Controlling the electron–hole separation in photo-assisted C–C coupling reactions catalysed by a RuO₂–ZnO heterojunction loaded with ultra trace palladium

Abstract

Zinc-oxide (ZnO) in combination with ruthenium-oxide (RuO₂) appeared to be an effective light harvesting heterojunction, controlling the electron (e⁻) and hole (h⁺) recombination rate. An ultra-trace amount of palladium (Pd) loaded into this RuO₂–ZnO hybrid heterojunction enabled a new, sustainable, cost-effective yet highly efficient non-free radical photocatalytic pathway for the Suzuki–Miyaura cross-coupling (SMCC) reaction, achieving superior selectivity and biaryl product yields over existing methods. Although research has aimed at improving photocatalysts' performance in SMCC, no study has yet been performed to understand such heterojunctions in the photocatalytic SMCC reaction. In the catalyst, ZnO retained its original role by absorbing UV-light, while RuO₂, on the other hand, played a crucial role as a quasi-metallic co-catalyst favouring charge-carrier transportation. This cooperative charge-management strategy resulted in prolonged charge separation, enhanced redox-activity, and maximized the utilization of Pd centres. The reaction proceeded with 100% selectivity and provided excellent yields (up to 98%) within a very short reaction time of 70 min in a greener methanol/water (MeOH/H₂O) solvent system. Some of the resulting SMCC keto-derivative products were reduced to synthesize a new range of important alcohol derivatives, further expanding the study. An in-depth kinetic study provided deeper insight into the reaction dynamics, while a systematic mechanistic analysis elucidated the charge carriers involved. Density functional theory (DFT) analysis was also performed to examine and compare the interactions involved in the photocatalysts. This work thus demonstrated a photostable, heterogeneous, recyclable, cost-effective, non-free radical and greener photocatalytic pathway for the SMCC reaction, highlighting the potential of multi-component semiconductor–metal architectures in the selective SMCC reaction with significantly low Pd loading.

---

1. Introduction

The activation of molecules through utilization of light as an energy source or driving organic transformation reactions through naturally or artificially available light sources has emerged as a greener approach and has a promising future for sustainable development.^1–5 Developing such greener synthetic methodologies has appeared to be a top priority in modern chemical research. This is because of the need for reducing energy consumption, minimizing waste, lowering the cost and improving the overall environmental footprint of widely practiced vital chemical reactions.⁶ Therefore, enhancing the sustainability of these chemical processes by developing efficient and sustainable photocatalytic pathways through harvesting light energy can have a wide impact in modern synthetic chemistry. While developing such photocatalytic processes, multiple factors can significantly influence both the reaction rate and the selectivity towards the desired product. The choice of photocatalyst, its light absorption characteristics, surface area, and most importantly the stability of both the catalyst and the reactants under photoirradiation conditions directly affect the overall transformation efficiency of those processes.^7,8 Apart from this, the extent of charge separation and recombination rate between the photogenerated charge carriers, namely electrons (e⁻) and holes (h⁺), plays a decisive role in determining the photocatalytic performance of a photocatalyst.⁹ Therefore, engineering of suitable light harvesting materials that maintain long-lived electron hole separation as well as reluctant to undergo photo-deactivation under photocatalytic conditions is highly demanded in the area of green chemistry. Such materials are crucial for developing greener and clean processes for fine chemical synthesis.

In this context, semiconducting materials like zinc oxide (ZnO) have attracted considerable attention for their high applicability in various photochemical reactions.¹⁰ ZnO serves as an excellent photoactive material owing to its suitable band-gap of ≈3.2 eV for absorption of UV light, high photostability, high surface area and intrinsic charge-transport properties.¹¹ Additionally, the presence of oxygen vacancies (OVs) in ZnO facilitates metal anchoring and provides actual active sites that can participate in the photocatalytic transformations. The presence of oxygen vacancies significantly affects the optical and electronic properties of the metal oxide and thus its photocatalytic performance.¹² Jiang et al. have reported that these OVs can act as anchoring or active sites to stabilize and bind metal species.¹² The high abundance, low cost, and environmentally benign nature of ZnO further support its use as a sustainable photoactive material in photocatalytic transformation reactions.¹³ However, ZnO alone often falls short of fulfilling all the requirements for efficient photoinduced organic transformation reactions. ZnO suffers from rapid recombination of photogenerated electron–hole pairs, which is the primary factor limiting its photocatalytic efficiency.¹⁴ This rapid recombination significantly reduces charge carrier lifetime and quantum efficiency, suppresses the availability of reactive species required for surface redox reactions, and prevents the effective migration of photogenerated charge carriers to active sites, leading to poor catalytic activity even under strong UV light irradiation.^14–16 The limited surface adsorption capacity of ZnO further hinders effective interfacial redox reactions.¹⁷ These inherent limitations significantly impede its photocatalytic efficiency under light irradiation. Therefore, various strategies such as metal and non-metal doping, noble metal deposition, defect engineering, and especially coupling with other semiconductors or metal oxides have been adopted to enhance light absorption and suppress rapid charge carrier recombination of ZnO. Such modifications promote efficient charge separation, improve carrier transport, and ultimately enhance the overall photocatalytic performance of ZnO-based systems. Therefore, constructing heterojunctions by coupling ZnO with suitable semiconductors or metal oxides has emerged as an effective strategy to suppress the issues with ZnO alone such as charge recombination, agglomeration, and photostability, and significantly boost photocatalytic performance. When combined with other metal oxides or suitable cocatalysts, the performance of ZnO based systems improved significantly in terms of product selectivity and more effective control over electron–hole separation and recombination rate of these charge carriers.^18,19 This type of control over the rate of recombination of the charge carriers (e⁻ and h⁺) is of utmost importance in designing efficient photocatalytic pathways. Notably, ZnO not only contributes to light absorption or charge separation but also to the development of effective heterojunctions with conductive or redox active co-catalysts to further boost the photoactivity of the catalyst system. For example, Yang et al. reported a Cu/ZnO photocatalyst for highly selective photosynthesis of bibenzyl via homo coupling.²⁰ Sun et al. also found that among the photocatalytic materials, ZnO based composites showed promising research prospects in photocatalytic reactions.²¹

Like ZnO, ruthenium (Ru) based metal oxides or metal complexes have also demonstrated excellent photoactivity in a wide range of photochemical transformations.^22–25 For instance, [Ru(bpy)₃]²⁺ complexes are very well-known molecular photocatalysts because of their strong visible-light absorption capability, long-lived excited states, and efficient photoredox behaviour.²⁶ The advantage of incorporating ruthenium oxide in designing photocatalytic heterojunctions lies in its excellent electrical conductivity, strong and faster redox activity, the ability to facilitate rapid charge transfer, and the capability to have synergistic interactions with other materials.²⁷ Ru exhibits excellent photoredox behaviour due to its ability to undergo rapid and reversible changes in oxidation states.²⁸ Because of these favourable photophysical and redox properties, numerous researchers have employed Ru-based catalysts in photo-electrocatalytic water splitting and other energy-conversion reactions.^29–31 However, despite the strong potential of RuO₂ as a conductive and redox-active cocatalyst, RuO₂–ZnO heterojunctions remain almost unexplored in photocatalytic organic transformation reactions.

This gap presents a significant opportunity, particularly for C–C bond-forming reactions such as the photocatalytic Suzuki–Miyaura cross-coupling (SMCC) reaction. Although the thermal SMCC reaction has been performed by many researchers using different catalysts of varied nature,^32–36 the reports for photocatalytic SMCC are very few. This scarcity arises from some intrinsic challenges associated with driving C–C coupling through light-induced pathways. Effective photocatalytic SMCC demands meticulously engineered semiconductor–metal catalyst systems that can sustain long-lived electron–hole separation, minimize the e⁻/h⁺ recombination rate, and efficiently harvest light energy to generate the energetic charge carriers required for catalysis.² Additionally, the catalyst must offer well-defined and accessible active sites capable of activating the C–X (X = halogen) bond which is one of the most energy-intensive and mechanistically critical steps in the Suzuki pathway. Apart from that, the actual active site required for SMCC is palladium (Pd), but its high cost remains a major limitation. In due course, although some catalysts have been introduced to perform the photocatalytic SMCC reaction, their effectiveness is still not good enough.^37–42 For example, Han et al. reported a Pd/SiO₂/3% Au/ZrO₂ photocatalyst but found very limited activity (low yield, 12 h reaction time) with no broader substrate scope study.⁴² Xiao et al. also reported a photocatalytic Suzuki cross-coupling reaction on Au–Pd alloy nanoparticles under visible light irradiation but only with 5 substrate molecules and a low % conversion starting from 25%.³⁷ In most of the reported systems, aryl iodides are preferred, but shifting towards more economical substrates like aryl bromides or chlorides is very essential. Also, pervasive reliance on the expensive and rare Pd metal in these reported protocols significantly upsurges the economic burden. One way to overcome these issues is to minimize the Pd loading by integrating trace amounts of Pd with a highly efficient light-harvesting heterojunction. Therefore, although extensively studied, the hunt for finding a suitable photocatalyst with favourable reaction conditions and with minimal utilization of costly Pd is still an ongoing process. Very recently, we reported a low content of Pd loaded into an Fe(III)-modified zeolite-Y catalyst for thermal SMCC reaction; however, this system required relatively high reaction temperatures, questioning the sustainability of the process.⁴³ Furthermore, most of the previously reported photocatalytic SMCC protocols rely on organic solvents such as toluene, acetonitrile, or DMF, which pose environmental and safety concerns due to their high volatility, toxicity, flammability, and disposal issues. Thus, replacing these with greener protic solvents like ethanol, methanol, or water is essential for a more sustainable photocatalytic SMCC process.

Therefore, in an attempt to develop a suitable photocatalyst that can harvest light energy to efficiently control the e⁻/h⁺ recombination by providing a suitable charge carrier pathway and can trigger the SMCC reaction, herein, we report a trimetallic composite of RuO₂–ZnO loaded with a trace amount of Pd in a greener methanol/water solvent system. To the best of our knowledge, photocatalytic SMCC reactions using such photoactive heterojunctions are not well known in the literature. Thus, this research work undertakes a comprehensive and systematic investigation as well as analysis, and obtains mechanistic insights into the photocatalyst and the associated photocatalytic pathway to elucidate the factors enabling an efficient but sustainable photocatalytic SMCC reaction.

