Pt Single-Atom Doping in Ag₂₉ Nanoclusters for Enhanced Band Bending and Z-Scheme Charge Separation in TiO₂ Heterojunction Photocatalysts

Abstract

In recent years, metal nanoclusters (NCs) with atomic-scale precision have emerged as novel photosensitizers for light energy conversion in metal cluster-sensitized semiconductor (MCSS) systems. However, conventional NCs often suffer from photodegradation after binding with semiconductors, limiting their long-term catalytic stability. Modifying NCs via single-atom doping provides an effective strategy to tailor their interfacial charge transfer behavior. In this study, PtAg₂₈ NCs were synthesized by doping Pt single atoms into Ag₂₉ NCs and subsequently loaded onto TiO₂ via electrostatic adsorption to construct composite photocatalysts. Systematic investigations revealed that Pt doping significantly enhances light absorption and promotes the formation of a direct Z-scheme heterojunction. The optimized PtAg₂₈/TiO₂ composite exhibits effective suppression of charge recombination. This enhanced charge separation efficiency, driven by pronounced band bending at the interface, leads to a remarkable hydrogen evolution rate of 14,564 μmol g⁻¹ h⁻¹. This work demonstrates the critical role of single-atom doping in regulating the photophysical properties of metal NCs and offers a feasible approach for designing highly efficient and stable metal-cluster-based photocatalytic systems.

1. Introduction

The conversion of solar energy into chemical energy represents a sustainable and promising strategy to address global energy challenges and environmental concerns [1,2]. Among various approaches, solar-driven photocatalytic water splitting has attracted significant interest due to its potential to produce clean hydrogen fuel using abundant solar energy [3,4]. In recent years, atomically precise metal nanoclusters (NCs) have emerged as promising materials for solar energy conversion owing to their unique molecular-like electronic structures, quantum confinement effects, and abundant active sites [5,6]. These properties, coupled with their tunable optical characteristics, make metal NCs ideal candidates for photocatalytic applications, particularly in hydrogen evolution reactions [7,8,9]. According to metal synergistic effects, doping foreign atom(s) as an effective strategy opens up immense opportunities for engineering the chirality, catalytic activity, luminescence, stability, and other properties of metal nanoclusters at the atom level [10,11,12]. Bimetallic metal nanoclusters are a class of versatile catalysts due to the synergistic effects of different metal components, allowing for enhanced catalytic activity, selectivity, and stability compared to their monometallic counterparts [13,14,15]. The synergistic effect in bimetallic nanoclusters is generally attributed to modifications in the electronic structure or the formation of collaborative sites due to the presence of multiple metal components [16,17,18]. Therefore, understanding the distribution of individual metal components is crucial for revealing the nature of these synergetic effects [19,20].

Moreover, unlike conventional photosensitizers, the catalytic capability of the metallic core of NCs allows them to act as active catalytic centers in reactions. The significant energy gap between the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) of metallic nanoclusters enables these nanoclusters to function as narrow-bandgap semiconductors [21]. Metal NCs function as emerging photosensitizers due to their distinct, molecule-like HOMO-LUMO energy levels, which enable efficient electron transfer to semiconductor substrates. For instance, integration of metal NCs with metal oxides to form heterostructures presents a significant potential for photocatalytic applications [22,23]. This combination extends the photoresponse range of wide-bandgap semiconductors and enhances the carrier intensity in composite photosystems, thus enabling the various forms of photoredox catalysis including photocatalytic pollution control, CO₂ reduction, water splitting, and selective organic transformation under visible light. Despite recent progress, several key obstacles remain, such as the scarcity of metal NC–semiconductor pairs with suitably aligned energy levels, extremely short-lived charge carriers, and difficulties in precisely regulating interfacial charge transfer pathways [24,25,26,27,28,29,30,31]. These factors collectively hinder the further development of metal NC-based photocatalysts. Therefore, synthesizing novel load-bearing materials with metallic NC-semiconductors is of significant importance.

Titanium dioxide (TiO₂) is one of the most widely investigated semiconductor materials for solar energy conversion and photocatalysis owing to its excellent chemical stability, low cost, non-toxicity, and suitable band alignment. In addition to TiO₂, various binary oxide ceramics such as ZnO, Al₂O₃, SiO₂, CeO₂, Fe₂O₃, and WO₃ have also been explored for applications in solar cells and photocatalytic systems; however, TiO₂ remains particularly attractive due to its robustness and versatility under diverse reaction conditions. Nevertheless, the wide bandgap and rapid charge recombination of pristine TiO₂ significantly limit its solar utilization efficiency, motivating extensive efforts to enhance its visible-light response and charge separation performance [32,33].

