the recovered samples underwent alkaline leaching to purify the noble metals. The regeneration and recycling of WO₃ were achieved through the acidification-calcination method. An economic assessment of the entire experimental process and technical methods demonstrated significant economic advantages of the technology used in this study. Compared to traditional methods, this technology offers efficient, green, and low-cost recovery and extraction of noble metals, which is significant for the recycling of noble metal resources in China.

2. Experimental methods

2.1 Materials

Polyvinyl pyrrolidone (C₆H₉NO), polyethylene glycol (H(OCH₂CH₂)_nOH), sodium hydroxide (NaOH), and ammonia solution (NH₃·H₂O) were all sourced from Tianjin Damao Chemical Reagent Factory. Ammonium metatungstate ((NH₄)₆H₂W₁₂O₄₀·xH₂O) was purchased from Sinopharm Chemical Reagent Co., Ltd. Hexachloroplatinic acid hexahydrate (H₂PtCl₆·6H₂O) and tetrachloroauric acid tetrahydrate (HAuCl₄·4H₂O) was purchased from Nangong Bo Le Metal Materials Co., Ltd. Anhydrous Ethanol (CH₃CH₂OH) was sourced from Tianjin Bohua Tong Chemical Co., Ltd.

2.2 Preparation of WO₃ nanoparticles

Dissolve 2.00 g of polyvinyl pyrrolidone (PVP K30) in 15.0 mL of deionized water to prepare solution A. Dissolve 1.478 g of ammonium metatungstate (AMT) in 10.0 mL of deionized water to prepare solution B. At room temperature, mix solution A with solution B, then sonicate for 30 min, followed by stirring for 2 h. Add 2.00 g of polyethylene glycol (PEG 1000) to the mixture, and continue stirring at room temperature for an additional 4 h. Transfer the resulting sol into a porcelain boat, dry it at 80 °C, and heat-treat at 600 °C for 1 h.

2.3 Experimental conditions for photodeposition of Au

Solutions of Au ions at initial concentrations of 1.0, 0.5, 0.1, 0.05 and 0.01 mM were introduced into a quartz glass reaction cell. Subsequently, 0.25 g of WO₃ nanoparticles were added for the photodeposition reaction. The suitable range of Au ion concentrations for photochemical reactions was determined based on experimental results. Ethanol (C₂H₅OH) at concentrations of 0, 0.1, 0.2, 0.5, 1.0 and 2.0 M was then added, alongside varying amounts of WO₃ (2.0, 1.0, 0.5, 0.25 g L⁻¹), with pH adjustments (2.0, 1.0, 0.5) to determine optimal experimental conditions.

After determining the optimal conditions for the photocatalytic reaction of Au ions, a mixed solution containing Pt and Au ions, each at a concentration of 0.1 mM, was prepared. Ethanol was added to achieve a concentration of 0.2 M, followed by the addition of 0.25 g of WO₃ nanoparticles, and the pH was adjusted to 1. Following photodeposition reaction, the samples were filtered, washed, and subsequently dried in a vacuum oven at 70 °C for 6 h.

2.4 Determination of Au ion and Pt/Au ion concentrations in solution

2.4.1 Determination of Au ion concentration. To determine the variation in Au ion concentration during deposition, a standard curve was established using spectrophotometry. Five standard solutions of Au ions with concentrations of 0.001 mM, 0.005 mM, 0.01 mM, 0.05 mM, and 0.10 mM were prepared, covering a concentration range from 0.01 to 1.0 mM (2–200 mg L⁻¹). The absorbance of these standard solutions at characteristic wavelengths was measured using a UV spectrophotometer to establish the standard curve and corresponding equation.

Fig. 1a illustrates the relationship between absorbance and wavelength for standard solutions with 5 different initial Au ion concentrations. The Au ion concentrations were 0.001 mM, 0.005 mM, 0.01 mM, 0.05 mM, and 0.1 mM, respectively. Absorption peaks were observed at both 227 nm and 313 nm, but the correlation at 227 nm was lower (R² = 0.97). Therefore, 313 nm was chosen as the characteristic peak. Fig. 1b depicts the relationship between time and absorbance at 313 nm for an Au ion concentration of 0.001 mM. The results indicated that the absorbance remained constant over time, suggesting that UV light had no effect on the sample during testing. This confirms the feasibility of using spectrophotometry to measure Au ion concentration. Fig. 1c illustrates the relationship between different initial Au ion concentrations and absorbance. The results demonstrated a linear relationship between absorbance and Au ion concentration, with an equation of y = 5.6906x + 0.0075 and a correlation coefficient of R² = 99.91%, indicating a very high correlation. In the experiment, the Au ion concentration at a specific time can be calculated based on the absorbance of the solution at 313 nm.

Fig. 1 (a) The relationship between absorbance and wavelength for various initial concentrations of Au ions, (b) the relationship between time and absorbance at a wavelength of 313 nm for an Au ion concentration of 0.001 mM and (c) the standard curve illustrating the relationship between absorbance and the concentration of Au ions at 313 nm.

2.4.2 Determination of Pt/Au ion concentration. Five standard solutions were prepared with Pt/Au ion concentrations of 0.001 mM, 0.005 mM, 0.01 mM, 0.05 mM, and 0.1 mM, covering a concentration range of 0.01–1.0 mM and 2–200 mg L⁻¹. The absorbance of these standard solutions was measured at characteristic wavelengths using a UV spectrophotometer to establish the standard curve and equation for Pt/Au mixed ions.

Fig. 2a illustrates the relationship between absorbance and wavelength for standard solutions with 5 distinct initial Au ion concentrations: 0.001 mM, 0.005 mM, 0.01 mM, 0.05 mM, and 0.1 mM. Absorption peaks are observed at 205 nm and 262 nm, corresponding to the characteristic peaks of Pt ions and Au ions, respectively, in the coexisting ion solution. Fig. 2b shows the relationship between time and absorbance at an Au ion concentration of 0.001 mM and a wavelength of 262 nm. The graph demonstrates a relatively stable curve, indicating that UV light exerted no effect on the test solution during the experiment, thereby confirming the feasibility of spectrophotometry for determining ion concentration. Fig. 2c shows the standard curve of Pt ion concentration versus absorbance in the Pt/Au ion coexisting solution. The equation is y = 26.3035x + 0.0175, with a variance of R² = 99.94%. Fig. 2d illustrates the standard curve of Au ion concentration versus absorbance in the Pt/Au ion coexisting solution. The equation is y = 41.7405x + 0.1926, with a variance of R² = 97.53%.

Fig. 2 (a) The relationship between absorbance and wavelength for different initial concentrations of Pt/Au ions, (b) the relationship between time and absorbance at 262 nm for a Pt/Au ion concentration of 0.001 mM, (c) the standard curve showing the relationship between absorbance and concentration for Pt ions at 205 nm and (d) the standard curve for the relationship between absorbance and concentration of Au ions at 262 nm.

It should be noted that all the metal ion concentrations measured by spectrophotometers had also been verified by ICP test, indicating the feasibility and reliability of the spectroscopy method.

2.5 Multiple deposition and reduction of Pt/Au

The concentration of Pt/Au ions was adjusted to 10.0 mM, ethanol to 1.0 M, and the pH was set to 1.0. This solution was placed in a photochemical reactor, with 0.25 g of WO₃ nanoparticles added, and the photodeposition reaction was conducted for 5 h. After the reaction, the mixture was filtered, and the solid product was vacuum-dried at 70 °C for 6 h before collection. The collected solid product was subsequently used as the photocatalyst for additional rounds under identical solution preparation and experimental conditions, repeating the above steps four or more times.

2.6 Alkaline leaching extraction of Au and Pt/Au and the regeneration of WO₃

2.6.1 Alkaline leaching extraction of Au and Pt/Au. Increase the Au/Pt ion concentration by 100 times for deposition, recover the obtained solid, and place it in a 100 mL beaker. Add 40 mL of 0.1 M NaOH solution and stir for 12 h. Collect the solid products, centrifuge at 10 000 rpm for 6 times, and dry them in a vacuum oven at 70 °C for 24 h to obtain the alkali-treated solid samples: Au and Pt/Au.