2. Experimental section

2.1 Synthesis of ZnO

ZnO was synthesized using a precipitation method. An aqueous solution of 100 mL of 0.5 M ZnCl₂·6H₂O was treated with an aqueous solution of 0.5 M KOH in a round bottomed (RB) flask until complete precipitation of Zn(OH)₂ was observed. The mixture was then stirred for 10 h maintaining a pH of 10–11 followed by filtration, washing with distilled water (DW), and drying at 100 °C. The obtained material was further calcined at 500 °C to get the white coloured ZnO powder.

2.2 Synthesis of PdO/RuO₂–ZnO

The PdO/RuO₂–ZnO material was synthesized by using a coprecipitation method followed by hydrothermal treatment. For that, 500 mg of ZnO was dispersed in DW in an RB flask, followed by addition of 10 mL of 0.01 M PdCl₂ and 50 mL of 0.05 M RuCl₃·H₂O aqueous solution as the source of palladium (Pd) and ruthenium (Ru), respectively. The mixture was stirred at 80 °C for 10 h. The mixture was then autoclaved for 6 h at 150 °C followed by filtration, washing with DW, and drying at 100 °C. The obtained material was further calcined at 550 °C to obtain ∼700 mg of the final black coloured PdO/RuO₂–ZnO material. A schematic representation of this synthesis process is shown in SI, Scheme S1.

2.3 Synthesis of PdO–ZnO

Similar to the PdO/RuO₂–ZnO material, 500 mg of ZnO was dispersed in DW in an RB flask, followed by addition of 25 mL of 0.01 M PdCl₂ aqueous solution as the source of palladium. The mixture was stirred at 80 °C for 10 h. The mixture was then autoclaved for 6 h at 150 °C followed by filtration, washing with DW, and drying at 100 °C. The obtained material was further calcined at 550 °C to obtain ∼520 mg of the final PdO–ZnO material.

2.4 Synthesis of RuO₂–ZnO

The RuO₂–ZnO catalyst was synthesized by following the same procedure as PdO–ZnO, with PdCl₂ replaced by 50 mL of 0.05 RuCl₃·H₂O as the metal precursor to get ∼700 mg of the final RuO₂–ZnO.

2.5 Synthesis of PdO–RuO₂

The PdO–RuO₂ catalyst was synthesized by a coprecipitation method. Initially 50 mL of 0.01 M PdCl₂ and 50 mL of 0.05 M RuCl₃·H₂O aqueous solution were taken in an RB flask as the source of palladium (Pd) and ruthenium (Ru), respectively. To this 0.05 M KOH solution was added dropwise until a pH of 10–11 was obtained. The mixture was then stirred at 80 °C for 10 h. The mixture was then autoclaved for 6 h at 150 °C followed by filtration, washing with DW, and drying at 100 °C. The obtained material was further calcined at 550 °C to obtain ∼320 mg of the final PdO–RuO₂ material.

2.6 Procedure for the photocatalytic SMCC reaction

The photocatalytic SMCC reaction was carried out in a standard photochemical reactor set up having a 500 mL reaction chamber made up of quartz. A 450 W medium-pressure mercury lamp (λ = 254 nm) was used as the source of UV irradiation. The lamp was encircled by a double jacketed quartz immersion well. The well was fitted with water inlet and outlet connections. Water was continuously circulated through the inlet and outlet of the double-jacketed quartz immersion well to control the temperature of the reaction, ensuring that any catalytic activity observed could be attributed solely to UV irradiation or photon energy, eliminating possible contributions from thermal energy. The photocatalytic reaction was carried out by taking 1 mmol of aryl halide (CHOC₆H₅Br), 1.2 mmol phenylboronic acid [PhB(OH)₂], and 3 mmol potassium carbonate (K₂CO₃) in a methanol/water solvent system (3 : 1, v/v) followed by the addition of 20 mg of the PdO/RuO₂–ZnO catalyst. The reaction mixture was then stirred under UV-light irradiation for 70 min. The product formed was separated by column chromatography and the % yield was calculated as isolated % yield.

3. Results and discussion

A comprehensive characterization of the synthesized catalysts was conducted through various spectroscopic and analytical methods prior to their catalytic application and the results are discussed in the succeeding sections. As discussed in the experimental section, a series of metal oxide-based catalysts, namely PdO–ZnO, RuO₂–ZnO, PdO–RuO₂ and the combined PdO/RuO₂–ZnO composites, were synthesized and duly characterized.

Powder X-ray Diffraction (PXRD) analysis was performed to investigate the structural information, crystallinity and phase compositions of the synthesized materials. The PXRD patterns of the pristine ZnO, PdO–ZnO, RuO₂–ZnO, and the hybrid PdO/RuO₂–ZnO catalysts are shown in Fig. 1a. The PXRD pattern of the PdO/RuO₂ catalyst is provided in Fig. S1, SI. Comparison of the obtained diffraction pattern of ZnO (Fig. 1a, black line) with standard JCPDS data (JCPDS 00-003-0888) confirmed the presence of a pure hexagonal wurtzite phase of ZnO.⁴⁴ In the case of the PdO–ZnO composite material, the PXRD pattern exhibited additional distinct peaks corresponding to (101), (110), (112), (103), and (211) planes of PdO (JCPDS 00-006-0515), along with the characteristic peaks of ZnO (Fig. 1a, red line). Similarly, apart from the characteristic peaks of ZnO, the PXRD pattern of RuO₂–ZnO also showed a few additional peaks corresponding to (110), (101), and (211) planes of RuO₂ (JCPDS 00-018-1139), Fig. 1a (green line). These observations indicated the successful formation of these ZnO supported composite binary metal oxides (Fig. 1a red line and green line). Notably, the PXRD pattern of the trimetallic PdO/RuO₂–ZnO catalyst displayed well-defined peaks attributed to all of its three components—PdO, RuO₂, and ZnO— as shown by the distinct (101), (110), and (211) planes of PdO; (110), and (211) planes of RuO₂; and (101), (102), (110), (103), and (112) planes of the wurtzite phase of ZnO (Fig. 1a, blue line). This observation thereby validated the formation of a true ternary PdO/RuO₂–ZnO composite material. The presence of sharp and well-defined peaks in all the XRD patterns clearly indicated the high crystallinity and well-ordered structural arrangement of the synthesized materials.⁴⁵ Further, the Scherrer equation was used to calculate the crystallite sizes of the nanomaterials.⁴⁵ The crystallite size of pristine ZnO was found to be 13.04 nm. However, upon the incorporation of PdO and RuO₂, the crystallite sizes increased significantly to 24.96, 24.27, and 22.99 nm for PdO–ZnO, RuO₂–ZnO, and PdO/RuO₂–ZnO, respectively. This might be due to Ostwald ripening.⁴⁶ This notable increase in the crystallite size indicated the successful integration of PdO and RuO₂ into the ZnO support matrix. Anandan et al. reported that this type of incorporation of nickel into ZnO nanoparticles led to a noticeable escalation in crystallite size, increasing from 31 nm in pristine ZnO to 45 nm in the Ni-doped samples.⁴⁷ Jha et al. also observed an increase in the crystallite size of ZnO-supported catalysts when incorporated with ruthenium (Ru) as compared to the neat ZnO.⁴⁸ Similarly, some other previous studies also reported that the incorporation of foreign metals or metal oxides into ZnO led to larger crystallite sizes compared to those of pure ZnO, highlighting the influence of dopants on structural growth and ordering.^49,50 It is also persistent to note that the peak position corresponding to the (101) plane of ZnO shifted from ∼36.14° in pristine ZnO to ∼36.24° in the PdO–ZnO, RuO₂–ZnO, and PdO/RuO₂–ZnO catalysts, Fig. 1b. The FWHM of this particular peak of pristine ZnO was also found to decrease upon introduction of the foreign metal oxides, reducing from 0.67 nm in pristine ZnO to ∼0.35–0.38 nm in the PdO–ZnO, RuO₂–ZnO, and PdO/RuO₂–ZnO composites. This again supported the crystallite growth induced by the incorporation of PdO and RuO₂ species. Aranganayagam et al. reported this type of decrease in FWHM values due to crystallite growth when pristine ZnO was incorporated with Ru.⁵¹ The intensity of the peaks for (110) and (101) planes of RuO₂ got significantly diminished in the PXRD pattern of the ternary PdO/RuO₂–ZnO composite implying modification of these planes as compared to the binary RuO₂–ZnO catalyst, Fig. 1b (green line and blue line). As mentioned earlier, the PXRD pattern of the unsupported PdO/RuO₂ catalyst is provided in SI, Fig. S1. The pattern confirmed the presence of both RuO₂ and PdO crystalline phases in the unsupported catalyst.

Fig. 1 a) PXRD patterns of the synthesized catalysts, b) enlarged view of the PXRD pattern in the 2θ values 26–40°, and c) Raman spectra of the synthesized catalysts.