In this report, Pt single atoms were successfully introduced into the well-known Ag₂₉ NCs. Then, they were deposited onto TiO₂ to study the correlation of core structure, interface interaction behaviors, and photocatalytic properties. PtAg₂₈ NCs were thus obtained, which represent a new type of bimetallic nanocluster with a precisely controlled atomic composition. The introduction of Pt single atoms into Ag₂₉ NCs not only modifies the electronic structure of the nanoclusters but also creates new active sites for photocatalytic reactions. The deposition of these PtAg₂₈ NCs onto TiO₂ further enhances their photocatalytic performance by facilitating efficient charge separation and transfer at the interface. Through a series of characterization techniques, such as X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), and UV-Vis absorption spectroscopy, we confirmed the successful doping of Pt atoms and the formation of a strong interaction between PtAg₂₈ NCs and TiO₂. Moreover, the photocatalytic activity of the composite photocatalysts was evaluated by measuring hydrogen production rates under simulated solar irradiation. The results demonstrate that the PtAg₂₈/TiO₂ composite exhibits significantly improved photocatalytic performance compared to pure TiO₂ and Ag₂₉/TiO₂, highlighting the importance of interfacial interactions and atomic-level engineering in designing efficient photocatalysts.

2. Results

2.1. The Synthesis and Efficient Loading of Metal Nanoclusters

Thiolate-protected Ag₂₉(S-Adm)₁₈ and PtAg₂₈(S-Adm)₁₈ NCs were synthesized using the previously reported method [34,35]. The synthetic process involved precise control over reaction conditions to ensure the formation of nanoclusters with desired atomic compositions and structures. To confirm the atomic precision of Ag₂₉ NCs and the successful incorporation of a single Pt atom into the cluster framework, their optical absorption properties were examined by ultraviolet–visible (UV–vis) spectroscopy and their purity was assessed via electrospray ionization mass spectrometry (ESI-MS). The as-synthesized Ag₂₉ nanocluster exhibits a strong absorption signal at 437 nm in the UV–vis spectrum consistent with previous reports, making it an ideal platform for subsequent modifications (Figure 1a). By introducing a controlled amount of Pt precursor during the synthesis, PtAg₂₈ NCs were successfully obtained. Compared with Ag₂₉, PtAg₂₈ NCs display characteristic absorption peaks at 440 nm and 540 nm, which agree well with previously reported PtAg₂₈ NCs, confirming the successful preparation of bimetallic nanoclusters and their unique optical absorption properties (Figure 1b). And as illustrated in the diagram, we studied the fluorescence of an alloy nanocluster. The fluorescence emission spectrum of the Ag₂₉ and PtAg₂₈ nanoclusters exhibits high fluorescence intensity at 841 nm and 700 nm. Compared to Ag₂₉ NCs, the excitation spectrum of PtAg₂₈ nanoclusters exhibits two main peaks at 336 nm and 445 nm. Meanwhile, the emission peak has a 141 nm blueshift. This blueshift suggests that Pt doping alters the electronic transition pathways and effectively tunes the emission energy levels of the nanoclusters. The chemical compositions of Ag₂₉ and PtAg₂₈ were confirmed by electrospray ionization mass spectrometry (ESI-MS) in positive ion mode. In the ESI-MS spectrum of Ag₂₉, a strong signal was observed at m/z 2396.8352 (Figure S1), corresponding to the [Ag₂₉(S-Adm)₁₈(PPh₃)₄]³⁺ species. Similarly, the ESI-MS spectrum of PtAg₂₈ exhibited a prominent peak at m/z 3637.6377, which was assigned to [PtAg₂₈(S-Adm)₁₈(PPh3)₄]²⁺ (Figure S2). The experimental and simulated isotopic distributions for both species exhibited an exact match, and the peak spacing corresponds precisely to the +3 and +2 charge states, further confirming the successful atom-level substitution of Ag by Pt and the high purity of the nanoclusters.

Efficient loading of these nanoclusters onto the TiO₂ surfaces was achieved via an electrostatic adsorption method. The interaction between the nanoclusters and TiO₂ was optimized by adjusting the deposition parameters, to ensure uniform distribution and strong adhesion [21]. To verify the adhesion of the clusters to TiO₂, we performed Fourier transform infrared (FTIR) spectroscopy (Figure 1c). Compared to TiO₂, the loaded photocatalytic materials exhibited new characteristic peaks centered at 2851 cm⁻¹, which were attributed to C–H stretching vibrations corresponding to the C–H bonds of the S–Adm ligands in Ag₂₉ and PtAg₂₈ NCs. An enhanced peak centered at 2106 cm⁻¹ was attributed to Ag-S stretching vibrations, which correspond to the C-S bonds of S-Adm in NCs as well as the Ag-S bonds formed with Ag. In the UV–vis DRS spectra (Figure 1d), Ag₂₉/TiO₂ and PtAg₂₈/TiO₂ exhibit enhanced visible-light absorption and retain absorption bands consistent with the free nanoclusters, confirming that the nanocluster structure is preserved on the TiO₂ surface and contributes to noticeable sensitization.