2.6.2 Regeneration of WO₃ nanoparticles. Add 3.0 M hydrochloric acid into supernatants from the alkali treatment, adjusting the pH to 1, and stir until fully reacted. Heat the mixture in a water bath at 60 °C for 3 h until the solution turns yellow. Transfer to an oven, dry at 80 °C for 12 h. Finally, place the sample in a muffle furnace at 600 °C for 2 h to obtain the regenerated WO₃ nanoparticles, marking as R-WO₃.

The second method for WO₃ regeneration: Take 1 g of the sample, add 20 mL of ammonia solution to dissolve, and stir at room temperature for 4 h to form solution A. Dissolve 2.0 g of PVP in 15.0 mL of deionized water to form solution B. Add solution B to solution A at room temperature, and stir at room temperature for 2 h. Add 2.0 g of PEG 1000 to the solution and stir at room temperature for 4 h. Pour the sol into a porcelain boat, dry at 80 °C, and then heat-treat it at 600 °C for 2 h to obtain the regenerated WO₃ nanoparticles, marking as RA-WO₃.

The regenerated WO₃ nanoparticles were used for photodeposition to enrich and extract Au. The whole experimental procedure is shown in Fig. 3.

Fig. 3 Experimental process for the recovery of Pt/Au and schematic diagram of WO₃ recycling and regeneration.

2.7 Characterization methods

X-ray diffraction (XRD) analysis employs a D/MAX-2400 X-ray diffractometer with a copper target K_α radiation source to identify the phases of the sample. Microstructure of the prepared powder was analyzed by a field emission scanning electron microscope (FESEM, JSM-6700), operating at 3–5 kV and 5–7 μA. Particle microstructure and distribution was observed by high-resolution transmission electron microscopy (HRTEM, JEM-2010). UV-vis spectrophotometer (UV756) was employed to detect ion concentration by measuring absorbance. X-ray photoelectron spectroscopy (XPS) employed a Thermo Scientific ESCALAB 250Xi instrument to determine the chemical states of surface elements and identify the forms of noble metal elements.

3. Results and discussion

3.1 The kinetic process of depositing noble metals Au and Pt/Au on nano WO₃

A photochemical method was utilized to treat low-concentration Au ion solutions (0.01–1.0 mM, 2–200 mg L⁻¹), leading to the deposition and extraction of the noble metal Au under ultraviolet light irradiation. The effect of varying initial concentrations of Au ions on the photodeposition of Au was investigated under conditions of pH 1, WO₃ concentration at 0.5 g L⁻¹, and ethanol concentration at 1.0 M, see Fig. 4a–e. During the initial 30 min, Au ions were adsorbed by WO₃ in the dark, with results indicating that WO₃ exhibited low adsorption efficiency for Au ions. At all five concentrations, complete photodeposition within 10 min, with Au particles located on the surface of WO₃, indicating a self-catalytic effect. Higher initial concentrations facilitated photodeposition, with the initially deposited Au acting as nucleation sites, thereby effectively reducing the overpotential.

Fig. 4 Photodeposition kinetic curves for different initial Au concentrations: (a) 1.0 mM, (b) 0.5 mM, (c) 0.1 mM, (d) 0.05 mM and (e) 0.01 mM. Photodeposition kinetic curves of Au: (f) different amounts of ethanol, (g) different pH values, (h) different amounts of WO₃, and photodeposition kinetic curves at an initial concentration of 0.1 mM for Pt/Au: (i) Pt and (j) Au.

The impact of varying ethanol addition amounts on the photodeposition of Au under initial conditions of 0.1 mM Au concentration, pH = 1.0, and 0.5 g L⁻¹ WO₃ concentration is shown in Fig. 4f. The total photoreduction time for 0.1 M ethanol is 20 min, whereas for 0.2 to 0.5 M ethanol, the time is 10 min. Ethanol acts as a hole scavenger, preventing electron–hole pair recombination and significantly accelerating the efficiency of electron photoreduction of Au.³⁸ The concentration of ethanol has a significant impact on the photodeposition reaction. At low ethanol concentrations, holes are not fully captured, resulting in partial recombination of electron–hole pairs and a reduced photodeposition efficiency. As ethanol concentration increases, the hole capture efficiency accelerates, thereby increasing the photodeposition efficiency. In industrial production, however, balancing economic efficiency requires optimal conditions that ensure complete hole capture while maintaining a high photodeposition efficiency. Therefore, the optimal ethanol concentration is determined to be 0.2 M.

The influence of varying pH levels on the photodeposition of Au is shown in Fig. 4g. At pH = 1.0, the total photoreduction time is 10 min, whereas it increases to 20 min at pH = 0.5 or 2. The pH value affects the surface charge properties of the photocatalyst WO₃, thus influencing the adsorption and reduction of Au ions. At pH = 1.0, both adsorption and reduction processes are optimal. Furthermore, pH also influences the state of Au ions in the solution. The pH value influences the surface charge properties of the photocatalyst WO₃, thereby affecting its adsorption and reduction of Au ions. According to the Nernst equation (Formula (1)) and the relationship between pH value and H⁺ concentration (Formula (2)),³⁹ the pH value influences the surface potential of WO₃. As the pH value increases, the surface potential decreases, leading to fewer photo-generated electrons and reduced photodeposition efficiency. However, within the pH range of 0.5 to 2.0, variations in pH have minimal impact on photodeposition efficiency.

E = E^Θ + 0.059 log[H⁺] (1)

pH = −log[H⁺] (2)

Fig. 5 presents the potential–pH diagram of the Au–H₂O system. Au exists stably in ionic form within the pH range of 0.5 to 2.0. The reduction potential from Au³⁺ to Au is 1.40 V, and the pH values in this range do not affect the stability of Au ions.⁴⁰

Fig. 5 Potential–pH diagram of Au–H₂O.⁴⁰

Fig. 4h illustrates the impact of varying WO₃ concentrations on the photodeposition of Au. For catalyst concentrations of 0.25 g L⁻¹ and 1.0 g L⁻¹, the time required for complete photoreduction is 20 min. However, a concentration of 0.5 g L⁻¹ results in a reduction time of only 10 min, yielding the most favorable results. At low catalyst concentrations, the photodeposition process is inefficient, conversely, excessively high catalyst amounts can obstruct light, impairing light absorption and consequently diminishing the photodeposition efficiency. Under optimal conditions, the reduction efficiency of Au achieved 97.5%. When the WO₃ catalyst content is low, the number of photoexcited electrons is insufficient to efficiently complete the photodeposition process. Conversely, when the catalyst content is high, it can block some light, affecting light absorption and thus reducing the photodeposition efficiency. However, within the range of 0.25 to 1.0 g L⁻¹, the amount of WO₃ has little effect on the photodeposition efficiency.

Fig. 4i and j presents photodeposition kinetics curve for Pt/Au at an initial concentration of 0.1 mM. Upon initiation of the photoreaction, the reaction efficiency increases rapidly. After 10 min of photoreaction, Pt ions are nearly completely photoreduced, reducing the concentration in the solution to 0.001 mM, corresponding to a reduction efficiency of 99%. After 2 h, the Pt ion concentration further decreases to 0.0004 mM, a reduction efficiency of 99.6%. For Au, after 10 min of photoreaction, the ions continue photoreduction at a slower efficiency compared to Pt ions. After 20 min, Au ions are nearly completely photoreduced, reducing the concentration to 0.001 mM, with a reduction efficiency of 99%. After 2 h, the Au ion concentration further decreases to 0.0005 mM, achieving a reduction efficiency of 99.5%. When Pt and Au ions coexist at low concentrations, the aforementioned experimental method still achieves high efficiency.

3.2 Structural characteristics of Au and Pt/Au deposited on nano WO₃

Fig. 6a shows the XRD results of Au/WO₃ after recovering WO₃ nanoparticles at different initial concentrations of Au ions (0.01 mM, 0.05 mM, 0.1 mM, 0.5 mM, 1.0 mM). No characteristic peaks of Au or its compounds were observed at Au ion concentrations of 0.01 mM and 0.05 mM, mainly due to the low concentration and insufficient quantity of Au ions. At Au ion concentrations of 0.1 mM, 0.5 mM, and 1.0 mM, a characteristic peak of Au (PDF#04-0784) associated with the (111) crystal plane was observed at 38.2°, showing a strong intensity for elemental Au at this plane. As the concentration of Au ions in the solution increased, the peak intensity of Au in the recovered solid gradually increased, while the relative peak intensity of WO₃ decreased, indicating a corresponding increase in the amount of Au loaded onto WO₃. As presented in Fig. 6b, the peaks of WO₃ after Au deposition shifted slightly to lower angles, with the degree of shift increasing at higher Au ion concentrations. This suggests that during the photodeposition process, a small number of Au atoms replaced W atoms under UV radiation. Given that the radius of Au atoms is larger than that of W, this caused lattice expansion, resulting in the peaks shifting to lower angles. No significant differences were observed between the WO₃ after photodeposition and the original sample, indicating that the phase of WO₃ remained stable during the photodeposition process, thus facilitating the stable deposition of Au ions.