Raman spectroscopy was further carried out to supplement the PXRD findings and the results are provided in Fig. 1c. Raman spectra of neat ZnO showed peaks at 331, 386, 439, 584, 648, 1086, and 1153 cm⁻¹ (Fig. 1c, blackline). The peak at 331 cm⁻¹ was attributed to the multi-phonon process associated with the E^high₂–E^low₂ second-order vibration mode of ZnO.⁵² The signal at 386 cm⁻¹ corresponded to the A^TO₁ mode, while the most intense signal at 439 cm⁻¹ was assigned to the E^high₂ vibrational mode, indicating the presence of hexagonal wurtzite structure of ZnO with high crystallinity.⁵² These results aligned with the PXRD data, confirming the wurtzite phase of ZnO. The peak at 584 cm⁻¹, associated with the A^LO₁ mode, reflected the presence of oxygen vacancies in the ZnO material.⁵³ The Raman spectrum of ZnO indicated the presence of lattice defects in the material, with the band at 648 cm⁻¹ attributed to a local vibrational mode arising from defect sites within the host lattice.⁵⁴ The higher wavenumber peaks at 1086 and 1153 cm⁻¹ were ascribed to A₁(TO + LO) phonon combinations at the A and H symmetry points and to the E₁(2LO) second-order mode in ZnO, respectively (Fig. 1c, blackline).⁵³ In the Raman spectra of PdO–ZnO (Fig. 1c, red line), apart from the signals for ZnO, a few additional distinct signals were observed at 275, 641 and 711 cm⁻¹ characteristic of the PdO species.^55,56 The signals at 275 and 711 cm⁻¹ were due to the χ₈ phonon modes of PdO species, while the signal at 641 cm⁻¹ was for the B_1g mode.⁵⁶ These peaks confirmed the successful incorporation of PdO onto ZnO and supported the formation of the PdO–ZnO composite material. Likewise, in the case of the RuO₂–ZnO material, the extra peaks observed at 495 and 604 cm⁻¹ were attributed to the E_g and A_1g vibrational modes characteristic of the first-order phonon bands of RuO₂ species (Fig. 1c, green line).⁵⁷ The Raman spectra of the trimetallic PdO/RuO₂–ZnO material displayed characteristic signals from all of the three metal oxide components indicating successful incorporation of all the three metal oxides in the ternary hybrid material (Fig. 1c, blue line). In the Raman spectra of PdO/RuO₂–ZnO, the peaks at 273 and 636 cm⁻¹ corresponding to χ₈ and B_1g modes of PdO, a small peak at 506 cm⁻¹ attributed to the E_g vibrational mode of RuO₂, and peaks at 430 and 1067 cm⁻¹ associated with the E^high₂ vibrational mode and A₁(TO + LO) mode of ZnO were clearly identified.^{52,53,55–57} Notably, some shifts in the Raman bands were observed across all Raman spectra implying the impact of PdO or RuO₂ loading on the vibrational modes of ZnO and vice versa. These shifts might have originated from the changes in crystallinity nature of the ZnO supported hybrid materials as compared to the pristine ZnO.⁵⁷ The size confinement effect triggered by the incident phonon in the ZnO matrix might have also influenced these shifts after the incorporation of PdO or RuO₂ over the ZnO structure.⁵⁷ An interesting observation from the Raman analysis was that the higher-wavenumber peaks at 1086 and 1153 cm⁻¹ in the pristine ZnO material corresponding to the A₁(TO + LO) phonon combinations and the E₁(2LO) second-order mode exhibited significant variations after incorporation of PdO or RuO₂ into the ZnO matrix, Fig. 1c. The intensity of these bands increased in RuO₂–ZnO but decreased in the case of PdO–ZnO and PdO/RuO₂–ZnO. Since these bands reflected multi-phonon scattering in ZnO, their intensity was quite sensitive to any changes in the electron–phonon interactions and defect states like lattice disorder, oxygen vacancies, and dopants.⁵⁸ In the RuO₂–ZnO material this increase in Raman intensity of the band at 1085 cm⁻¹ might be due to the transfer of electrons between RuO₂ and ZnO (n-type enhancement) (Fig. 1c, green line). This led to the formation of an interfacial local electric field and strengthened the Fröhlich interaction (LO phonon–electron coupling).⁵⁹ This enhancement of the electron–phonon coupling further promoted the multi-phonon Raman scattering and amplified the higher-order Raman bands at 1085 cm⁻¹ in the Raman spectra of RuO₂–ZnO.⁵⁹ Conversely, the reduced intensity of the combination of E₁(LO) and A₁(LO) modes in PdO–ZnO and PdO/RuO₂–ZnO catalysts might have arisen from lattice irregularities, particularly oxygen-related defects, that disrupt the displacement of ions along the ab-axis for the E₁ mode and along the c-axis for the A₁ mode (Fig. 1c, red line and blue line).^58,60 The Raman spectra for the unsupported PdO/RuO₂ catalyst showed the corresponding bands characteristic of both PdO and RuO₂ phases of the metal oxides confirming the coexistence of both metal oxides in the material, Fig. S2, SI.

To gain additional insights into the chemical states of the constituent elements, X-ray Photoelectron Spectroscopy (XPS) was also performed on the synthesized catalysts and the spectra are provided in Fig. 2, S3 and S4. The deconvoluted spectrum of the Pd 3d region of the PdO/RuO₂–ZnO catalyst exhibited a pair of relatively low-intensity peaks ascribed to the 3d_5/2 and 3d_3/2 spin–orbit components with binding energy (B.E.) values of 336.2 eV and 341.8 eV, respectively, Fig. 2a. These values align well with those reported for Pd(II) or PdO, confirming its presence in the catalyst.^43,61–63 The low intensity of the Pd 3d peaks indicated a low concentration of PdO species in the sample, resulting in noisy and poorly resolved peaks, Fig. 2a. The Ru 3d spectrum of the same catalyst exhibited peaks at 280.6 and 284.8 eV, attributed to the 3d_5/2 and 3d_3/2 states of Ru(IV) in RuO₂, while the low intensity peaks at 280.1 and 284.1 eV were for the 3d_5/2 and 3d_3/2 states of Ru(0) species in the sample, Fig. 2b.^63–65 This implied that RuO₂ was also associated with some low content of Ru(0) species. The presence of peaks at B.E. values 1022.5 eV and 1045.6 eV in the XPS of Zn 2p for the 2p_3/2 and 2p_1/2 states confirmed the presence of Zn in the +2 oxidation state in ZnO, Fig. 2c.⁶³ The O 1s spectrum revealed three distinct peaks at 529.1, 530.4, and 531.6 eV, corresponding to lattice oxygen, oxygen vacancies or defects, and adsorbed oxygen species in the synthesized material, respectively, Fig. 2d.^63–65 The XPS analyses of the other synthesized materials are presented in the SI (Fig. S3 and S4), together with their corresponding B.E. values and respective assignments of the oxidation states summarized in Table S1. The B.E. values provided in Table S1 confirmed the presence of Pd(II) and Zn(II) in the palladium oxide supported on zinc oxide material (PdO–ZnO), and Ru(IV) along with Zn(II) in the ruthenium oxide supported on zinc oxide material (RuO₂–ZnO), consistent with the structural features revealed by PXRD and Raman analyses.

Fig. 2 X-ray photoelectron spectra of a) Pd 3d, b) Ru 3d, c) Zn 2p, and d) O 1s of PdO/RuO₂–ZnO.

Further to evaluate the interfacial interaction and charge transfer due to integration of the constituent metal-oxides in the ternary hybrid material, the XPS spectra of pristine PdO, RuO₂, and ZnO were recorded and compared with that of the integrated hybrid ternary PdO/RuO₂–ZnO composite. The XPS of pure PdO showed Pd 3d_5/2 and Pd 3d_3/2 peaks at ∼336.8 and ∼342.2 eV, confirming the Pd²⁺ oxidation state in pure PdO, while the XPS of pure RuO₂ displayed Ru 3d_5/2 at ∼280.5 eV and Ru 3d_3/2 at 284.75, corresponding to Ru⁴⁺ of RuO₂, Fig. S5a and b.⁶⁶ The XPS of pure ZnO exhibited Zn 2p_3/2 and Zn 2p_1/2 peaks centered at ∼1021.9 and ∼1045.1 eV, respectively, characteristic of Zn²⁺ in the ZnO lattice, Fig. S5c.⁶⁶ The B.E. values are compared in Table S2, SI. It was found that the B.E. values for Zn in the hybrid ternary catalyst were comparatively higher than that of the pure ZnO. Similarly, B.E. values of Ru were slightly higher in the ternary catalyst than that of pure RuO₂, indicating a decrease in local electron density around Zn and Ru species.⁶⁷ In contrast, the BE values in the case of Pd for the PdO/RuO₂–ZnO catalyst were found to be lower in comparison to pure PdO, signifying an increase in electron density at the Pd sites of PdO in the PdO/RuO₂–ZnO catalyst.⁶⁸ This opposite trend suggested directional charge transfer from ZnO and RuO₂ to PdO upon heterostructure formation. These observations also implied that both Ru and Zn participated in transfer of electron density from RuO₂ and ZnO towards PdO across an interfacial junction formed in the ternary catalyst.⁶⁷ Upon interfacial contact due to the difference in Fermi level of PdO and ZnO, Fermi level equilibration might have driven this electron transfer from ZnO and RuO₂ towards PdO, favouring the electron hole separation.^69,70 Thus, this change in the B.E. values implied that Pd served as an electron sink in the heterostructure, functioning as an active cocatalytic center in the photocatalyst.⁷⁰

Field Emission-Scanning Electron Microscopy (FE-SEM) analysis was performed to investigate the surface morphology and dispersion of PdO and RuO₂ species over the ZnO support matrix in the synthesized materials, Fig. 3. In all cases, the ZnO support matrix appeared as aggregated rutile-like flakes, mostly nearly-spherical, showing a textured surface due to the presence of PdO or RuO₂. PdO and RuO₂ appeared as small nanoparticles dispersed over these ZnO flakes, indicating their successful incorporation into the support matrix. For the PdO/RuO₂–ZnO composite (Fig. 3a and b), a dense nanoparticles' surface was evident, where the fine PdO and RuO₂ domains were spread over the ZnO framework, promoting more heterogeneous surface exposure. In the case of PdO–ZnO (Fig. 3c and d), comparatively finer and more evenly distributed nanoparticles were observed decorating the ZnO surface, forming an aggregated nanostructure. On the other hand, the RuO₂–ZnO catalyst (Fig. 3e and f) exhibited some rod-like RuO₂ species spreading over the ZnO surface, highlighting a distinct growth behaviour. The FE-SEM images of the unsupported PdO/RuO₂ catalyst are provided in Fig. S6.

Fig. 3 FE-SEM images of (a and b) PdO/RuO₂–ZnO, (c and d) PdO–ZnO, and (e and f) RuO₂–ZnO.

Transmission Electron Microscopy (TEM) analysis was also carried out to gain further insights in the morphology and distribution of PdO and RuO₂ species over the ZnO support, Fig. 4. The TEM images confirmed the presence of small square and rod-like PdO and RuO₂ nanoparticles deposited over the nearly-spherical ZnO support matrix in the PdO/RuO₂–ZnO material, Fig. 4a–h. Elemental contrast clearly distinguished the heavier Pd and Ru species from the lighter ZnO background, evidencing their dispersion across the matrix. The fringe pattern of the PdO/RuO₂–ZnO catalyst supported the presence of the (100) plane of PdO (JCPDS 00-006-0515), (211) and (222) planes of RuO₂ (JCPDS 00-018-1139), and (101), (105), and (200) planes of ZnO (JCPDS 00-003-0888), representing the successful coexistence of all three metal oxides in the ternary catalyst, Fig. 4d. The TEM images of PdO–ZnO (Fig. 4i–l) and RuO₂–ZnO (Fig. 4m–o) also revealed the presence of the respective metal oxides on the ZnO surface, as disclosed by the elemental contrast with the heavier metals appearing as darker spots compared to the light coloured ZnO background. The images also displayed lattice fringes across all the synthesized materials, highlighting the crystalline nature of the materials, Fig. 4. The PdO/RuO₂ material was also analysed by TEM analysis and the images are provided in Fig. S7.