To further demonstrate the successful loading of metal nanoclusters, we analyzed the morphology of the photocatalyst using transmission electron microscopy (TEM). It can be observed that TiO₂ exhibits a disc-like morphology within the 20–50 nm range (Figure 2a,c). By directly comparing the TEM images, it is evident that Ag₂₉ and PtAg₂₈ NCs are incorporated onto and supported on the external surface of the TiO₂. In addition, no apparent aggregation and size change for metal NCs in Ag₂₉/TiO₂ and PtAg₂₈/TiO₂ can be observed (Figure 2b,d). The metal nanoclusters are nearly indistinguishable because of their tiny size and low loading. Analysis of the particle size distribution indicated that the diameters of Ag₂₉ NCs and PtAg₂₈ NCs were 1.61 nm and 1.57 nm, respectively. This finding is consistent with the lack of characteristic metal nanocluster peaks in the XRD pattern of the composite (Figure S3). Elemental mapping (Figure 2e–l) demonstrates homogeneous distribution of Ti, O, Ag, and S across the composite, while Pt is detected in trace amounts, supporting the retention of the single Pt atom within the nanocluster framework. Collectively, these results confirm that the nanoclusters are efficiently and stably immobilized on the TiO₂ surface while preserving their atomically precise structures.

2.2. Charge Transfer in Z-Scheme Heterojunctions

To further elucidate the chemical state of Ag₂₉/TiO₂ and the effect of Pt single-atom introduction on it, X-ray photoelectron spectroscopy (XPS) analysis was conducted. The full spectrum of the photocatalysts verifies the presence of Ag, O, and Ti in the samples (Figure S4a), with no detectable impurity peaks. The absence of Pt signal in the full spectrum can be attributed to its low content. Inductively coupled plasma (ICP) analysis of the optimal PtAg₂₈/TiO₂ photocatalyst reveals that the contents of Ag and Pt are 0.92 wt.% and 0.06 wt.%, respectively (Figure S4b). XPS fine spectra were used to investigate the charge transfer dynamics between Ti, Ag, and Pt. First, we investigated the effect of Pt single atoms on Ag in the metal NCs. Compared to the parent Ag₂₉, PtAg₂₈ NCs exhibit a negative shift in the Ag 3d XPS binding energy (Figure 3a). This is primarily attributed to Pt single atoms acting as doping sites, enabling more efficient electron capture from ligands or interfaces. Through Pt-Ag interactions, a portion of the electron density is delocalized onto surrounding Ag atoms [36,37,38]. A negative binding energy shift is observed in the Ag 3d signal for the Ag₂₉/TiO₂ composite relative to the Ag₂₉ NCs, indicative of electron donation from TiO₂ to the nanoclusters (Figure 3b). More importantly, this phenomenon is also confirmed in the PtAg₂₈/TiO₂ system, where a similar shift is detected compared to the pristine PtAg₂₈ NCs (Figure 3c). When TiO₂ comes into contact with NCs, the Ti 2p XPS signal of Ag₂₉/TiO₂ and PtAg₂₈/TiO₂ exhibits a shift to a higher binding energy than that of TiO₂ (Figure 3d). These parallel results from the Ag₂₉/TiO₂ and PtAg₂₈/TiO₂ systems collectively demonstrate that electron transfer from TiO₂ to the nanoclusters is a general interfacial interaction. This charge transfer pathway signifies the formation of downward band bending at the heterojunction interface in both composites, which provides the driving force for carrier separation [39]. The shift of these peaks is well proven in the Pt-Ag system; the perfect contact between Ag NCs and TiO₂, and the probable formation of Z-scheme heterojunction, may be significant for the movement of photogenerated carriers [40,41,42].