Fig. 6 (a) XRD patterns of Au/WO₃ after photodeposition with different initial Au concentrations, (b) enlarged view of the dotted section, (c) XRD pattern of Pt/Au/WO₃, (d) SEM image of WO₃ nanoparticles after Au deposition and (e) Pt/Au deposition, (f) EDS surface elements distribution map after photodeposition of Au on WO₃, and (g) Pt/Au on WO₃.

As described in Fig. 6c, the XRD characterization results of Pt/Au/WO₃ obtained after photodeposition a Pt/Au solution with an initial concentration of 0.01 mM. The diffraction peaks corresponding to standard WO₃ (PDF#20-1324) and standard Au (PDF#04-0784) are evident in this figure. Elemental Au primarily crystallizes along the (111) plane. No diffraction peaks corresponding to Pt are observed, mainly due to the low concentration of Pt in the solution and the small size and good dispersion of the Pt particles post-deposition. As presented in Fig. 6d, the SEM image of WO₃ nanoparticles after Au deposition, while Fig. 6e shows the SEM image after Pt/Au deposition. The samples are uniformly dispersed and exhibit a granular morphology, with the size of the deposited Au particles estimated to be below 20 nm. In contrast, the Pt particles are even smaller, and agglomeration is more pronounced in their presence.

Fig. 6f shows the EDS surface scan distribution of Au deposited on WO₃ nanoparticles, revealing a W content of 29.0 wt%, an O content of 70.8 wt%, and an Au content of 0.2 wt%. Despite the low Au content, it is uniformly distributed across the surface of WO₃. Fig. 6g illustrates the EDS surface scan distribution following the photodeposition of Pt/Au, with a W content of 40.2 wt%, an O content of 57.5 wt%, a Pt content of 1.1 wt%, and an Au content of 1.2 wt%. Both Pt and Au are uniformly distributed on the surface of WO₃, indicating successful incorporation of Pt and Au onto the WO₃ particles.

Fig. 7a presents the transmission electron microscopy (TEM) characterization of WO₃ nanoparticles following Au deposition. The image demonstrates that the particle sizes are heterogeneous, with a maximum dimension of approximately 80 nm and the smallest particles measuring less than 20 nm. Fig. 7b displays a high-resolution transmission electron microscopy (HRTEM) image, illustrating smaller particles deposited onto larger ones. The HRTEM characterization depicted in Fig. 7b reveals the crystal structure of WO₃ with deposited Au. In the fast Fourier transform (FFT) image of the smaller nanoparticles, illustrated in inset Fig. 3, the corresponding [1̄1̄0] zone axis of Au exhibits crystal planes of (1̄11̄) and (11̄1̄). In regions ① and ②, the corresponding zone axis of WO₃ is [001], associated with crystal planes of (200) and (02̄0), where a limited number of impurity particles are observed. In region ④, the corresponding zone axis of WO₃ is [11̄0], featuring crystal planes of (002) and (110). Fig. 7b confirms the presence of Au following the photodeposition reaction, with diffraction results from region ③ verifying that the smaller particles are indeed Au, indicating that Au is deposited onto the surface of the larger WO₃ particles. The average grain size of WO₃ is measured at 75 nm, while the minimum grain size of Au is 10 nm.

Fig. 7 (a) TEM image and (b) HRTEM image of WO₃ nanoparticles after Au deposition: ①–④ are FFT images of different regions, (c) TEM image and (d) HRTEM image after Pt/Au deposition: ⑤–⑧ are FFT images of different regions, (e) TEM surface elements distribution map after photodeposition of Au, and (f) TEM surface elements distribution map after photodeposition of Pt/Au, (g) XPS survey spectrum of WO₃ nanoparticles after Au deposition, (h) Au 4f, (i) W 4f and (j) O 1s, and (k) XPS spectrum of WO₃ nanoparticles after Pt/Au deposition, (l) Pt 4f and (m) Au 4f.

Fig. 7c presents a transmission electron microscopy (TEM) image following the deposition of Pt/Au, revealing small particles aggregating at one end of the larger particles. Fig. 7d displays a high-resolution transmission electron microscopy (HRTEM) image of the sample. The HRTEM image in Fig. 7d characterizes the crystal structure of WO₃ with the deposited Pt/Au. The fast Fourier transform (FFT) patterns of the larger nanoparticles, as illustrated in inset patterns of the larger nanoparticles, as illustrated in inset Fig. 7⑥ and ⑦, correspond to the [110], [5̄12], and [62̄3̄] zone axes of WO₃, with the associated crystal planes being (002), (22̄0), (112), (02̄1), (130), and (1̄02). The region in inset Fig. 7⑤ illustrates the zone axes of WO₃ and Au, where the zone axes of WO₃ are [2̄30] and [2̄10], corresponding to the crystal planes (120), (002), (002), and (322), while the zone axis of Au is [210], corresponding to the crystal planes (022̄) and (11̄1̄). The region in inset Fig. 7⑧ corresponds to the [310] zone axes of Pt and Au, with the associated crystal planes being (002), (131), (13̄1̄), and (002̄). Additionally, a small quantity of impurity particles is present, with their specific composition and elemental characteristics pending further investigation. Fig. 7d effectively demonstrates the presence of Pt/Au following the photodeposition reaction. Diffraction analysis of region 8 indicates that the composition comprises pure Pt and Au, with no formation of solid solutions or, and is supported on larger WO₃ particles.

Fig. 7e shows the TEM mapping distribution after the photodeposition of Au, with yellow regions indicating the presence of Au. Au is uniformly distributed across the mapping area and grows on the WO₃ surface, occasionally showing epitaxial growth at the edges of WO₃. The small yellow regions in panel (a) represent growing Au nanoparticles, corresponding to inset Fig. 7③. Fig. 7f presents the TEM mapping distribution after the photodeposition of Pt/Au, with yellow regions indicating Au and showing a uniform overall distribution. The purple sections indicate the presence of Pt, with some areas appearing in an agglomerated form, consistent with the EDS mapping results and corresponding to inset Fig. 7⑧.

Fig. 7g presents the XPS spectrum of WO₃ nanoparticles following the deposition of Au. In the sample, only the elements O, W, and Au were detected, along with a C peak used for calibration, indicating the absence of other impurities. Fig. 7h displays the Au 4f spectrum, where the two peaks corresponding to Au 4f_7/2 and 4f_5/2 are observed at 84.5 eV and 88.2 eV, respectively, thereby confirming the presence of elemental Au.^41–43 The research findings indicate that Au predominantly exists as elemental Au in the Au/WO₃ composite material, with no other chemical states identified. In Fig. 7i, the W4f_7/2 and W4f_5/2 peaks are observed at 36.1 eV and 38.1 eV, respectively, indicating the presence of W⁶⁺, which denotes hexavalent tungsten in WO₃. Additionally, characteristic energy loss peaks of WO₃ are also observed. Fig. 7j presents the O 1s peak, which represents O²⁻ and corresponds to the oxygen in WO₃. At the O 1s peak of 532.8 eV, smaller peaks corresponding to water molecules and hydroxyl groups are observed.⁴⁴ The binding energy of W 4f increased by approximately 0.4 eV when compared to the normal peak position. In contrast, the binding energy of O 1s increased by 0.3 eV. This observation suggests that the presence of Au results in a slight increase in the binding energies of both W and O.

Fig. 7k presents the XPS survey spectrum of WO₃ nanoparticles following the deposition of Pt/Au. In addition to the calibration peak C 1s, only the elements W, O, Pt, and Au were detected, with no other impurities identified. Fig. 7l illustrates the Pt 4f spectrum, with the two peaks of Pt 4f_7/2 and Pt 4f_5/2 observed at 71.6 eV and 74.9 eV, respectively. Due to the small size of the Pt particles, some of the Pt is oxidized to form PtO, resulting in corresponding peaks at 73.1 eV and 76.6 eV. Fig. 7m shows the Au 4f_7/2 and Au 4f_5/2 peaks, which are located at 84.5 eV and 88.2 eV, respectively,^41–43 given the significant and stable size of Au particles, no compounds other than the elemental form were detected. The results demonstrate that WO₃ nanoparticles effectively deposited Pt/Au onto their surface.