Fig. 4 TEM images of (a–h) PdO/RuO₂–ZnO, (i–l) PdO–ZnO, and (m–o) RuO₂–ZnO.

The distinct morphological and structural characteristics observed from FE-SEM and TEM analyses of the synthesized materials also emphasize the significance of the structure–activity relationship in the designed catalysts. As observed from both the FE-SEM and TEM images of the ternary PdO/RuO₂–ZnO catalyst, the dense and almost homogeneous dispersion of palladium and ruthenium oxide nanoparticles over the zinc oxide matrix resulted in the formation of abundant nanoscale heterointerfaces and hence more actual catalytically active sites, Fig. 3a and b and 4a–h. These types of heterointerfaces are structurally favourable for efficient charge separation and interfacial charge migration in photocatalytic reactions.⁷¹ Apart from that, the intimate lattice contact and high crystallinity observed in the ternary PdO/RuO₂–ZnO system suggested strong electronic coupling among the constituent oxides, essential for minimizing recombination losses and enhancing redox efficiency in photocatalytic reactions.⁷² As evident from the TEM images, the presence of rod- and cube-like nanostructures in the PdO/RuO₂–ZnO ternary catalyst was also structurally beneficial for photocatalytic applications. Such well-defined morphologies generally promote photocatalytic performance by providing shorter diffusion pathways for photogenerated charge carriers, thereby reducing the probability of bulk recombination.^73,74 In addition, anisotropic and faceted structures as described by their respective miller indices and lattice fringes in the TEM description typically exposed more reactive surface planes and enhance surface-active site availability, Fig. 4d.⁷⁵ Moreover, the surface texturing in the PdO/RuO₂–ZnO catalyst provided greater exposure of catalytically active sites and improved adsorption capability, Fig. 3a and b. The well-resolved lattice fringes observed in the ternary catalyst also indicated the presence of exposed crystalline facets, Fig.4d. Such defined crystal planes often possess higher surface energy and enhanced catalytic reactivity, which can facilitate improved adsorption and surface redox processes.⁷⁶ In contrast, the binary catalysts namely PdO–ZnO and RuO₂–ZnO did not exhibit similarly distinct lattice fringe patterns, implying fewer exposed high-energy facets and less defined interfacial crystallinity. This structural difference suggested a comparatively lower density of highly reactive surface planes in the binary systems, potentially contributing to reduced catalytic performance. The comparatively fewer, agglomerated and less integrated interfaces in the binary systems indicated limited pathways for charge transfer.⁷⁷ The aggregation of metal oxide domains observed in the TEM images of the binary catalysts indicated a possible reduction in catalytic efficiency. This is because such agglomeration decreased the effective surface area and active site accessibility, and thus had the possibility of limiting the formation of efficient heterointerfaces.⁷⁸ This increased the likelihood of charge recombination and diminishing photocatalytic performance in the binary catalysts as compared to the ternary one. These observations highlighted the significance of the structure–activity relationship of the synthesized materials and the possibility of the role of nanoscale architecture, interfacial density, and dispersion uniformity in photocatalytic performance.

The EDX analysis of the PdO/RuO₂–ZnO catalyst further validated the morphological features observed in the SEM and TEM images. The signals corresponding to Zn in the EDX colour mapping (Fig. 5e) confirmed the presence of nearly-spherical ZnO domains as already ascribed in the SEM and TEM images, Fig. 5a and e. The coloured EDX mapping also validated that the rod/square-like structures observed in the micrographs were not associated with ZnO but were instead recognized to be of the heavier metal oxides of Pd and Ru. This was clearly evidenced by the colour mapping (Fig. 5b–e), displaying the elemental presence of only Ru and Pd with no Zn element detected in the region highlighted by the yellow-marked lines. Furthermore, EDX analysis of regions outside the nearly spherical domain of ZnO, as illustrated by the yellow highlighted circle in Fig. 5e, showed no Zn signals. This again confirmed that the nanoparticles in those areas correspond to dispersed RuO₂ or PdO species of the PdO/RuO₂–ZnO catalyst. All of these observations provided strong confirmation that the PdO and RuO₂ species were successfully deposited on the ZnO surface and were responsible for the distinct rod/square-like features, supplementary to the SEM and TEM results. The EDX spectrum of PdO/RuO₂–ZnO revealed a Zn content of 50.24 wt%, Ru at 24.04 wt%, and Pd at a very low loading of 0.4 wt%, Fig. 5g. These results were in line with the deliberate ultra trace loading of Pd intended in this catalyst system. ICP analysis of the PdO/RuO₂–ZnO catalyst also supported this observation, showing the ultra-trace amount of Pd (0.009 ppm). The ICP results for the PdO/RuO₂–ZnO catalyst are provided in the SI, Table S3. As discussed earlier, the XPS spectra also vouched for these low Pd loading results, demonstrated by the distinctly low-intensity Pd peaks. The EDX analysis of the other catalysts, namely PdO–ZnO, RuO₂–ZnO and PdO/RuO₂, is provided in the SI, Fig. S8–S10. The analysis confirmed the presence of the respective elements in these synthesized catalysts.

Fig. 5 a) Electron image, b) EDX layered image, EDX color mapping of c) Ru, d) Pd, e) Zn, and f) O of the PdO/RuO₂–ZnO catalyst, g) EDX spectrum of PdO/RuO₂–ZnO with the inset showing respective weight %, and h–k) AFM images of the PdO/RuO₂–ZnO catalyst.

Atomic force microscopy (AFM) analysis was also employed to investigate the surface topography of the PdO/RuO₂–ZnO catalyst. The AFM images, as shown in Fig. 5(h–k), revealed a rough and heterogeneous surface decorated with nanoscale protrusions that could be attributed to the PdO and RuO₂ nanoparticles anchored onto the ZnO support matrix. The higher vertical scale observed in the AFM images of the PdO/RuO₂–ZnO catalyst reflected high surface roughness due to protruding nanoparticles and hence greater surface area and catalytically active sites, expecting high catalytic activity of the catalyst. The AFM images of the other catalysts are provided in SI, Fig. S11–S13. Some isolated aggregated large peaks in the AFM topography map were observed for the other catalysts indicating a localized protrusion on the sample surface, typically an agglomerate of nanoparticles. Such aggregation likely concentrated the active sites of the catalysts but might have reduced their accessible surface area, Fig. S11–S13.

Following the morphological and compositional confirmation, the UV-vis diffused reflectance spectra (UV-DRS) of the synthesized catalysts revealed distinct absorption features arising from the incorporation of PdO and RuO₂ species into the ZnO matrix, Fig. 6a. Pure ZnO exhibited a sharp absorption edge at ∼362 nm, originating from a charge-transfer process from the valence band to the conduction band of ZnO (O_2p → Zn_3d), (Fig. 6a, black line).^79,80 In contrast, upon PdO or RuO₂ loading, an additional absorption band at around ∼224 nm was observed for all other three materials namely PdO–ZnO, RuO₂–ZnO, and PdO/RuO₂–ZnO, Fig. 6a. These bands were ascribed to O²⁻→metal (Pd²⁺ or Ru⁴⁺) charge transfer transitions in these materials.^81,82 Furthermore, lower energy absorption bands were also observed with absorption peaks at ∼421 nm for PdO–ZnO, ∼477 nm for RuO₂–ZnO, and a further red-shift peak at ∼526 nm for the ternary PdO/RuO₂–ZnO system, Fig. 6a. This was because of the enhanced absorption intensity of these modified catalysts in the visible region (400–700 nm) compared to the pristine ZnO. This behaviour could be attributed to interfacial electronic coupling between ZnO and the narrow band gap oxides (PdO and RuO₂) in the modified catalysts, which introduced defect-mediated intermediate energy levels.¹⁶ In the modified catalysts incorporation of PdO and RuO₂ introduced additional electronic or defect states near the conduction or valence band edges and promoted interfacial charge-transfer transitions and hence enhanced the visible-light absorption.¹⁶ This red shift could also be ascribed to the synergistic interaction between the metal oxide components in the modified catalysts further establishing the formation of a heterojunction at their interfaces.⁸³ The ternary PdO/RuO₂–ZnO system benefited the most from synergistic interfacial interaction, and thus exhibited the most pronounced visible absorption at ∼526 nm, indicating effective band structure modulation and improved photoresponse relative to the binary counterparts. Further, to have a better visualization, the band gaps of the synthesized materials were calculated using Tauc plots and the same has been produced in Fig. 6b–e. The band gaps of the synthesized materials revealed negligible changes in the values upon incorporation of PdO and RuO₂ into the ZnO matrix. While pristine ZnO exhibited a band gap of 3.11 eV, the introduction of PdO slightly increased the band gap to 3.74 eV in the PdO–ZnO material. Similarly, RuO₂ incorporation resulted in a slightly higher value of 3.37 eV in the RuO₂–ZnO material. The combined PdO/RuO₂–ZnO system displayed a shift, reaching 3.78 eV, suggesting a favourable synergistic electronic interaction between the metal oxides (PdO or RuO₂) and the ZnO matrix. Importantly, the improved visible response originated from defect-mediated and interfacial charge-transfer transitions or possible Burstein–Moss-type band filling effects rather than intrinsic band gap narrowing, again confirming effective heterojunction formation without compromising the fundamental band structure of ZnO in the modified catalysts.^16,83–85 The UV-DRS spectrum of the unsupported PdO/RuO₂ catalyst is shown in Fig. S14. The Tauc plot of PdO–RuO₂ identified a bandgap of value of 2.26 eV, Fig. S15.

Fig. 6 a) UV-DRS spectra of all the synthesized materials, direct band gaps of b) PdO/RuO₂–ZnO, c) PdO–ZnO, d) RuO₂–ZnO, and e) pristine ZnO, and f) PL spectra of all the synthesized materials with the inset showing an enlarged view of the spectra for PdO/RuO₂–ZnO (blue line), RuO₂–ZnO (green line), and PdO–ZnO (red line).