Furthermore, to further confirm the formation of the Z-scheme Ag/TiO₂ composite heterojunction, in situ electron spin resonance (EPR) techniques were employed to monitor the capture of superoxide radicals (·O₂⁻) by 5,5′-dimethyl-1-pyrroline -N-oxide (DMPO) capturing ·O₂⁻ and in situ photoluminescence (PL) of terephthalic acid (PTA) capturing hydroxyl radicals (·OH). These were employed to investigate the radical generation behavior of the photocatalyst. Essentially, no DMPO-·O₂⁻ signal was detected under dark conditions, while the DMPO-·O₂⁻ signals of all samples were enhanced under irradiation (Figure 4a). The intensity of the ·O₂⁻-DMPO signal is obviously stronger for the PtAg₂₈/TiO₂ sample than for TiO₂ and Ag₂₉/TiO₂, implying that more ·O₂⁻ radicals are generated in PtAg₂₈ NCs. This indicates that the more negative LUMO position of PtAg₂₈ NCs is conducive to reduction reactions. In the PL experiment, the PtAg₂₈/TiO₂ sample exhibited the highest PL intensity, indicating that this sample generated the most ·OH radicals during the photocatalytic process (Figure 4b). The ·OH signal detected in TiO₂ is higher than that in the two Ag NCs. This indicates that the VB position of TiO₂ is more suitable for oxidation reactions to occur. Both results occurred because more electrons were transferred to the metal NCs in the composites, which were then captured by O₂ to produce ·O₂⁻, while more holes oxidized H₂O on TiO₂ to produce ·OH. Such coexistence is characteristic of a Z-scheme mechanism, where interfacial recombination eliminates low-energy charge carriers while retaining high-energy electrons and holes at their respective components. In contrast, a type-Il heterojunction would weaken either the oxidation or reduction capability due to charge migration to energetically less reactive band edges. Therefore, the electron–hole separation mode of the photocatalyst is inferred to the Z-scheme mechanism [43,44,45,46,47]. After doping with Pt atoms, there is an accumulation of electrons due to larger band bending. Meanwhile, the results of EPR and PL demonstrate that PtAg₂₈/TiO₂ exhibits enhanced separation of photo-excited charge carriers.

When NCs contact a semiconductor, the interaction between the two leads to charge redistribution at the interface due to the difference in their Fermi levels. So ultraviolet photoelectron spectroscopy (UPS) analysis was employed to determine the valence band maximum (VBM) and work function (Φ) variation in the heterojunction material. The valence band maxima for TiO₂, Ag₂₉/TiO₂, and PtAg₂₈/TiO₂ were measured at 2.95 eV, 1.11 eV, and 0.83 eV, respectively (Figure 5a). The work functions of TiO₂, Ag₂₉/TiO₂, and PtAg₂₈/TiO₂ were determined using the equation Φ = hν − Ecutoff (where Ecutoff is the secondary electron cutoff (SECO) position and hν is 21.2 eV), yielding values of 3.06, 3.26, and 3.53 eV, respectively (Figure 5b) [48]. The difference in work function (ΔΦ) between TiO₂ and PtAg₂₈/TiO₂ (0.47 eV) is greater than the difference with Ag₂₉/TiO₂ (0.20 eV), indicating that Pt single-atom doping in PtAg₂₈ NCs alters the energy level structure of Ag, reducing electron binding energy and conferring a stronger electron storage capacity than Ag₂₉ NCs. This process is consistent with our XPS and EPR test results, increasing the electron density on the Ag surface and inducing downward bending of the energy bands in Ag₂₉/TiO₂ (Φ = 0.05 eV) and PtAg₂₈/TiO₂ (Φ = 0.15 eV). The conduction band position (CB) of TiO₂ was determined via Mott–Schottky measurements conducted using a standard three-electrode system (Figure 5c). Additionally, the HOMO–LUMO gaps of TiO₂, Ag₂₉ NCs, and PtAg₂₈ NCs were calculated from Tauc plots (Figure 5d, Figures S5 and S6) to be 3.0 eV, 1.95 eV, and 1.79 eV, respectively. The absolute vacuum energy is translated to redox potential versus the normal hydrogen electrode (NHE) [49]. The energy level alignment and charge transfer pathways within the heterojunction were further elucidated, as illustrated in Figure 5e.