3.3 Multiple photodeposition of Pt/Au

The initial concentration of Pt/Au was 10.0 mM, and the XRD characterization after 5 photodeposition cycles is shown in Fig. 8a. During the first deposition of Pt/Au, peaks of WO₃ (PDF#20-1324), Pt (PDF#87-0646), and Au (PDF#04-0784) were observed. Both Pt and Au preferentially grew on the (111) plane, but the peak intensity of Au was higher than that of Pt. The Pt/Au/WO₃ nanoparticles after the first deposition were used as the photocatalyst for the second deposition of Pt/Au. From the XRD patterns, it can be seen that during the second and subsequent depositions of Pt/Au, the peaks of WO₃ gradually decreased in number and size, due to the decreasing proportion of WO₃. By the fifth deposition, the peaks of WO₃ were no longer observable, and only the peaks of Au and Pt were present. The relative content of (111) plane grains of Au and Pt decreased. This phenomenon indicates that when the relative content of WO₃ is higher, Au and Pt are more likely to grow on the (111) plane, when the relative content of WO₃ decreases, the content of (111) plane grains of Au and Pt also decreases. With the increase in deposition cycles, the peak intensity of Pt gradually surpassed that of Au. Additionally, as the number of depositions increased, the deposition efficiency of Pt decreased more slowly compared Au. As demonstrated in Fig. 8b, the Au peak slightly shifts to a higher angle, whereas the Pt peak remains unchanged. Given that Pt and Au have similar atomic radii,⁴⁵ and comparable electro-negativities (ranging from 2.1 to 2.5),^46,47 a small amount of Pt replaces Au at some lattice sites during their co-deposition. Due to Pt's smaller atomic radius compared to Au, lattice contraction occurs, which results in the Au peak shifting to a higher angle. The EDS and surface distribution images following 5 photodeposition cycles at an initial Pt/Au concentration of 10.0 mM are presented in Fig. 8c and d. The results show that the W content is 0.05 wt%, O content is 0.30 wt%, Pt content is 55.97 wt%, and Au content is 43.68 wt%. The Pt content is slightly higher than that of Au, aligning with the XRD results from the fifth deposition. However, because of the uneven distribution and agglomeration of Pt and Au, they cannot be directly observed.

Fig. 8 (a) XRD pattern after 5 photodeposition cycles with an initial Pt/Au concentration of 10.0 mM and (b) enlarged view, (c) EDS map and surface spectrum after 5 photodeposition cycles with an initial Pt/Au concentration of 10.0 mM, (d) energy spectrum, and photodeposition kinetics with an initial Pt/Au concentration of 10.0 mM: (e) Pt and (f) Au.

The photodeposition kinetic curves for an initial Pt/Au concentration of 10.0 mM are presented in Fig. 8e and f. For Pt ions, the kinetic curves from the first three depositions indicate a progressive increase in reaction efficiency. This occurs because, during the reaction, the active sites transition from WO₃ to Pt/Au/WO₃, which enhances hole capture capacity and accelerates the reaction efficiency. However, in the fourth and fifth depositions, the reaction efficiency decreases due to the saturation of active sites, impeding the photocatalytic reaction. For Au ions, the kinetic curves from the first three depositions also indicate an increasing reaction efficiency. Compared to pure Au ions, the Pt/Au/WO₃ active sites inhibit Au aggregation, promote dispersed growth, reduce the size of Au flakes, and improve light utilization and reaction efficiency. However, during the fourth and fifth depositions, the reaction efficiency decreases because the increased Au content expands the Au flake area, leading to greater light reflection, reduced light utilization, and ultimately impeding the photocatalytic reaction.

Based on the changes in the concentration of noble metal ions during each deposition, the cumulative deposition amounts of Au and Pt on WO₃ over 5 depositions were calculated, as shown in Table 1. The concentrations of Pt and Au ions were both 10 mM, with each deposition lasting 8 h. After the first deposition, the deposition amounts of Au and Pt were 3.14 g g⁻¹ WO₃ and 3.89 g g⁻¹ WO₃, respectively. The amount of noble metals was more than three times that of the catalyst. After the fifth deposition, the deposition amounts of Au and Pt were 15.70 g g⁻¹ WO₃ and 19.47 g g⁻¹ WO₃, respectively. This demonstrates that even after multiple cycles, WO₃ still has a strong deposition capacity for noble metals.

Table 1 The cumulative deposition amounts of Au and Pt on WO₃ over 5 cycles

Noble metals	The cumulative loading amount of noble metals on WO3 (g per g WO3)
	1 st	2 nd	3 rd	4 th	5 th
Au	3.14	6.28	9.42	12.56	15.70
Pt	3.89	7.79	10.88	15.57	19.47

---

3.4 Alkaline leaching extraction of Au and Pt/Au

Following the alkaline leaching treatment, the supernatant containing sediment was analyzed using ICP testing. As shown in Table 2, the concentration of Au in the supernatant of WO₃ nanoparticles deposited with Au after alkaline leaching is 0.0001 g L⁻¹ (0.1 ppm), which is negligible. At this stage, WO₃ is dissolved by NaOH, resulting in the formation of Na₂WO₄, with a W concentration of 1.9872 g L⁻¹. After alkaline leaching, the concentration of Au in the supernatant of WO₃ nanoparticles deposited with Pt/Au is 0.0027 g L⁻¹, while the Pt concentration is 0.0110 g L⁻¹. In comparison, the dissolution amount of WO₃ is lower in this case.

Table 2 ICP testing of the supernatant after alkaline leaching

Solid samples	Concentrations of various elements in the solution after alkaline leaching
	Pt (g L^−1)	Au (g L^−1)	W (g L^−1)
WO3 nanoparticles with deposited Au	—	0.0001	1.9872
WO3 nanoparticles with deposited Pt–Au	0.0110	0.0027	2.3000

---

The Au recovered and extracted after alkali leaching is shown in Fig. 9a, and the sample appears golden yellow. Fig. 9c shows the XRD spectrum of the Au recovered and extracted after alkali leaching. From the figure, it can be seen that the measured peaks correspond to standard Au (PDF#04-0784), corresponding to the (111), (200), (220), and (311) crystal planes, confirming the sample as pure Au, with the peak of the (111) crystal plane being the strongest. According to the Debye–Scherrer formula,⁴⁸ the grain sizes of the crystal planes are approximately 49.3 nm for (111), 39.4 nm for (200), 40.9 nm for (220), and 36.9 nm for (311). Fig. 9b shows the Pt/Au recovered after alkali leaching, appearing gray and indicating a mixture of Pt and Au. Fig. 9d displays the XRD pattern of Au and Pt recovered after alkali leaching. The measured peaks correspond to standard Au (PDF#04-0784) for the (111), (200), (220), and (311) crystal planes, confirming the elemental form of Au. Three sets of Pt peaks (PDF#87-0646) appear, corresponding to the (111), (200), and (220) crystal planes, indicating a higher Pt content compared to Au. Both Pt and Au predominantly grow on the (111) crystal plane. The calculated grain sizes reveal that the Au (111) crystal plane in the mixture is approximately 43.7 nm, while the Pt (111) crystal plane is approximately 11.8 nm.

Fig. 9 Photographs of (a) Au, (b) Pt/Au, and XRD patterns of (c) Au, (d) Pt/Au recovered after alkaline leaching, (e) XPS spectra of Au extracted by alkaline leaching, and XPS spectra of Pt/Au extracted by alkaline leaching: (f) Au 4f and (g) Pt 4f.

Fig. 9e displays the X-ray photoelectron spectroscopy (XPS) of elemental Au extracted after alkaline leaching, with peaks corresponding to Au 4f_7/2 and Au 4f_5/2, having binding energies of 84.5 eV and 88.2 eV, respectively. This confirms that the product extracted by alkaline leaching is elemental Au and indicates the absence of other oxidation states. Fig. 9f and g present the XPS spectra after alkaline leaching extraction of Pt/Au, with Fig. 9f representing Au 4f and Fig. 9g representing Pt 4f. The peaks in Fig. 9f correspond to Au 4f_7/2 and Au 4f_5/2, with binding energies of 84.5 eV and 88.2 eV, confirming the presence of elemental Au, The peaks in Fig. 9g correspond to Pt 4f_7/2 and Pt 4f_5/2, with binding energies of 71.6 eV and 74.9 eV, confirming the presence of elemental Pt.^41,49,50 Due to the extremely small size of Pt particles, a portion of the Pt surface oxidizes to form platinum oxide (PtO), which has a binding energy of 72.6 eV.