Based on the observed band gap values, the materials were considered to be promising candidates for UV light assisted photocatalytic reactions. However, apart from the band gaps, the recombination rates of photogenerated electron–hole (e⁻–h⁺) pairs in the photocatalyst system play a crucial role in determining photocatalytic efficiency. Photoluminescence (PL) analysis was carried out to evaluate these recombination rates.⁸⁶ PL spectroscopy is widely recognized as a reliable approach to evaluate the recombination rates of photogenerated electron–hole (e⁻–h⁺) pairs in photocatalysts.⁸⁶ Previous studies showed that a reduction in the peak intensity of the PL spectra would reflect a minimized degree of e⁻–h⁺ recombination rate thereby reflecting a better photocatalyst.⁸⁷ Accordingly, PL measurements were carried out for the synthesized materials at an excitation wavelength of ∼224 nm to gain insights into their charge carrier recombination behaviour and the results are provided in Fig. 6f. Compared to the pristine ZnO, PdO–ZnO, and RuO₂–ZnO, the PL intensity of the trimetallic PdO/RuO₂–ZnO catalyst was significantly lowered, implying the superior capability of this material to act as a photocatalyst by suppressing the e⁻–h⁺ recombination rate. The PL spectrum of the unsupported PdO/RuO₂ catalyst also showed a higher PL intensity as compared to the trimetallic PdO/RuO₂–ZnO catalyst, Fig. S16 and S17. This suggested that the presence of the ZnO support matrix was also crucial in suppressing charge-carrier recombination. A bar diagram showing the PL intensity vs. expected photocatalytic efficiency of the synthesized materials is provided in SI, Fig. S17.

4. Computational study

A density functional theory (DFT) analysis was conducted to examine the interaction between PdO, RuO₂ and PdO/RuO₂ composite clusters adsorbed on ZnO surfaces in the gas phase using the Gaussian 16 software package.⁸⁸ The calculations were performed using the B3LYP hybrid functional in conjunction with the density fitting triple zeta valence with double polarization (def2-TZVPP) basis set.^89,90

At the initial stage, the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) isosurface of PdO, RuO₂ clusters and the ZnO surface was calculated as depicted in Fig. S18. The HOMO–LUMO analysis of the ZnO surface revealed that the electron density of the HOMO was primarily located on the oxygen (O) atoms, whereas the LUMO electron density was concentrated on the zinc (Zn) atoms. Similarly, the HOMO–LUMO analysis of PdO and RuO₂ clusters also revealed that the electron density of the HOMO was located on oxygen and the LUMO on Pd and Ru atoms. As the HOMO in the ZnO surface was mainly located at the O atoms, so it was more preferable for Pd and Ru atoms of PdO and RuO₂ nanoclusters to interact with the O atoms of the ZnO surface as compared to the Zn atoms. The adsorption energies of PdO, RuO₂, and the PdO/RuO₂ composite at the ZnO surface were found to be −119.94, −104.52, and −203.27 kcal mol⁻¹, respectively as shown in Table S4. These findings further indicated that while both PdO and RuO₂ nanoclusters exhibit strong individual adsorption on the ZnO surface, the PdO/RuO₂ composite demonstrates a significantly enhanced stability with an adsorption energy of −203.27 kcal mol⁻¹. The HOMO–LUMO isosurfaces plots for the adsorption of PdO, RuO₂ and the PdO/RuO₂ composite clusters on ZnO surfaces is presented in Fig. 7.

Fig. 7 Optimized structures, HOMO, and LUMO isosurfaces plots, respectively for the adsorption of (a–c) PdO on ZnO, (d–f) RuO₂ on ZnO, and (g–i) PdO–RuO₂ composite clusters on ZnO surfaces.

4.1 Partial density of states (PDOS) analysis

The partial density of states (PDOS) analysis was carried out using the Multiwfn program to examine the contributions of various atomic orbitals to the electronic structure.⁹¹ From the PDOS plots for Pd in PdO and the O atom of the ZnO surface in the PdO–ZnO material, it was found that there was significant overlapping between the Pd 4d and O 2p orbitals in the energy range just below the Fermi level, Fig. S19. It indicated strong hybridization and predicted covalent bonding characteristics between the Pd and O atoms in this region. The presence of the Pd 5p orbital at higher energies (above the Fermi level) and its lack of overlap with oxygen (O) states suggested no significant contribution to bonding. Similarly, the Pd 5s orbital showed some overlap with O 2p and 3p states indicating some weaker hybridization. Overall, the PDOS analysis revealed that the electronic interactions between Pd and O are dominated by Pd 4d and O 2p orbital hybridization, which played a key role in determining the electronic structure and bonding nature in PdO and the ZnO surface.

Again, the PDOS analysis for Ru in RuO₂ and the O atom of the ZnO surface in RuO₂–ZnO indicated significant overlapping between the Ru 4d and O 2p orbitals, predicting strong hybridization and covalent bonding between Ru and O atoms, Fig. S20. The Ru 5p orbital showed a sharp, unoccupied peak just above the Fermi level, indicating no contribution to bonding. The Ru 5s orbital overlapped weakly with O 2p and 3p states, showing limited but present hybridization. The Fermi level's position near a high intensity peak of the Ru 4d state also implied a high density of available states at the Fermi level, further suggesting metallic or semi-metallic behaviour. Overall, the plot highlighted strong Ru–O orbital interactions primarily via Ru 4d and O 2p, which played a key role in determining the electronic structure and conductivity of RuO₂ when interfaced with the ZnO surface.

The PDOS analysis for Pd in the PdO/RuO₂ composite and O atom of the ZnO surface in the PdO/RuO₂–ZnO material is shown in Fig. S21. The Zn 3d orbital exhibited a sharp, intensely bound peak with very minimal overlapping with the Pd 4d orbitals predicting covalent bonding between Pd and O atoms. The Pd 5p orbital was prominently well above the Fermi level indicating that they remained unoccupied and did not contribute to bonding. The Fermi level itself positioned at the edge of the Pd 4d band, suggesting limited metallic character and indicating that Pd contributes more significantly to the electronic states near the conduction edge, while Zn's contribution is primarily from its deep-lying 3d orbitals.

Thus, PDOS analysis of these three systems revealed distinct electronic interactions, with the PdO/RuO₂ composite supported over the ZnO surface (PdO/RuO₂–ZnO) emerging as the most favourable based on direct metal–metal orbital coupling. In the PdO/RuO₂–ZnO material, a clear overlap was observed between the Pd 4d and Zn 3d orbitals indicating strong d–d hybridization. This type of interaction is particularly significant since metal–metal d–d orbital coupling can enhance electron delocalization, support metallic conductivity, and improve charge transport across interfaces, especially in nano-heterostructures. Cartesian coordinates of all the species are given in Table S5.

5. Catalytic study

To assess the catalytic efficiency of the synthesized materials, the catalytic investigation commenced with the development of standard reaction conditions to evaluate the activity of the synthesized catalysts in the photocatalytic Suzuki–Miyaura cross-coupling (SMCC) reaction. The synthesized PdO/RuO₂–ZnO, PdO–ZnO, and RuO₂–ZnO catalysts in the present work exhibited suitable band gaps comparable to other wide band gap semiconductors, enabling efficient light absorption in the UV region. The PL analysis of these systems suggested the capability to promote effective photoinduced charge separation and recombination that were essential for driving photocatalytic reactions. Therefore, owing to their favourable electronic properties and the synergistic interaction between PdO, RuO₂ and ZnO species, the catalysts were employed for the photocatalytic SMCC reaction.

For that, a reaction mixture containing aryl halide (CHOC₆H₅Br, 1 mmol), phenylboronic acid [PhB(OH)₂, 1.2 mmol], and cesium carbonate (Cs₂CO₃, 3 mmol) in a methanol and water solvent system (CH₃OH/H₂O in 3 : 1, v/v) was stirred in the presence of 15 mg of the synthesized catalysts under UV light irradiation for a duration of 20 min, Fig. 8a. During this screening of the synthesized catalysts for the photocatalytic SMCC reaction, the PdO/RuO₂–ZnO catalyst exhibited the highest activity, affording a yield of 45% of the desired biaryl product surpassing the performance of all other investigated catalysts, Fig. 8a. Although the yield was relatively modest, the reaction proceeded with remarkable selectivity with no traces of side-products, signifying a positive direction of the reaction. Because, despite the fact that photocatalytic reactions are generally energy efficient, the photocatalytic SMCC reaction involves various challenges such as protodeboronation of boronic acids, dehalogenation of aryl halides through single-electron transfer (SET), oxygen interference, and radical–radical recombination pathways. So, even though the product yield under the initial reaction conditions was not so satisfactory, the high selectivity with no formation of side products with the PdO/RuO₂–ZnO catalyst in this photocatalytic reaction was quite remarkable. Therefore, to improve the yield, the reaction was further investigated by optimizing various key reaction parameters, including the solvent system, base, irradiation time, and catalyst amount taking PdO/RuO₂–ZnO as the test catalyst. In photocatalytic SMCC reactions, solvent optimization is vital because solvent polarity, proticity, and miscibility strongly influence the photocatalyst's excited-state lifetime, electron-transfer efficiency, and stability thereby governing both yield and selectivity. Our earlier experimental results suggested that a 3 : 1 ratio of organic solvent: water exhibited the best results as compared to only organic solvents.^43,92 Accordingly, in this study, we initially selected a methanol/water (CH₃OH/H₂O) solvent system in a 3 : 1 ratio (v/v) and proceeded to optimize the organic component while maintaining the same ratio. As anticipated, the CH₃OH/H₂O solvent system outperformed all other solvent systems tested during optimization providing the highest yield of 45% of the desired product 3ab (Scheme 1) and hence was selected for subsequent optimization studies, Fig. 8b. Furthermore, the reaction was evaluated with various bases (Cs₂CO₃, K₂CO₃, KHCO₃, and Na₂CO₃) and 3 equivalents of K₂CO₃ as a base was identified as optimal for this reaction with a 58% yield of the desired product, Fig. 8c. Irradiation time in photocatalytic reactions also critically influences the yield and selectivity by controlling the extent of catalyst excitation, reactive intermediate generation, and the balance between productive and competing pathways. Therefore, the reaction was monitored at varying irradiation times (Fig. 8d), and a 70 min duration was identified as optimal for this photocatalytic SMCC reaction, affording the desired product in 84% yield. The catalyst amount was also optimized taking the already optimized reaction parameters. The study revealed that 20 mg of the PdO/RuO₂–ZnO catalyst afforded the highest yield (97%) of the desired biaryl product, 3ab (Scheme 1). Thus, through systematic optimization of catalysts, solvent system, base, irradiation time, and catalyst loading, the photocatalytic SMCC reaction using the PdO/RuO₂–ZnO catalyst was transformed from a low-yielding process into an efficient, high yielding reaction, achieving up to 97% yield of the desired biaryl product under green photocatalytic conditions. The optimization parameters were viz. methanol/water solvent system in a 3 : 1 (v/v) ratio, 3 equivalents of K₂CO₃ as a base, 70 min of UV-light irradiation time (450 W medium pressure mercury lamp, λ =254 nm), and 20 mg of the PdO/RuO₂–ZnO catalyst affording the maximum product yield of 97% (product 3ab, Scheme 1).