2.3. Optical Performance and H₂ Yield

To evaluate the performance of the photocatalyst in water-splitting reactions, triethanolamine served as a sacrificial agent, and the H₂ evolution rate was monitored across the full spectral range. Upon loading and increasing the nanoclusters amount, the activity gradually improves and reaches a maximum at a loading amount of 1 wt.% (Figure S7). The investigation into the effect of triethanolamine (TEA) concentration revealed that a value of 10 vol.% was optimal for maximizing the photocatalytic activity (Figure S8). As shown in Figure 6a, pristine TiO₂ generates H₂ of 1365 μmol g⁻¹ h⁻¹, while Ag₂₉/TiO₂ exhibits a roughly threefold enhancement in hydrogen production rate (3511 μmol g⁻¹ h⁻¹). Furthermore, the Pt-doping modification of Ag₂₉ NCs yielded a PtAg₂₈/TiO₂ photocatalyst with an enhanced H₂ evolution rate of 14,564 μmol g⁻¹ h⁻¹, which can be attributed to the high intrinsic activity of Pt cocatalysts and the electron transfer effect between Pt-Ag within the heterojunction material for HER. To verify the full-spectrum-driven photocatalytic reaction, we tested the effect of different wavelengths of light irradiation on the hydrogen evolution activity of the photocatalyst PtAg₂₈/TiO₂ (Figure 6b). First, no noticeable hydrogen evolution was observed when PtAg₂₈ NCs were selectively excited with light of wavelengths longer than 400 nm, indicating that photo-excitation of the nanoclusters alone does not provide sufficient driving force for proton reduction. In contrast, pronounced hydrogen production was achieved under 365 nm irradiation, where TiO₂ is effectively excited and abundant photogenerated holes are generated in its valence band. Notably, if a conventional Type-Il heterojunction were operative, visible-light excitation of the nanoclusters would be expected to drive electron transfer to the conduction band of TiO₂ and thus enable hydrogen evolution; however, this behavior was not observed. Instead, the clear wavelength-dependent response demonstrates that hydrogen evolution requires simultaneous participation of both components, thereby supporting a direct Z-scheme heterojunction in which photogenerated electrons in the LUMO of the nanoclusters and holes in the valence band of TiO₂ are retained as the active redox species. The reaction results are consistent with the ·OH species capture phenomenon observed in the in situ PL experiment described on the previous page, as well as the findings of greater band bending. To further compare the redox capabilities of different photocatalysts, we employed a COD (chemical oxygen demand) analyzer and chromatography-mass spectrometry (GC-MS) to determine their COD degradation rates in reaction solutions and analyze the oxidation product species, respectively. Notwithstanding a progressively accelerating TEA degradation rate over time, the PtAg₂₈/TiO₂ catalyst consistently delivered the highest COD removal and H₂ yield among all samples after two hours of reaction (Figure S9). GC-MS results reveal a suite of TEA degradation intermediates, tracking the pathway from triethylamine to diethylamine, methyldiethylamine, dimethylethanolamine, and ethylene glycol (Figure S11a,b). This succession of products demonstrates the stepwise breakdown of the TEA carbon skeleton via oxidation, yielding simpler organic acids, CO₂, and H₂O [50,51]. Subsequently, stability tests were carried out for photocatalysts and the results are shown in Figure 6c,d. Notably, the H₂ evolution rate of PtAg₂₈/TiO₂ remained stable without noticeable decrease during a 10 h continuous operation or after five successive cycles. This performance stands in sharp contrast to the Ag₂₉/TiO₂ catalyst, which suffered a 12% activity loss under the same testing conditions, thereby highlighting the exceptional stability imparted by the Pt-doped structure. The stability of PtAg₂₈/TiO₂ stems from Pt atoms occupying the core center of nanoclusters, forming more stable bond lengths with surrounding Ag atoms. This prevents decomposition or agglomeration of the photocatalyst during light cycles caused by core structural distortion.

Based on the above results, the exceptional structural stability of PtAg₂₈/TiO₂, which originates from the rigid Pt-centered kernel, effectively mitigates structural distortion during photocatalysis. This remarkable enhancement stems from the synergistic combination of two factors: the intrinsic structural stability provided by the Pt-centered kernel and the facilitated interfacial charge separation [52]. The robust structure and greater band bending of PtAg₂₈ NCs suppress charge recombination at the NCs/TiO₂ interface, enabling separated charges to be sustained for longer periods [11,53]. To further substantiate this conclusion, we evaluated the photoelectrochemical properties of the catalysts. The transient photocurrent response and electrochemical impedance spectroscopy (EIS) Nyquist plots were measured. As anticipated, the PtAg₂₈/TiO₂ electrode generated the strongest and most stable photocurrent density under simulated sunlight irradiation, demonstrating the most efficient generation and separation of charge carriers (Figure 7a). Consistently, its EIS Nyquist plot displayed the smallest arc radius, indicating the lowest charge transfer resistance at the electrode–electrolyte interface and the most rapid interfacial reaction kinetics (Figure 7b). Furthermore, steady-state photoluminescence (PL) spectroscopy provided complementary evidence. The PtAg₂₈/TiO₂ composite exhibited the weakest PL emission intensity among the three samples (Figure 7c). The quenching of the PL signal unequivocally confirms the suppression of radiative recombination, as the photogenerated carriers are more effectively separated and utilized in photocatalytic reactions rather than recombining to emit light. Time-resolved photoluminescence (TRPL) spectroscopy provides direct kinetic evidence for the enhanced charge separation (Figure S12). The average fluorescence lifetime of the Ag₂₉/TiO₂ composite (0.707 ns) is significantly shorter than that of pristine TiO₂ (1.434 ns), indicating effective interfacial charge transfer that quenches the TiO₂ emission, consistent with the formation of a charge-transfer pathway. Crucially, the PtAg₂₈/TiO₂ exhibits a markedly recovered and prolonged lifetime of 1.191 ns. This demonstrates that the Pt-doped interface establishes a more efficient charge separation channel. Within the framework of a direct Z-scheme mechanism, the optimized interface facilitates rapid recombination of useless charges, thereby effectively suppressing the radiative recombination of the useful high-energy carriers and leading to the observed longer apparent PL lifetime. This result correlates directly with its superior photocatalytic hydrogen evolution performance.