Fig. 10a and b display scanning electron microscope (SEM) images of the recovered Au elements following alkaline leaching. The Au elements appear as relatively uniform spherical particles, though they exhibit adhesion and agglomeration. Fig. 10c and d present SEM images of the recovered Pt/Au mixture following alkaline leaching. The images reveal significant agglomeration. Due to the smaller size of Pt particles, their agglomeration is more pronounced than that of Au particles. In the high-magnification images, two particle sizes can be observed: the larger particles are likely Au, while the smaller ones are Pt. Fig. 10e and f shows the energy-dispersive X-ray spectroscopy (EDS) mapping and overall spectra of the recovered Pt/Au mixture following alkaline leaching. The figure indicates the presence of Pt, Au, W, and O elements, with Pt and Au having higher contents of 83.24 wt% and 16.54 wt% respectively, and W and O having lower contents of 0.09 wt% and 0.03 wt% respectively. Because the Pt content is higher than that of Au, both elements exhibit agglomeration in certain areas, leading to uneven content distribution in the mapped region. Consequently, the Pt content is higher, and the Au content is lower, resulting in an overall Pt/Au mixture proportion of 99.88%.

Fig. 10 (a and c) SEM images of Au recovered after alkaline leaching, (b and d) SEM images of Pt/Au, (e) EDS surface elements distribution mapping and overall spectrum of Pt/Au recovered after alkaline leaching and (f) energy spectrum, and (g) TEM image of Au recovered after alkaline leaching, (h) SAED image, (i) TEM image of Pt/Au and (j) HRTEM image: ① and ② are FFT images of different regions.

The TEM images of Au and Pt/Au, recovered after alkali leaching, are presented in Fig. 10g and i, while the SAED images are shown in Fig. 10h and j. Au exists as nanoparticles with a grain size of approximately 40 nm, exhibiting uniform size and distribution. When Pt and Au coexist, they appear aggregated, making them difficult to distinguish clearly. Fig. 10h and j are the SAED images of Au and Pt/Au, respectively. Fig. 10g corresponds to different crystal planes of Au, specifically (111), (200), (220), (311), (222), and (400), which are consistent with the XRD results of Au. No W or other impurities were detected, indicating that the sample is relatively pure. The FFT image in Fig. 10j corresponds to different crystal planes of Pt, while the crystal planes of WO₃ are (110) and (200). The crystal zone axis of Pt is [1̄1̄0], with the corresponding crystal planes being (1̄11̄) and (11̄1̄). No Au was observed in the field of view, possibly due to its uneven distribution and agglomerated growth. At this stage, WO₃ still exists in the recovered Pt/Au mixture, likely due to the incomplete dissolution of WO₃ during the alkali leaching process, leaving a small amount of WO₃ residue on the sample surface.

The efficiency of alkaline leaching is determined by dividing the mass of the element after leaching by its mass before leaching and then multiplying by 100%. The mass of the element before leaching is obtained by subtracting 0.25 g of WO₃ from the initial mass prior to alkaline leaching. When calculating for the Pt/Au mixture, the remaining WO₃ after alkaline leaching must also be deducted. As shown in Fig. 10f, the purity of the Pt/Au mixture is 99.88%. To calculate the alkaline leaching efficiency, the extraction efficiency of the Pt/Au mixture must be adjusted by its purity of 99.88%. Table 3 shows that the alkaline leaching efficiencies for Au and Pt/Au are 95.49% and 93.73%, respectively, indicating high leaching efficiency. Table 4 indicates that the overall extraction efficiency of the products is the product of photodeposition efficiency and alkaline leaching efficiency. Following photocatalytic reduction and alkaline leaching, the overall extraction efficiencies for Au and Pt/Au are 93.10% and 93.26%, respectively, with total efficiencies exceeding 93%, demonstrating high extraction efficiency.

Table 3 Alkaline leaching efficiency of different products

Extracted products	Before alkaline leaching (g)	After alkaline leaching (g)	Alkaline leaching efficiency
Au	1.1348	0.8449	95.49%
Pt/Au	3.3663	2.9244	93.73%

---

Table 4 Overall extraction efficiency of different products

Extracted products	Photodeposition efficiency	Alkaline leaching efficiency	Overall efficiency
Au	97.5%	95.49%	93.10%
Pt/Au	99.5%	93.73%	93.26%

---

3.5 The regeneration of WO₃ and its photodeposition performance

To recycle WO₃ and enhance its performance, two preparation methods were employed. The addition of ammonia water during the calcination process (RA-WO₃) generates more pores, resulting in a larger specific surface area and enhanced photocatalytic activity. To investigate the photocatalytic performance of RA-WO₃ prepared with ammonia water, a control experiment was conducted using WO₃ prepared without ammonia water (R-WO₃). Fig. 11a presents the XRD patterns of WO₃ produced by calcination with ammonia water and direct calcination of H₂WO₄, along with their Au photo-deposition, corresponding to the standard WO₃ peaks (PDF#20-1324) and Au peaks (PDF#04-0784), respectively. Regardless of the addition of ammonia water, the prepared WO₃ matched the standard card (PDF#20-1324). However, the peak intensities differed: RA-WO₃ exhibited a higher peak at the (200) crystal plane, while R-WO₃ showed higher peaks at the (001) and (020) crystal planes. This suggests that the performance of WO₃ may vary due to differences in crystal planes resulting from different preparation methods.

Fig. 11 (a) XRD patterns of RA-WO₃ and R-WO₃ after photodeposition of Au, (b) EDS surface elements distribution images and overall spectrum of RA-WO₃ after photodeposition of Au, (c) EDS surface elements distribution images of R-WO₃ after photodeposition of Au, (d) TEM and (e) HRTEM images of RA-WO₃, (f) TEM and (g) HRTEM images of R-WO₃, and (h) photodeposition kinetics of WO₃ regenerated by different methods.

Fig. 11b and c display the EDS surface scan and overall spectrum of WO₃ after photodeposition of Au, produced by ammonia water calcination and direct calcination of H₂WO₄, respectively. The Au appears in an aggregated form. In Fig. 11b, the W content is 24.2 wt%, the O content is 71.3 wt%, and the Au content is 4.5 wt%. In Fig. 11c, the W content is 20.2 wt%, the O content is 75.1 wt%, and the Au content is 4.7 wt%. After the photodeposition of RA-WO₃, the surface Au content is slightly higher. As observed in Fig. 11b and c RA-WO₃ is larger and more irregular, appearing in flaky, granular, and bulk forms with strong aggregation. In contrast, R-WO₃ is smaller and more regular, primarily consisting of granular and flaky forms. There is no significant difference in the photodeposition of Au on WO₃ prepared by these two methods.

Fig. 11d and e display the TEM and HRTEM images of WO₃ produced by calcination with ammonia water, whereas Fig. 11f and g show the TEM and HRTEM images of WO₃ produced by direct calcination of H₂WO₄. As illustrated in Fig. 11d, the morphology of WO₃ is irregular, predominantly comprising spherical, rod-like, and block-like shapes. Fig. 11e reveals a measured lattice fringe spacing of 0.335 nm, corresponding to the (120) plane spacing of WO₃ (PDF#20-1324). As shown in Fig. 11f, WO₃ primarily consists of blocky and triangular shapes, with grain sizes around 100 nm. Fig. 11g shows a measured lattice fringe spacing of 0.241 nm, corresponding to the (103) plane spacing of WO₃ (PDF#20-1324).

Fig. 11h illustrates the photodeposition kinetics curves of WO₃ prepared by different methods. All three types of WO₃ were fully reduced after 20 min of reaction. After 10 min of photodeposition, the reaction efficiency of R-WO₃ nanoparticles and RA-WO₃ were relatively high, with slightly higher reduction efficiency, whereas the reaction efficiency of R-WO₃ was slower, exhibiting a slightly lower reduction efficiency. After 20 min of reaction, the reaction efficiency of WO₃ prepared by the three methods converged to essentially the same value. After 60 min of reaction, the concentration of Au ions in the solution decreased to 0.0025 mM, representing a reduction efficiency of 97.5%, indicating that almost all Au ions were reduced. RA-WO₃ did not exhibit any significant advantage over R-WO₃.