Fig. 8 Graphical illustration showing the optimization studies of (a) % yield with different catalyst systems, (b) % yield across different solvent systems, (c) % yield with different bases (d) % yield over different reaction times, (e) % yield at varying catalyst loadings, and (f) optimized conditions and schematic diagram of the reaction set up used.

Scheme 1 Substrate scope for the SMCC reaction of various aryl halides with different boronic acids. The yields are isolated % yield after column chromatography.

With these optimized parameters established, the research work was expanded to a broad substrate scope study, Scheme 1. Excellent to good yields (78–98%) were consistently obtained across a wide range of substrates. These results confirmed the efficiency of the photocatalyst, highlighting its potential for sustainable and light-driven synthetic transformations. The isolated products were characterized using NMR spectroscopy, and the corresponding NMR spectra and all other details of characterization of the organic molecules are provided in the SI, section 23.

5.1 Controlled experiments

Several controlled experiments were carried out for the photocatalytic SMCC reaction using the PdO/RuO₂–ZnO catalyst in MeOH/H₂O under UV light irradiation and the same has been discussed in this section. Initially, to assess the influence of light irradiation, the reaction was carried out under completely dark conditions in the absence of any light irradiation keeping all other optimized parameters unchanged. The absence of product formation suggested the inevitable role of light irradiation in driving this photocatalytic SMCC process. Further, a blank reaction performed in the absence of the PdO/RuO₂–ZnO catalyst under otherwise identical conditions also showed no detectable formation of the cross-coupled product. This clearly established that neither the base nor UV irradiation alone was sufficient to drive the transformation, thereby confirming that catalytic activity arises exclusively from the PdO/RuO₂–ZnO photocatalyst in the presence of UV irradiation. The reaction was also performed without the base, keeping all other optimized parameters the same, but again no product formation was observed. The photocatalytic SMCC reaction was also carried out under visible-light irradiation keeping all other optimized parameters the same. However, no significant product formation was observed under visible light, Fig. S22. This observation was consistent with the measured band gap of the catalyst (3.78 eV), which allows absorption only in the UV region, highlighting the crucial role of UV light in driving the photocatalytic process. Subsequently, to evaluate the individual contributions of each component of the PdO/RuO₂–ZnO catalyst, controlled reactions were carried out using the PdO–ZnO, RuO₂–ZnO, and PdO/RuO₂ catalysts as discussed in the catalytic study section, Fig. 8a. As shown in Fig. 8a, these binary systems showed only very low to moderate yields under the test reaction conditions. Even under the optimized conditions developed for the trimetallic PdO/RuO₂–ZnO catalyst, their activity did not improve significantly. This outcome clearly suggested that the superior photocatalytic efficiency was not achievable with the binary counterparts alone, and instead arose from the cooperative interplay of PdO, RuO₂, and ZnO in the trimetallic PdO/RuO₂–ZnO system. In the trimetallic composite, plausibly the close integration and synergistic interaction of PdO, RuO₂, and ZnO facilitated more efficient charge carrier separation and suppressed the electron–hole recombination rate, thereby ensured the availability of a higher density of reactive charge carriers to drive the cross-coupling process. The DFT study also suggested the superior stability of the ternary PdO/RuO₂–ZnO catalyst with an adsorption energy of −203.27 kcal mol⁻¹ as compared to the binary catalysts, Table S4.

This catalytic finding also aligned well with the PL analysis presented in the results and discussion section. As mentioned in the results and discussion section, the band gaps of the three materials, namely PdO/RuO₂–ZnO, PdO–ZnO, and RuO₂–ZnO, were well-suited for efficient UV-light absorption. However, the ternary PdO/RuO₂–ZnO catalyst exhibited the lowest PL intensity among all other synthesized materials, reflecting its superior ability to suppress the electron–hole recombination rate to the lowest, Fig. S17. Although the PL intensity of the RuO₂–ZnO catalyst was nearly comparable to the ternary catalyst, the non-availability of the actual Pd active sites might have resulted in reduced activity of this catalyst. Pd had this title of the actual active site in the SMCC reaction because the entire catalytic cycle was fundamentally based on a Pd-centred orchestration. The key mechanistic step or the rate determining step of SMCC, the oxidative addition step, occurs exclusively on Pd. Therefore, the inactivity of the RuO₂–ZnO catalyst towards the SMCC reaction was reasonable. However, interestingly this RuO₂–ZnO heterojunction appeared to be a suitable hybrid material that can control the electron–hole recombination rate very efficiently as suggested by the PL analysis. In contrast to this RuO₂–ZnO heterojunction catalyst, the PdO–ZnO catalyst showed a higher PL intensity, indicating a higher rate of charge-carrier recombination which plausibly was the reason behind its low catalytic activity despite the presence of Pd centres, Fig. S17. The unsupported PdO/RuO₂ catalyst, with a band gap of 2.26 eV and notably higher PL intensity, was also not suitable for the UV-irradiated photocatalytic SMCC process, displaying inferior activity. Overall, the coexistence of the RuO₂–ZnO heterojunction to control the electron–hole recombination rate along with the actual Pd active sites proved to be essential for achieving enhanced performance in this photocatalytic SMCC process and thus vouched for the higher activity of the PdO/RuO₂–ZnO catalyst in comparison to the other synthesized catalysts. The high activity achieved with the minimally loaded Pd in the PdO/RuO₂–ZnO catalyst eliminated the need to further optimize the amount of Pd during the synthesis of the catalyst. The current ultra trace Pd content already represented a practical lower limit, beyond which reduction of Pd content was not meaningful. Conversely, increasing the Pd amount would unnecessarily escalate the cost of the catalyst system and would compromise the sustainability advantage of the system.

6. Mechanistic investigations

To gain insight into the photocatalytic pathway and to clarify the roles of photogenerated charge carriers in the SMCC reaction involving the PdO/RuO₂–ZnO catalyst, a series of experiments were performed. To outline the roles of different charge carriers and possible radical intermediates in the catalytic cycle, silver nitrate (AgNO₃) was employed as an electron (e⁻) scavenger, KI as a hole (h⁺) scavenger, and TEMPO and DMPO as radical scavengers.^93–95 The addition of AgNO₃ to the reaction mixture of this photocatalytic SMCC reaction almost completely suppressed the reaction, confirming that conduction-band (CB) electrons were absolutely necessary for the reaction to occur. These electrons probably generated catalytically active Pd(0) species in the catalytic cycle. Similarly, the introduction of potassium iodide (KI), a hole scavenger, to the reaction mixture led to a moderate inhibition of the reaction, suggesting that the holes present in the valence-band (VB) also contributed to the reaction mechanism but were not the sole driving force. Notably, when both AgNO₃ and KI were added to the reaction mixture, the reaction was entirely quenched implying the complementary requirement of both charge carriers (e⁻ and h⁺) to complete a successful catalytic cycle. In contrast, the introduction of TEMPO or DMPO as radical trapping agents, each added individually to the reaction mixture had no effect on the reaction outcome. It thereby suggested the non-involvement of any radical intermediates in the catalytic cycle. Collectively, these observations validated a plausible mechanism wherein electrons were essential in activation of Pd for an efficient oxidative addition step and the holes promoted boronic acid activation, together enabling a catalytic cycle as shown in Scheme 2. On the basis of these findings, a mechanistic pathway is proposed and designed as outlined below.

Scheme 2 Plausible reaction mechanism of the photocatalytic SMCC reaction in the presence of the PdO/RuO₂–ZnO catalyst.

At first, upon irradiation with UV light, the PdO/RuO₂–ZnO photocatalyst got activated, leading to the generation of photogenerated electrons (e⁻) and holes (h⁺) at the conduction band (CB) and valence band (VB) of the catalyst, respectively (Scheme 2). The photogenerated electrons were then plausibly transferred to the Pd(II) centres present in the PdO phase of the aforementioned catalyst. This electron transfer process reduced Pd(II) to Pd(0), generating the actual catalytically active species required in an SMCC reaction. This in situ reduction step is strongly feasible as supported by the XPS analysis. XPS analysis had revealed a negative shift in the Pd binding energy in the ternary system as compared to the pristine PdO, indicating increased electron density around Pd due to interfacial electronic interaction. This observation suggested that the in situ reduction to Pd(0) is very much favourable in the ternary catalyst. In addition, UV-DRS analysis also suggested the formation of defect-induced intermediate energy states in the ternary catalyst, which enhanced interfacial charge-transfer transitions and improved charge separation efficiency. Notably, palladium species are well known to act as effective electron sinks due to their favourable redox potential and high electron affinity.^96,97 Thus, in this PdO/RuO₂–ZnO system, the PdO phase could facilitate rapid trapping of photogenerated electrons, which not only promoted the in situ formation of Pd(0) active sites but also suppressed the recombination of electron–hole pairs. The Ru phase within the catalyst containing both Ru(0) and Ru(IV) might have also provided a fast electron-transfer pathway by shuttling its oxidation states, enabling smooth reduction of Pd(II) to Pd(0). The formation of these Pd(0) centres facilitated the cleavage of the carbon–halogen bond of the aryl halide via oxidative addition as shown in step 1, Scheme 2. In the meantime, the photogenerated holes in the VB were also directed towards the RuO₂ part of the PdO/RuO₂–ZnO catalyst making it electrophilic.⁹⁸ RuO₂, owing to its delocalized 4d band and multivalent Ru centres (0 and +4 oxidation states) in the PdO/RuO₂–ZnO catalyst, could effectively stabilize both photogenerated electrons and holes. The presence of Ru species in both Ru(0) and Ru(IV) oxidation states in the PdO/RuO₂–ZnO catalyst increased the ability of the Ru centres of the catalyst to shuttle its oxidation states. This shuttling of oxidation states, in turn, might have also helped the acceptance of holes generated in the VB during the photoexcitation process by the Ru phase of the catalyst, and influenced the activation of boronic acid as shown in step 4. The electrophilicity obtained from the holes by the Ru phase of the catalyst promoted the cleavage of the C–B bond in the activated boronic acid species [PhB(OH)₃]⁻, step 4, Scheme 2. Jiao et al. have previously reported this type of interaction leading to bond cleavage in boronic acids in an Pd@NiO₈₀/SiC catalysed photocatalytic SMCC reaction.⁹⁸ In the meantime the presence of the RuO₂–ZnO heterojunction effectively maintained a minimal rate of recombination of electron–holes and thereby helped the overall catalytic cycle. Thereafter, transmetalation occurred between the Pd(II)–aryl intermediate generated in step 1 and the activated boronic species generated in step 4, furnishing the intermediate species A, Scheme 2. Lastly, reductive elimination from species A resulted in the coupling of both the coordinated aryl groups producing the desired biaryl product, step 5, Scheme 2. DFT projected density-of-states (PDOS) analysis of the PdO/RuO₂–ZnO catalyst revealed significant Pd 4d–Zn 3d orbital overlapping, indicating strong d–d hybridization that improved the interfacial charge transfer making shuttling of Pd redox states easier. The Pd 3d–Zn 3d orbital overlapping might have created a shared electronic channel at the interface, allowing photogenerated electrons to flow more freely toward Pd(II) centres. This d–d hybridization could lower the electron-transfer resistance and increased the electron density around Pd sites. As a result, Pd(II) species anchored on this electronically coupled interface might receive electrons more efficiently, making their reduction to Pd(0) easier as required in the catalytic cycle to afford an efficient oxidative addition step. This electronic synergy accelerated oxidative addition, facilitated electron-mediated Pd regeneration, and enhanced the rate of the photocatalytic SMCC process.