The photocatalytic mechanism for H₂ evolution over the PtAg₂₈/TiO₂ composite is illustrated in Figure 7d. The incorporation of a single Pt atom into the Ag₂₉ framework induces profound electronic and structural optimizations. Crucially, Pt doping narrows the HOMO-LUMO gap from 1.94 eV (Ag₂₉ NCs) to 1.79 eV (PtAg₂₈ NCs), thereby enhancing visible-light harvesting. More importantly, it causes a significant negative shift of the LUMO from −0.83 eV to −0.96 eV (vs. NHE), endowing the nanocluster with substantially stronger reducing power for proton reduction. This intrinsic property ensures that photogenerated electrons persist long enough to engage in surface reactions. At the interface, the PtAg₂₈ NCs induce a larger downward band bending (Φ_BB = 0.15) on the TiO₂ surface compared to their Ag₂₉ counterpart, creating a stronger driving force for charge separation. These synergistic advantages collectively promote a direct Z-scheme charge transfer pathway. The electrons in the CB band of TiO₂ efficiently recombine with the holes in the HOMO of the PtAg₂₈ NCs. This process preserves the high-energy electrons in the negatively shifted LUMO of the PtAg₂₈ NCs for H₂ evolution, while the powerful holes in TiO₂ oxidize the sacrificial agent. Moreover, the introduction of the Pt core has reinforced the structure of Ag₂₉ NCs. The stable core, enhanced recovery capability, and excellent interfacial charge separation collectively contribute to the exceptional photocatalytic hydrogen production activity of the PtAg₂₈/TiO₂ catalyst.

3. Discussion

This study demonstrates that the superior hydrogen evolution performance of PtAg₂₈/TiO₂ arises from the synergistic effects of Pt single-atom doping. Pt doping narrows the HOMO-LUMO gap and enhances the reducing power of the nanocluster. More importantly, it strengthens the interfacial band bending with TiO₂, establishing a direct Z-scheme charge transfer pathway. This mechanism effectively separates photogenerated carriers: high-energy electrons in PtAg₂₈ drive proton reduction, while holes in TiO₂ oxidize the sacrificial agent. Consequently, the composite achieves a high H₂ evolution rate of 14,564 μmol g⁻¹ h⁻¹ and excellent stability, demonstrating the critical role of atomic-level engineering in photocatalyst design.

4. Materials and Methods

4.1. Chemicals and Reagents

Titanium dioxide (TiO₂, P₂₅), silver nitrate (AgNO₃), 1-Adamantanethiol (C₁₀H₁₆S), triphenylphosphine (C₁₈H₁₅P), tetraphenylphosphonium bromide (C₂₄H₂₀BrP), dichloromethane (CH₂Cl₂), methanol (CH₃OH), ethanol (CH₃CH₂OH), triethanolamine (C₆HNO₃), anhydrous sodium sulfate (Na₂SO₄), and Nafion reagent (C₇HF₁₃O₅S·C₂F₄). The purity of all chemicals was chemical grade.

4.2. Synthesis of Photocatalysts

4.2.1. Synthesis of Ag₂₉ NCs

In a 20 mL amber glass vial, 94 mg of silver nitrate (AgNO₃) was dissolved in a mixture of 5 mL methanol and 10 mL ethanol under stirring. Then, 50 mg of 1-adamantanethiol (S-Adm) was added under vigorous stirring at 1200 rpm, resulting in a color change from colorless and transparent to milky white. After stirring vigorously for 20 min, 100 mg of triphenylphosphine (PPh₃) was introduced and stirring was continued for another 15 min, during which the solution turned from milky white back to colorless and transparent. Next, 10 mg of sodium borohydride (dissolved in 1 mL of ice-cold water) was added. The colorless transparent solution immediately turned yellow and gradually deepened to a dark brown over time. The reaction mixture was aged at ambient conditions for 6 h. The supernatant was then subjected to rotary evaporation to obtain a solid product, which was washed multiple times with n-hexane to ensure complete removal of any unreacted compounds. Finally, the purified cluster was dried under vacuum overnight to afford the target product, Ag₂₉(S-Adm)₁₈.