3.6 Mechanism of photodeposition of Au and Pt/Au on nano WO₃

When light irradiates the semiconductor WO₃, electrons in the valence band are excited by external energy and transition to the conduction band. These negatively charged electrons can reduce Au ions to elemental Au, which deposits on the surface of WO₃, forming new active sites of Au/WO₃.⁵¹ Ethanol in the solution acts as a hole scavenger, reacting with holes to produce CO₂ and H₂O. Additionally, H₂O in the solution reacts with holes to produce O₂, thereby preventing the recombination of electrons and holes. This allows more electrons to participate in the reduction reaction, increasing electron utilization efficiency (Fig. 12a).^52–54

Fig. 12 Mechanism diagram of (a) Au photodeposition on WO₃ nanoparticles, and (b) Pt/Au photodeposition on WO₃ nanoparticles.

The mechanism of photo-deposition of Au on WO₃ nanoparticles can be primarily divided into three parts. First, upon light excitation, WO₃ undergoes an electron transition, resulting in the formation of holes in the valence band, as illustrated in eqn (3).

During a reduction reaction in the solution, Au ions react with electrons to form elemental Au, as illustrated in eqn (4). WO₃ reacts elemental Au to form a catalyst, Au/WO₃. Subsequently, Au ions interact with Au/WO₃ to regenerate elemental Au, as depicted in eqn (5).

Subsequently, an oxidation reaction occurs in which both ethanol and water capture holes, producing by-products such as CO₂ and O₂, as illustrated in eqn (6) and (7).

The overall reaction is represented in eqn (8). Under light irradiation and with WO₃ as the catalyst, Au ions as the raw material and ethanol as the hole scavenger ultimately produce elemental Au, CO₂, and O₂.

The model and formula for photodeposition of Au on WO₃ nanoparticles elucidate the mechanisms of electron transition and Au ion reduction during the photodeposition process, clearly demonstrating that the photodeposition method can extract noble metal ions from the solution, producing only CO₂ and O₂ as by-products, and without generating impurities, thereby ensuring the purity of the extracted noble metal Au is maintained.

Fig. 12b illustrates the mechanism of photo-deposition of Pt and Au on WO₃ nanoparticles. When light irradiates the semiconductor WO₃, electrons in the valence band are excited and transition to the conduction band. The excited electrons reduce Pt and Au ions to elemental Pt and Au, respectively, which deposit on the surface of WO₃, forming new active sites Pt/WO₃ and Au/WO₃. At the Pt/WO₃ active sites, electrons from WO₃ combine with holes on Pt, enabling more electrons on Pt to participate in the reaction, thereby reducing Au ions to Au and Pt ions to Pt, promoting Pt cluster growth. Compared to elemental Pt, the presence of Au hinders Pt aggregation, weakening its tendency to aggregate and resulting in Pt–Au or Pt–Pt growth modes.^55,56 At the Au/WO₃ active sites, electrons from WO₃ combine with holes on Au, enabling more electrons on Au to participate in the reaction, thereby reducing Pt ions to Pt and Au ions to Au, promoting Au cluster growth. Compared to elemental Au, the presence of Pt hinders Au aggregation, weakening its tendency to aggregate and resulting in Au–Pt or Au–Au growth modes. Ethanol in the solution acts as a hole scavenger, reacting with holes in WO₃ to produce CO₂ and H₂O, thereby preventing electron–hole recombination, allowing more electrons to participate in the reduction reaction and enhancing electron utilization efficiency.⁵⁷

3.7 Principle of alkaline leaching extraction of Au and Pt/Au

The principle of the alkaline leaching process is illustrated in eqn (9), where WO₃ reacts with NaOH to form Na₂WO₄, while noble metals remain unreacted and stay in the solution as solids. These noble metals can then be extracted via centrifugal separation. Fig. 13 illustrates the potential–pH diagram of the W-H₂O system.⁵⁸ It can be observed that WO₃ exists stably when the pH value is less than 4, whereas it converts to WO₄²⁻ when the pH value is greater than 4.⁵⁹

WO₃ + NaOH → Na₂WO₄ + H₂O (9)

Fig. 13 Potential–pH diagram of W-H₂O.⁵⁸

During the alkaline leaching process, the alkaline substances used are NaOH and NH₄·H₂O. Experiments indicate that noble metals dissolve when NH₄·H₂O is used for alkaline leaching, but do not dissolve when NaOH is used. Consequently, NaOH is selected as the reagent for alkaline leaching.

3.8 Economic estimation

Table 5 presents the economic benefit calculation for producing one gram of noble metal Au. To ensure smooth technical operation and control the concentration of Au ions in the experiment, chloroauric acid, priced at 295 RMB per g, was used as the source of Au. However, in actual production, the raw materials primarily come from industrial wastewater containing noble metals, leading to lower raw material costs and higher economic benefits. According to statistics, the cost of experimental techniques and materials is only 5 RMB, while the market price of nano Au is approximately 1200 RMB per g. In comparison, the price of gold is nearly 700 RMB per g (as 2024), indicating its higher economic value. Pt black and nano Au are extensively used in aviation, aerospace, missiles, rockets, nuclear energy, microelectronics, chemical engineering, gas purification, and the metallurgical industry. Therefore, the use of WO₃ as a photocatalyst for the enrichment and extraction of noble metals from industrial waste liquids is feasible. Moreover, this technology is cost-effective, requiring fewer consumables and reagents, thus making it suitable for industrial production processes.

Table 5 Consumables and costs required to extract 1 g of Au

Materials	Unit price (¥)	Usage	Price (¥)
Anhydrous ethanol	12/500 mL	60 mL	1.44
Ammonium metatungstate	217/100 g	0.37 g	0.81
Polyvinylpyrrolidone	55/250 g	0.5 g	0.11
Polyethylene glycol	38/500 g	0.5 g	0.04
Sodium hydroxide	68/500 g	0.2 g	0.03
Filter paper	22/50 sheets	1 sheet	0.50
Centrifuge tube	32/200 pieces	5 pieces	0.80
Electricity cost	0.6/degree	2 degrees	1.20
Total	—	—	4.93

---

Table 6 presents a comparison of various literature regarding the recycling of precious metals using different methods. Techniques such as solvent extraction and displacement are associated with relatively high costs and lack environmental friendliness, whereas methods like adsorption exhibit relatively low efficiency. By contrast, the photocatalytic method demonstrates both economic viability and high efficiency.

Table 6 Comparison of different methods for recycling precious metals

Methods	Materials	Precious metals	Initial concentration	Times	Efficiency	Cost	Ref.
Displacement	Si	Pt/Pd/Au	1 mM	5 min	99.9%	Medium	60
	Cu	Au	100 mg L^−1	7 min	93%	Medium	61
Extraction	Diethyl carbonate	Au	100 mg L^−1	5 min	73%	Higher	62
	4-HAP	Au	200 mg L^−1	2 min	99.7%	Medium	63
Adsorption	Activated carbon	Au	50 mg L^−1	24 h	97%	Lower	64
	Biomass	Pt/Pd/Au	100 mM	50 h	95%	low	65
Electrolysis	—	Au	0.125 mM	2 h	94.2%	Lower	66
	Zn	Au	10 mg L^−1	30 min	99.59%	Medium	67
Photocatalysis	ZnO	Pd	0.3 mM	15 min	>99%	Lower	33
	ZnO	Ag	0.15–1.2 mM	<30 min	>90%	Lower	68
	WO3	Au	0.1 mM	10 min	99%	Lower	This work

---

4. Conclusion

This study proposes a green, efficient, and low-cost semiconductor photochemical recycling technology for the enrichment and extraction of low-concentration noble metals. This method can complete the enrichment of noble metals within 20 min, with a deposition efficiency of over 99% and a total recovery efficiency of over 93%, suitable for noble metal concentrations ranging from 0.01 to 10 mM. The extracted Pt particles have a diameter of approximately 5 nm, and the Au particles have a diameter of about 60 nm, both with a purity of 99.9%. The maximum loading capacity of the WO₃ semiconductor for Au is 15.7 g g⁻¹, and for Pt is 19.47 g g⁻¹. Due to the high photocatalytic activity of the semiconductor, it can efficiently reduce and enrich extremely low concentrations of noble metal ions, making it suitable for various noble metals and their mixed solutions. The recycling and regeneration of WO₃ semiconductor nanoparticles can be achieved by direct acidification and calcination of the alkaline leaching solution, maintaining stable photodeposition performance. Economic calculations indicate that this semiconductor photochemical recycling technology is low-cost and has potential for industrial application. The efficient separation of noble and base metals will be investigated in future.