This mechanistic pathway again decisively accounted for the superior performance of the trimetallic PdO/RuO₂–ZnO catalyst over its binary counterparts. Only in this trimetallic catalytic system could the charge transfer towards the PdO and RuO₂ domain of the catalyst and dual-site activation be effectively justified. These features were inherently lacking in the bimetallic counterparts. The binary systems, namely PdO–ZnO and RuO₂–ZnO, could not proceed with an efficient photocatalytic pathway for the SMCC reaction, as the photogenerated charge carriers (e⁻/h⁺) were either not utilized properly or recombined rapidly, leading to only moderate yields. Similarly, the PdO/RuO₂ catalyst without ZnO as the semiconductor support lacked an efficient semiconducting platform for charge generation (e⁻/h⁺) and separation under UV light irradiation. This might have restricted the simultaneous utilization of electrons and holes in the catalytic cycle of the SMCC reaction. Thus, this mechanism provided a clear rationale for the lower catalytic performances of the binary counterparts as compared to the trimetallic PdO/RuO₂–ZnO catalyst.

7. Kinetic study

To gain insight into the reaction dynamics, the photocatalytic SMCC process between phenyl bromide (PhBr) and 4-(diphenylamino)phenylboronic acid was closely monitored using UV-visible (UV-vis) spectroscopy. The advancement of the reaction was tracked by collecting small aliquots of the reaction mixture at each 10 min of time interval up to 70 min, and then recording their UV vis spectra. Initially i.e. at time (t) = 0 min, the UV-vis absorption spectrum was completely dominated by the reactants namely PhBr and 4-(diphenylamino)phenylboronic acid, indicating the absence of the desired biaryl product, 3as, Scheme 1. As the reaction proceeded, a gradual change in the absorption profile was observed, reflecting the formation of the biaryl product 3as. The UV-vis spectra of the aliquots measured at a time interval of 10 min are provided in Fig. 9a. Some clear shifts and changes in the absorption profile were observed in these spectra. The rate of the reaction was derived from the absorbance (A) profile of the reaction mixture over time. The plot of ln[A] vs. time (min) showed a linear trend, confirming that the reaction followed first-order kinetics. The rate constant was found to be 2.54 × 10⁻² min⁻¹, Fig. 9b.

Fig. 9 a) UV-vis absorption spectra of the aliquots recorded at various time intervals, and b) first-order rate kinetics plot showing the linear dependency.

To understand the influence of different substituents on the aryl halide substrate molecules and subsequently on the SMCC reaction, a Hammett analysis was also performed. For this purpose, the formation of products from aryl bromides bearing electron-withdrawing groups (EWG) (viz. p-CHO and p-CN) and electron-donating groups (EDG) (viz. p-OMe and p-Me) with 4-(diphenylamino)phenylboronic acid was monitored using UV-vis spectroscopy, Fig. 10. In all the cases, the reaction followed first-order kinetics, as evidenced by the linear relationship observed in the ln[A] vs. time plots, Fig. S23. The reaction rates were determined for each substituted aryl halide and a Hammett plot of relative rate [log(k_X/k_H)] vs. the Hammett constant (σ) was plotted as provided in SI, Fig. S24. The calculated logarithmic values of the relative rate [log(k_X/k_H)] for the aforementioned aryl bromides with EWG and EDG along with their standard σ-value are provided in Table S6.⁹⁹ The plot exhibited a positive slope of ρ value + 0.297, signifying that aryl bromides with electron-withdrawing substituents at the para position accelerated the reaction than those with electron-donating substituents.¹⁰⁰ This positive slope also suggested that the rate determining step favoured substrates with electron-withdrawing groups (EWG). This observation again aligned with and supported the mechanism study stating the oxidative addition of the aryl bromide to the Pd(0) centre as the rate-determining step of the reaction. Electron-deficient aryl bromides (i.e. aryl bromides with EWG) facilitated this step by stabilizing the transition state, whereas electron-rich aryl bromides (i.e. aryl bromides with EDG) slowed down the reaction. These results thus provide a clear insight into the influence of substituent effects on the kinetics of a photocatalytic SMCC reaction and highlighted the importance of electronic factors in determining reaction efficiency.

Fig. 10 UV-vis spectra of the reaction mixture recorded at different intervals for (a and b) aryl bromides bearing electron-withdrawing substituents [a) p-CHO, b) p-CN] and (c and d) aryl bromides with electron-donating substituents [a) p-Me, b) p-OMe] in the reaction with 4-(diphenylamino)phenylboronic acid.

8. Expanding the utility of the SMCC products: conversion to alcohols

To further establish the synthetic relevance of the present photocatalytic SMCC products, we investigated their subsequent transformation into a new range of valuable alcohol derivatives. In particular, representative biaryl products (3be to 3bj, Scheme 1) bearing carbonyl functional group were subjected to NaBH₄-mediated reduction under mild conditions as shown in Scheme 3. The reactions proceeded very efficiently to result in complete conversion of the SMCC products into their corresponding secondary alcohols. Interestingly, this simple one-step transformation furnished versatile new alcohol derivatives that are important in many fields. The alcohol products obtained through this transformation are highly valuable synthetic intermediates in many important processes. Such alcohol motifs are extensively utilized across organic synthesis as key functional handles for diverse operations—including esterification, etherification, oxidation, protection–deprotection sequences, and incorporation into more complex molecular frameworks.¹⁰¹ They are very crucial in the preparation of pharmaceuticals, fine chemicals, agrochemicals, and assorted bioactive compounds, where alcohol functionalities often serve as the essential entry points for further derivatization.¹⁰¹ The ease with which the SMCC products could be converted into these important intermediates significantly enhanced the practical reach of the developed protocol. Thus, the successful reduction of the biaryl carbonyl groups highlighted two important aspects. Firstly, it demonstrated the chemical robustness of the SMCC products, as they tolerated further transformations without degradation of the biaryl framework. Secondly, it established a straightforward route to access structurally diverse new alcohol derivatives under SMCC reaction conditions, well known for their broad functional group tolerance.

Scheme 3 Synthetic expansion of SMCC products: transformation into new alcohol derivatives.

9. Green chemistry metrics

The sustainability of the developed photocatalytic SMCC protocol was also assessed through various green chemistry metrics. Green chemistry metrics were typically employed as quantitative measures to evaluate the sustainability, efficiency, and environmental footprint of chemical products and processes.¹⁰² Catalytic processes for the production of fine chemicals are particularly important in modern chemistry due to their high selectivity, lower energy requirements, and reduced waste generation, even though they require complex synthesis procedures and high purity standards. Consequently, a thorough evaluation of green metrics in catalytic reactions has become essential for guiding the design and implementation of more sustainable chemical processes. Therefore, to evaluate these aspects in the present photocatalytic SMCC reaction, metrics such as reaction yield, atom economy (AE), E-factor (E), process mass intensity (PMI), stoichiometric factor (SF), and reaction mass efficiency (RME) were considered and calculated using the formulae as provided in SI, eqn (S1)–(S6).¹⁰² Among the series of 41 substrate molecules investigated, as shown in Scheme 1, the best performing substrate (product 3as, Scheme 1) in terms of green chemistry metrices exhibited a isolated yield of 92%, reaction yield (ε) of 0.92, with an atom economy (AE) of 0.59, an E-factor (E) of 0.74, and a process mass intensity (PMI) of 2.85. The stoichiometric factor (SF) was also calculated to be 1.12 with the ideal value of SF as 1, and the reaction mass efficiency (RME) was calculated to be 0.35. To the best of our knowledge, these values were very good for this type of C–C coupling protocol. Above all these, the energy requirement in this protocol was also greener, as the reaction proceeded under light-driven conditions. Furthermore, this catalytic reaction was found to proceed through a non-free radical pathway, thereby avoiding the generation of unstable and hazardous radical intermediates. This was highly advantageous from the green chemistry viewpoint, as radical-based transformations typically require harsh initiators, high energy inputs, and are often accompanied by undesirable side reactions and byproduct formation. The absence of a free radical mechanism in the present system thus enhanced the reaction selectivity, improved the atom economy, and minimized waste generation making it sustainable for large-scale applications.

10. Comparison with previously reported Pd based heterogenous photocatalysts for the SMCC reaction

In order to compare the relative performance of the PdO/RuO₂–ZnO catalyst in the photocatalytic SMCC reaction, a comparative analysis with previously reported Pd based photocatalytic systems was carried out. It was found that compared to previously reported photocatalytic systems, the PdO/RuO₂–ZnO catalyst offered several unique features and significant improvements. A survey of literature reports (as shown in Table 1) clearly indicated that although only a few photocatalysts had been designed for the photocatalytic SMCC reaction, the majority of them were found to be effective only for activated substrates such as aryl iodides, while displaying poor or negligible reactivity with the aryl bromide substrate molecules. Even when successful with aryl bromides, most photocatalysts demanded stringent reaction conditions such as use of additives like CTAB, stronger bases like NaOH, higher reaction temperature than room temperature, the necessity of inert reaction atmospheres (like Ar or N₂), further complicating their applicability. Additionally, the reported reaction times were typically very prolonged (up to 24 h in some cases) and even under such conditions the substrate scope study remained quite limited with low yield in many cases.