4.2.2. Synthesis of PtAg₂₈ NCs

In a 20 mL amber glass vial, dissolve 22 mg of AgNO₃ and 8 mg of potassium tetrachloroplatinate (K₂PtCl₄) in 2 mL of methanol under vigorous stirring at 1200 rpm. After complete dissolution, introduce 33 mg of 1-adamantanethiol (S-Adm) dissolved in 1 mL of methanol. Continue stirring vigorously at 1200 rpm for 30 min. Then, add 150 mg of triphenylphosphine (PPh₃) dissolved in 2 mL of dichloromethane (CH₂Cl₂) and stir for an additional 10 min. Subsequently, add 10 mL of CH₂Cl₂ and stir for another 10 min. Next, add 4 mg of tetraphenylphosphonium bromide (PPh₄Br) dissolved in 0.5 mL of methanol. After 5 min, introduce 9 mg of sodium borohydride (NaBH₄) dissolved in 1 mL of ice-cold water. Allow the reaction mixture to age at room temperature for 6 h. The supernatant is then rotary evaporated to obtain a solid product, which is washed multiple times with n-hexane to remove any unreacted compounds completely. The purified cluster is dried under vacuum overnight to afford the final product, PtAg₂₈(S-Adm)₁₈.

4.2.3. Synthesis of Ag NCs/TiO₂

Ag NCs/TiO₂ was prepared via electrostatic adsorption. First, 2.5 mg of the cluster powder was dissolved in 1 mL of methanol and then vigorously stirred into a TiO₂ dispersion (50 mg of TiO₂ in 25 mL of methanol). The mixture was sealed and stirred overnight. During this period, the light brown color of the solution faded, and the originally colorless titanium dioxide turned dark brown, indicating complete adsorption of the clusters onto the TiO₂ support. The solid was collected by centrifugation, washed several times with ethanol, and dried under vacuum.

4.3. Characterization

The samples’ crystal phase structures were analyzed by XRD-2000 X-ray diffraction (XRD, XRD-2000, Dandong Tongda Science & Technology Co., Ltd., Dandong, Liaoning, China). The structure of the catalyst was observed by transmission electron microscopy (TEM, Talos F200X, Thermo Fisher Scientific, Waltham, MA, USA). Analysis of valence and surface elements of samples was conducted by X-ray photoelectron spectroscopy, which was mainly used to analyze the degree of energy band bending and electron flow direction (XPS, PHI5400, Physical Electronics Inc., Chanhassen, MN, USA). UV–vis diffuse reflectance spectroscopy (DRS) was performed by a UV–vis spectrophotometer using Ba₂SO₄ particles as a reference sample, which was mainly for analyzing the characteristic peaks of the sample and verifying whether the sample was prepared successfully (UV–vis, Shimadzu UV-2600, Shimadzu Corporation, Kyoto, Japan). Electrospray ionization (ESI) time-of-flight mass spectra were obtained from an AB Sciex Triple TOF 5600+ system. Photoluminescence spectra were obtained by fluorescence spectrometry, which was mainly for the analysis of electron–hole complexes in samples (PL, Shimadzu-RF-6000, Shimadzu Corporation, Kyoto, Japan). Ultraviolet photoelectron spectroscopy was performed using the He I UV source, which was mainly used for sample energy band structure analysis (UPS, Thermo Fisher Scientific, East Grinstead, UK). Analysis of the surface structure of samples was conducted by electron scanning electron microscopy (SEM, JEOL Ltd., Tokyo, jSM-6460, Japan).

The electrochemical workstation was used to test the photoelectrochemical properties of catalysts, and was mainly used to analyze the carrier transport efficiency of the samples (CHI660, CH Instruments Inc., Shanghai, China). The FTO glass coated with catalyst was used as the working electrode. First, 1 mL of ethanol, 80 μL of Nafion reagent, and 5 mg of TiO₂ (p₂₅) were mixed and sonicated for 30 min, then dispersed on 0.25 cm² of FTO glass by spin-coating method. The sensitization process was conducted by immersing a TiO₂ film electrode in Ag NC aqueous solution for 24 h. The photoelectrochemical test took place in a three-electrode system with 0.5 M aqueous Na₂SO₄ solution as electrolyte, and the Ag/AgCl electrode and Pt plate as the reference electrode and counter electrode, respectively.