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Author contributions

All listed authors participated in this work and the detailed contribution is described as follows: Faqi Zhan designed the experiment and draft the manuscript. Ruixin Li and Haiyan Zhao conducted experiments and organized data. Yisi Liu were responsible for layout and grammar correction. Min Zhu and Yuehong Zheng assisted to analyze the data. Dalin Chen provided partial financial support for the project. The whole study was conducted under the guidance of Prof. Peiqing La, and Jie Li.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study was supported by the Gansu Province Science and Technology Major Project (24ZD13GA018), the Gansu Provincial Department of Education Young Doctor Support Project (2023QB-036), the China Postdoctoral Science Foundation (2022MD723787), the fund of the State Key Laboratory of Nickel and Cobalt Resources Comprehensive Utilization (GZSYS-KY-2021-023), and the Tamarisk Outstanding Young Talents Program of Lanzhou University of Technology (062202).

References

1. T. Q. Tran, J. K. Y. Lee and A. Chinnappan, et al. , Strong, lightweight, and highly conductive CNT/Au/Cu wires from sputtering and electroplating methods, J. Mater. Sci. Technol., 2020, 40, 99–106.
2. H.-Q. Fan, D.-D. Shi and X.-Z. Wang, et al. , Enhancing through-plane electrical conductivity by introducing Au microdots onto TiN coated metal bipolar plates of PEMFCs, Int. J. Hydrogen Energy, 2020, 45(53), 29442–29448.
3. H. Chen, L. Zou and E. Hossain, et al. , Functional structures assembled based on Au clusters with practical applications, Biomater. Sci., 2024, 12, 4283–4300.
4. M. She, Z. Wang and J. Chen, et al. , Design strategy and recent progress of fluorescent probe for noble metal ions (Ag, Au, Pd, and Pt), Coord. Chem. Rev., 2021, 432, 213712.
5. S. K. Padamata, A. S. Yasinskiy and P. V. Polyakov, et al. , Recovery of noble metals from spent catalysts: a review, Metall. Mater. Trans. B, 2020, 51(5), 2413–2435.
6. R. Herrala, Z. Wang and J. Vapaavuori, et al. , Recovery of gold as nanoparticles from gold-poor Au–Cu–Cl solutions, J. Phys. Chem. C, 2023, 127(32), 16099–16109.
7. H. Yan, F. Liu and J. Zhang, et al. , Facile synthesis and environmental applications of noble metal-based catalytic membrane reactors, Catalysts, 2022, 12(8), 816.
8. M. Xue, W. Mao and J. Chen, et al. , Application of Au or Ag nanomaterials for colorimetric detection of glucose, Analyst, 2021, 146, 6726–6740.
9. Z. Jiang, J. Duan and X. Guo, et al. , Synthesis of Au/lignin-tannin particles and their anticancer application, Green Chem., 2021, 23, 6945–6952.
10. S. Saha, H. Basu and S. Singh, et al. , A biogenic hydrogel to recover Au(III) from electronic waste, J. Environ. Manage., 2024, 363, 121384.
11. Y. Xiang, X. Chen and C. Cao, et al. , High performance and selectivity recovery of Au(III) from waste solution using the RFU resin, React. Funct. Polym., 2020, 154, 104637.
12. H. Li, M. Ye and Z. Liu, et al. , Selective Pt recovery from spent catalyst enabled by hierarchical porous poly(imine dioxime)/polyethylenimine composite membrane for recycled Pt/C catalyst, Sep. Purif. Technol., 2023, 310, 123125.
13. I. Korolev, S. Spathariotis and K. Yliniemi, et al. , Mechanism of selective gold extraction from multi-metal chloride solutions by electrodeposition-redox replacement, Green Chem., 2020, 22(11), 3615–3625.
14. K. Yliniemi, P. Halli and I. Korolev, et al. , Effect of impurities in precious metal recovery by electrodeposition-redox replacement method from industrial side-streams and process streams, ECS Meet. Abstr., 2018, MA2018-01(20), 1292.
15. K. He, J. Tang and H. Weng, et al. , Efficient extraction of precious metal ions by a membrane emulsification circulation extractor, Sep. Purif. Technol., 2019, 213, 93–100.
16. S. Galyaltdinov, V. Brusko and A. Khannanov, et al. , Oxidatively modified carbon as a promising material for gold extraction, Diamond Relat. Mater., 2024, 142, 110826.
17. J. He and A. Kappler, Recovery of precious metals from waste streams, Microb. Biotechnol., 2017, 10(5), 1194–1198.
18. Y. Xiang, Y. Liu and M. Li, et al. , The recovery of Au(III) by hydrogel-like beads, Hydrometallurgy, 2023, 215, 105964.
19. X. Wan, L. Kleemola and L. Klemettinen, et al. , On the kinetic behavior of recycling precious metals (Au, Ag, Pt, and Pd) through copper smelting process, J. Sustainable Metall., 2021, 7, 920–931.
20. J. Ding, Y. Chai and Q. Liu, et al. , Selective deposition of silver nanoparticles onto WO₃ nanorods with different facets: the correlation of facet-induced electron transport preference and photocatalytic activity, J. Phys. Chem. C, 2016, 120(8), 4345–4353.
21. L. E. Gomes, A. C. Nogueira and M. F. Da Silva, et al. , Enhanced photocatalytic activity of BiVO₄/Pt/PtO_x photocatalyst: the role of Pt oxidation state, Appl. Surf. Sci., 2021, 567, 150773.
22. Y. Chen, S. Lai and W. Wu, et al. , SiC Substrate/Pt nanoparticle/graphene nanosheet composite photocatalysts for hydrogen generation, ACS Appl. Nano Mater., 2024, 7(8), 8958–8968.
23. S. Chen, A. M. Abdel-Mageed and C. Mochizuki, et al. , Controlling the O-vacancy formation and performance of Au/ZnO catalysts in CO₂ reduction to methanol by the ZnO particle size, ACS Catal., 2021, 11(15), 9022–9033.
24. G. Xie, R. Jin and P. Ren, et al. , Boosting CO₂ hydrogenation to methanol by adding trace amount of Au into Cu/ZnO catalysts, Appl. Catal., B, 2023, 324, 122233.
25. K. Chang, X. Zhang and Y. Hua, et al. , Facet-dependent photocatalytic conversion of methane to C1 oxygenates with Au/TiO₂ nanoparticles, ACS Appl. Nano Mater., 2024, 7(18), 21453–21462.
26. M. Bijl, K. R. G. Lim and S. Garg, et al. , Controlling nanoparticle placement in Au/TiO₂ inverse opal photocatalysts, Nanoscale, 2024, 16, 13867–13873.
27. Y. Yao, H. Cheng and G. Zhong, et al. , Progress in the construction strategy of noble metal active sites for zeolite-based PNA and VOCs catalysts, Green Energy Environ., 2024 DOI:10.1016/j.gee.2024.06.001.
28. N. Kunthakudee, T. Puangpetch and P. Ramakul, et al. , Light-assisted synthesis of Au/TiO₂ nanoparticles for H₂ production by photocatalytic water splitting, Int. J. Hydrogen Energy, 2022, 47(56), 23570–23582.
29. J. F. S. Fernando, M. P. Shortell and K. L. Firestein, et al. , Photocatalysis with Pt-Au-ZnO and Au-ZnO hybrids: effect of charge accumulation and discharge properties of metal nanoparticles, Langmuir, 2018, 34(25), 7334–7345.
30. X. Zheng, T. Yu and S. Liu, et al. , High-efficiency photocatalyst based on Au nanoparticles loaded on defective ZnO nanorods, Vacuum, 2024, 225, 113236.
31. F. Conte, I. Rossetti and G. Ramis, et al. , Low metal loading (Au, Ag, Pt, Pd) photo-catalysts supported on TiO₂ for renewable processes, Materials, 2022, 15(8), 2915.
32. N. Posachayanan, P. Liwetpitayat and A. Thoumrungroj, et al. , Highly-efficient recovery of silver from industrial cyanide-based plating effluent on TiO₂/activated carbon composites, Chemosphere, 2024, 143614.
33. M. Muscetta, R. Andreozzi and R. Marotta, et al. , Recovery of palladium(II) from aqueous solution through photocatalytic deposition in presence of ZnO under UV/visible-light radiation, J. Environ. Chem. Eng., 2021, 9(6), 106523.
34. S.-I. Motoda, S. Uematsu and T. Shinohara, Influence of impurities in TiO₂ coatings on electrode potential of photocatalytic anode assembling to marine microbial fuel cell, ECS Meet. Abstr., 2011, MA2011-02(4), 239.
35. N. C. D. Nath, S. Y. Choi and H. Park, et al. , (Invited) stand-alone photoreduction of carbon dioxide on copper oxide nanowires powered by WO₃/dye dual absorber, ECS Meet. Abstr., 2016, MA2016-02(49), 3691.
36. G. Mineo, L. Bruno and E. Bruno, et al. , WO₃ nanorods decorated with very small amount of pt for effective hydrogen evolution reaction [J/OL], Nanomaterials, 2023, 13(6), 1071.
37. I. Székely, M. Baia and K. Magyari, et al. , The effect of the pH adjustment upon the WO₃–WO₃·0.33H₂O-TiO₂ ternary composite systems' photocatalytic activity, Appl. Surf. Sci., 2019, 490, 469–480.
38. A. N. Tuama, L. H. Alzubaidi and M. H. Jameel, et al. , Impact of electron–hole recombination mechanism on the photocatalytic performance of ZnO in water treatment: a review, J. Sol-Gel Sci. Technol., 2024, 110, 792–806.
39. C. M. Krowne, Nernst equations and concentration chemical reaction overpotentials for VRFB operation, J. Electrochem. Soc., 2023, 170(10), 100534.
40. N. P. Finkelstein and R. D. Hancock, A new approach to the chemistry of gold, Gold Bull., 1974, 7(3), 72–77.
41. A. Y. Fedorov, A. V. Bukhtiyarov and M. A. Panafidin, et al. , Thermally Induced surface structure and morphology evolution in bimetallic Pt-Au/HOPG nanoparticles as probed using XPS and STM, Nanomaterials, 2024, 14(1), 57.
42. A. Mellor, A. Wilson and C. L. Pang, et al. , Photoemission core level binding energies from multiple sized nanoparticles on the same support: TiO₂(110)/Au, J. Chem. Phys., 2020, 15(2), 024709.
43. Abdallah M. S., Aboona B. E. and Adam J., et al. , Measurement of ⁴_∧H and ⁴_∧He binding energy in Au+Au collisions at √sNN = 3 GeV, Phys. Lett. B, 2022, 834, 137449.
44. T. Govindaraj, C. Mahendran and R. Marnadu, et al. , The remarkably enhanced visible-light-photocatalytic activity of hydrothermally synthesized WO₃ nanorods: an effect of Gd doping, Ceram. Int., 2021, 47(3), 4267–4278.
45. H. Tandon, P. Ranjan and T. Chakraborty, et al. , Computation of absolute radii of 103 elements of the periodic table in terms of nucleophilicity index, J. Math. Chem., 2020, 58(5), 1025–1040.
46. S. Wang, B. Cheng and X. Fang, et al. , Electronegativity-dependent Pt anchoring and molecule adsorption for graphene-based supported Pt single atom, J. Mol. Model., 2024, 30(5), 138.
47. V. Kumari, T. Singh and S. Devi, et al. , Atomic electronegativity based on hardness and floating spherical gaussian orbital approach, J. Math. Chem., 2022, 60(2), 360–372.
48. S. Karaoglu and S. Yolcular, Optimization of hydrogen generation process from the hydrolysis of activated Al-NaCl-SiC composites using Taguchi method, Int. J. Hydrogen Energy, 2022, 47(66), 28289–28302.
49. D. Wang, X. Cui and Q. Xiao, et al. , Electronic behaviour of Au-Pt alloys and the 4f binding energy shift anomaly in Au bimetallics- X-ray spectroscopy studies, AIP Adv., 2018, 8(6), 065210.
50. N. Omar, P. Cloutier and C. Ramseyer, et al. , Binding energies of CD₄ and fragment species to Pt(111): implications for measurements of anion electron stimulated desorption, J. Chem. Phys., 2024, 161(6), 064705.
51. Y. Zhao, K. Jiang and W. Wang, et al. , Simple synthesis of Au-WO₃ nanoparticles with enhanced photocatalytic performance, J. Electron. Mater., 2024, 53(7), 4250–4260.
52. D. Li, R. Chen and S. Wang, et al. , Unraveling the kinetics of photocatalytic water oxidation on WO₃, J. Phys. Chem. Lett., 2020, 11(2), 412–418.
53. F. Zhan, G. Wen and R. Li, et al. , A comprehensive review of oxygen vacancy modified photocatalysts: synthesis, characterization, and applications, Phys. Chem. Chem. Phys., 2024, 26(15), 11182–11207.
54. R. Li, F. Zhan and G. Wen, et al. , Facile synthesis of a micro-nano-structured FeOOH/BiVO₄/WO₃ photoanode with enhanced photoelectrochemical performance, Catalysts, 2024, 14(11), 828.
55. J. Ran, H. Liu and H. Dong, et al. , Accurate quantification of TiO₂(B)'s phase purity via Raman spectroscopy, Green Energy Environ., 2023, 8(5), 1371–1379.
56. J. Chen, R. Yin and G. Chen, et al. , Selective capture of Tl₂O from flue gas with formation of p–n junction on V₂O₅-WO₃/TiO₂ catalyst under working conditions, Green Energy Environ., 2023, 8(1), 4–9.
57. A. L. Kozlovskiy and M. V. Zdorovets, Study of the photocatalytic activity of irradiated WO₃ microparticles, Appl. Phys. A, 2020, 126(8), 638.
58. M. J. N. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions [B], 1974.
59. N. Cui, W. Li and Z. Guo, et al. , Electrocatalytic performance of carbon supported WO₃-containing Pd-W nanoalloys for oxygen reduction reaction in alkaline media, Catalysts, 2018, 8(6), 225.
60. K. Fukuda, S. Yae and N. Fukumuro, et al. , Noble metal recovering by electroless displacement deposition on silicon powder, ECS Meet. Abstr., 2013, MA2013-01(24), 977.
61. I. Korolev, P. Altınkaya and P. Halli, et al. , Electrochemical recovery of minor concentrations of gold from cyanide-free cupric chloride leaching solutions, J. Clean. Product., 2018, 186, 840–850.
62. S. Raiguel, L. Gijsemans and A. Van Den Bossche, et al. , Solvent extraction of gold(III) with diethyl carbonate, ACS Sustainable Chem. Eng., 2020, 8(36), 13713–13723.
63. B. T. Khogare, G. S. Kamble and A. N. Kokare, et al. , Development of novel solvent extraction method for determination of gold(III) using 4-heptylaminopyridine: Application to alloys and environmental analysis, J. Environ. Chem. Eng., 2016, 4(3), 3075–3083.
64. Y. Jiang, Y. Chen and F. Zi, et al. , Making untreated carbon effective in cleaner thiosulfate system: a new and high-efficiency method including gold adsorption and desorption, J. Clean. Product., 2022, 334, 130185.
65. Q. Zhu, J. Zhu and K. Huang, Selectivity of modified pomegranate peel for gold, platinum, and palladium: adsorption behavior, mechanism, and recovery, J. Environ. Chem. Eng., 2024, 12(2), 111902.
66. H. Wang, Y. Wu and N. Deng, et al. , Development of a bipolar electrochemical flow microreactor for recovery of valuable metals from mixed solutions, Chem. Eng. J., 2020, 382, 121907.
67. Q. Zhang, Y. Nie and Q. Wang, et al. , Recovery of gold ions from thiosulfate solution using an electrogenerative process, Process Saf. Environ. Prot., 2024, 192, 1515–1524.
68. M. Muscetta, L. Clarizia and M. Race, et al. , A novel green approach for silver recovery from chloride leaching solutions through photodeposition on zinc oxide, J. Environ. Manage., 2023, 330, 117075.

---