4.4. Free Radical Capture and Identification

The generation of hydroxyl radicals was indirectly confirmed by measuring the fluorescence spectrum of 2-hydroxy terephthalic acid at an emission wavelength of 435 nm using a fluorescence spectrophotometer (Shimadzu-RF-6000, Shimadzu Corporation, Kyoto, Japan). Specifically, 10 mg of catalyst was added to a solution containing 2 × 10⁻³ mol NaOH and 5 × 10⁻⁴ mol terephthalic acid. The mixture was irradiated under simulated sunlight for 20, 40, 60, 80, and 100 min, respectively. After irradiation, the solution was filtered through a 0.2 μm membrane, and the concentration of 2-hydroxy terephthalic acid was quantified by fluorescence measurement at an excitation wavelength of 315 nm.

For electron spin resonance (EPR) analysis, a 5 mg sample was dispersed in 0.5 mL methanol, mixed with 10 μL of 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as a spin-trapping agent, irradiated for 5 min, and then measured. The generation of superoxide radicals (·O²⁻) was investigated by performing full-spectrum excitation EPR measurements on TiO₂, Ag₂₉/TiO₂, and PtAg₂₈/TiO₂ under both dark and illuminated conditions.

4.5. Photocatalytic H₂ Generation Tests

A total of 10 mg of the chemical powder was suspended in a mixture of 73 mL of distilled water and 7 mL of triethanolamine. The air was evacuated by passing N₂ gas and stirring continuously for 15 min prior to the reaction. During the reaction, the hydrogen production vial was maintained at 25 °C using a circulating water bath throughout the entire irradiation period. The reaction vial was irradiated with a 100 mW-cm⁻¹ xenon lamp (CHFXM500, Beijing Porphyry Technology Co., Ltd., Beijing, China) and a 500 μL volume of gas in the reaction vial was extracted using a microsyringe, and the amount of hydrogen production was tested in gas chromatography. The photocatalytic hydrogen evolution was carried out under continuous irradiation for a total duration of 2 h. Hydrogen yield in the conversion gas chromatography measurement (GC-9890B, for Shanghai Linghua Instrument Co., Ltd., Shanghai, China) was measured by hydrogen production under the conditions of a TCD, thermal conductivity detector, column with a 5 A molecular sieve, column chamber temperature of 50 °C, inlet temperature of 100 °C, detector temperature set to 80 °C, and carrier gas flow rate of 20 mL-min⁻¹ during the collection period, using N₂ gas as the carrier gas yield measures the reduction performance and activity of the catalyst in photocatalytic engineering.

4.6. Determination of COD

Samples were collected from the photocatalytic reaction at 0, 30, 60, and 120 min. Each sample was filtered three times through a 0.2 μm membrane. Then, 2.5 mL of the filtrate was transferred to a digestion tube, mixed with digestion solution, and digested at 165 °C for 10 min. After digestion, the tube was removed and cooled in air for 2 min. Next, 2.5 mL of deionized water was added, and the mixture was shaken thoroughly, followed by cooling in water for another 2 min. The absorbance of the solution was measured using a portable COD analyzer with deionized water as the blank. The COD value of the wastewater was determined based on a pre-established calibration curve.

$$\mathrm{COD}\mathrm{degradation}\mathrm{rate}={\displaystyle \frac{{\mathrm{COD}}_{\mathrm{before}}-{\mathrm{COD}}_{\mathrm{after}}}{{\mathrm{COD}}_{\mathrm{before}}}}$$

(1)

where COD_before is the COD content prior to reaction, and COD_after is the COD content after reaction.

5. Conclusions

In summary, both Ag₂₉ and Pt-doped PtAg₂₈ NCs were successfully immobilized onto TiO₂ via electrostatic adsorption, forming a Z-scheme heterojunction system that effectively suppresses charge recombination. XPS and UPS studies reveal the band structures and charge transfer pathways of the different materials. Pt single-atom doping increases the electron density on the Ag surface in PtAg₂₈/TiO₂. Bonding with TiO₂ induces greater band bending, providing a higher driving force for carrier separation and effectively suppressing charge recombination. Photoelectrochemical and PL tests confirm that the PtAg₂₈/TiO₂ composite exhibits superior charge separation efficiency and excellent stability, demonstrating optimal activity, with an H₂ yield of 14,564 μmol g⁻¹ h⁻¹. This represents a 4.1-fold increase over Ag₂₉/TiO₂ and a 10.7-fold increase over pure TiO₂. The composite’s exceptional stability underscores the pivotal role of the central Pt single atoms in the photocatalysis of metallic nanocrystals. This advancement is significant in the preparation of structurally precise and customizable heterogeneous photocatalysts.