Table 1 Comparative performance of reported photocatalysts and the present PdO/RuO₂–ZnO system in the photocatalytic Suzuki–Miyaura cross-coupling reaction

Sl. No	Catalyst system	Aryl halide used	Reaction conditions	Time	Yield (%)	No. of substrates	Ref.
1	Pd/SiO2/3%Au/ZrO2	Aryl iodides	1,4-Dioxane, Cs2CO3,40 °C, Ar, visible light	12 h	Up to 63.1	7	42
2	PdCu/UTCN	Aryl bromide	EtOH : H2O, K2CO3, 300 W xenon lamp, Ar	12 h	14.7	1	38
3	g-CN(G)–AgPd	Aryl bromides	K2CO3 (5 eq.), white LED, EtOH/H2O	15 min– 1 h	80–97	6	103
4	Pd/CN–xCu	Aryl iodide	K3PO4, EtOH/H2O, N2, 420 nm light	50 min	98	1	104
5	Pd3P/CdS	Aryl bromides	K2CO3 (5 eq.), EtOH/H2O (3 : 2), 25 W white LED lamp	12 h	72–84	4	105
5	Pd/Ce0.7Fe0.3O2−δ	Aryl bromides	NaOH, visible light	6 h	10.4–99.8	9	106
6	Pd@BD-g-C3N4-35	Aryl bromides	K2CO3, H2O, UV-visible	4 h–5 h	41–77	6	107
7	Pd-1/C3N5	Aryl bromides	K2CO3, ethanol, visible light	Up to 7 h	79–98	13	108
8	Pd@TCN-4	Aryl bromides	K2CO3, Ar atmosphere, LED, EtOH	6 h	80–84	2	40
9	NiFe2O4@TiO2@PDA–Pd	Aryl bromides	K2CO3, EtOH/H2O under sunlight	2 h	76–80	2	109
10	Fe3O4@SiO2@TiO2@Schiffbase@Pd(0)	Aryl bromides	K2CO3, PEG-400, LED	Up to 24 h	70–97	8	110
11	Pd@B–BO3	Aryl bromide	K2CO3, DMF/H2O, white LED lamp	Up to 16 h	76–96	3	41
12	Au–Pd–ZrO2	Aryl bromides	NaOH, CTAB, H2O, light energy 0.5 W cm^−2	3 h	25–87	5	37
13	PdO/RuO 2 –ZnO	Aryl bromides	K2CO3, CH3OH/H2O, UV-vis light	70 min	78–98	41	This work

---

In sharp contrast, the PdO/RuO₂–ZnO catalyst developed in this research work exhibited an excellent balance of efficiency, greener prospects, low cost, and broad applicability. The reaction set up was entirely additive-free, requiring no surfactants, special ligands, or inert atmospheres, thereby simplifying the reaction conditions and reducing both cost and waste generation. The optimized reaction time was very short (70 min), considerably shorter than most of the reported heterogeneous photocatalytic systems. Importantly, the catalyst enabled cross-coupling of a remarkably broad range of aryl bromide substrates (41 examples), representing the widest substrate tolerance reported to date for a photocatalyst-based SMCC reaction. The high yields achieved across this wide substrate scope further highlighted the robustness of the system. Notably, the excellent catalytic performance of the PdO/RuO₂–ZnO catalyst was achieved with Pd present only at ultra-trace levels. The low Pd incorporation not only reduced the material cost but also enhanced the overall sustainability profile of the catalyst, reinforcing its suitability as a practical alternative to conventional Pd-rich catalytic systems. This minimal Pd usage aligned with the broader objective of creating a cost-effective and resource-efficient catalytic platform without compromising the performance. Thus with a very low Pd loading, additive-free milder reaction conditions, shorter reaction time, high % yield, and an unprecedentedly broad substrate scope, this research work clearly demonstrated that the PdO/RuO₂–ZnO photocatalyst surpassed the activity of most of the previously reported Pd based heterogeneous photocatalysts.

11. Catalyst reusability and heterogeneity study

To further explore the stability and reusability of the catalyst across multiple numbers of cycles, a recyclability study was performed. For that, after the initial run of the reaction, the catalyst was segregated from the reaction mixture by a simple filtration method. In order to remove any unwanted impurities, the recovered catalyst was thoroughly washed with suitable organic solvents, hot distilled water, and ethanol, and then dried in a vacuum oven before reusing it in each subsequent cycle. The recyclability investigation was carried out under the already optimized reaction conditions, employing p-bromobenzaldehyde as the test substrate. The recyclability study was extended up to the 8th catalytic cycle; however, a noticeable decline in photocatalytic activity was observed from the 7th cycle onward, as clearly illustrated in the bar diagram, Fig. S25. No significant decrease in photocatalytic efficiency was detected up to the 6th cycle, indicating that the PdO/RuO₂–ZnO catalyst exhibited good stability and reusability for the SMCC reaction, with six effective recyclable cycles. The recycled catalyst after the 6th catalytic cycle was characterized with Raman, SEM, and TEM analysis and the results are provided in SI, Fig. S26 and S27. Further, to assess the true heterogeneous nature of the PdO/RuO₂–ZnO photocatalyst, a hot filtration test was also performed by removing the catalyst from the reaction mixture after 30 min of reaction time under the optimized reaction conditions. The filtrate was then allowed to react further under the identical reaction conditions. However, no considerable increase in the product yield was observed, thereby ruling out the possibility of catalysis by leached metal species, Fig. S28. Thus, this observation confirmed that the reaction predominantly proceeded via a heterogeneous pathway and the catalyst was a truly heterogeneous one.

12. Conclusion

In summary, the PdO/RuO₂–ZnO heterostructured catalyst demonstrated excellent photoactivity for the C–C coupling reaction by delivering excellent yields, exceptional selectivity, and a broad substrate scope under mild and greener reaction conditions within a short reaction time of 70 min. Notably, the incorporation of only an ultra-trace amount of Pd on the RuO₂–ZnO heterostructure was sufficient enough to induce a pronounced improvement in the catalytic performance. The presence of RuO₂ as a redox mediator helped the charge carrier transportation process, assisting the reduction of Pd(II) to Pd(0) as well as facilitating the cleavage of the C–B bond of aryl boronic acids. ZnO retained its original role of generating the photogenerated charge carriers essential for the photocatalytic process. The synergistic interactions between the metal oxide components in the ternary catalyst significantly amplified the overall activity of the catalyst, enabling rapid formation of the desired biaryl products under UV-light irradiation. The catalyst also exhibited highly efficient and controlled electron–hole separation, minimizing recombination losses and thus providing a key factor behind its superior photocatalytic behaviour. DFT analysis reinforced these observations by revealing a clear orbital overlap between Pd 4d and Zn 3d states, indicating strong d–d hybridization that supports smooth flow of charge carriers throughout the catalytic cycle. Apart from these, the catalyst's exceptionally low Pd loading, ability to efficiently harness light energy, non-free radical pathway, shorter reaction time, and the ability to maintain broad substrate scope compatibility highlighted its potential as a sustainable and cost-effective alternative to traditional thermally driven SMCC protocols using Pd rich catalytic systems. The reaction had occurred in a true heterogeneous way and the catalyst was highly recyclable up to the 6th catalytic cycle. Overall, this study established PdO/RuO₂–ZnO as a robust, efficient, and sustainable photocatalyst for C–C bond-forming transformation with proper utilization of the RuO₂–ZnO heterojunction, offering a promising platform for future developments in green heterogeneous photocatalysis.

Author contributions

Tonmoy J. Bora: conceptualization, methodology, validation, formal analysis, investigation, data curation, visualization, writing – original draft. Donguk Kim: investigation, data curation. Young-Bin Park: resources, validation. Gaurisankar Phukan: methodology, software, data curation. Shamim Islam: formal analysis, data curation. Arpita Devi: validation. Kusum K. Bania: conceptualization, resources, writing – original draft, supervision, project administration, funding acquisition.

Conflicts of interest

There are no conflicts to declare.

Data availability

Data will be made available on request.

Supplementary information (SI): materials, physical measurements, procedure for dark-phase and room-light SMCC reaction, characterization of synthesized materials, computational details, recyclability and heterogeneity study, characterization of the recycled catalyst, and ¹H and ¹³C NMR spectra of synthesized organic compounds. See DOI: https://doi.org/10.1039/d5cy01590a

Acknowledgements

KKB thanks ANRF for research grants CRG/2023/000957, EEQ/2023/000350 and TAR/2023/000221. KKB also thanks Department of Science and Technology (DST), Govt. of India for the PURSE research grant with file No. SR/PURSE/2024/240. TJB and AD acknowledge the DST, Govt. of India for the INSPIRE Fellowship (No. DST/INSPIRE Fellowship/2021/IF210170 and No. DST/INSPIRE Fellowship/2020/IF200100). The authors also thank SAIC Tezpur University for analytical facilities.

References

1. Y. Wang, Y. Liu, J. Gan, L. Hong, J. Shen, Z. Zhang, H. Zheng and X. Wang, ACS Sustainable Chem. Eng., 2024, 12, 14505–14513.
2. S. B. Beil, S. Bonnet, C. Casadevall, R. J. Detz, F. Eisenreich, S. D. Glover, C. Kerzig, L. Næsborg, S. Pullen, G. Storch and N. Wei, JACS Au, 2024, 4, 2746–2766.
3. A. Quintavalla, D. Carboni and M. Lombardo, ChemCatChem, 2024, 16, e202301225.
4. F. Zhao, I. Ahmad, H. Bayahia, S. AlFaify, K. M. Alanezi, M. Q. Alfaifi, M. D. Ali, Y. Y. Ghadi, I. Ali and T. L. Tamang, Int. J. Hydrogen Energy, 2024, 80, 659–685.
5. I. Ahmad, M. Q. Alfaifi, Y. A. Alassmy, M. M. Abduljawad, S. A. Alshuhri, B. N. Alotaibi, A. Aslam, T. L. Tamang, F. Farahim and T. Ali, Int. J. Hydrogen Energy, 2025, 103, 456–466.
6. S. Karim, H. M. Saw and Y. P. Ting, Green Chem., 2024, 26, 4112–4126.
7. F. Xu and B. Weng, J. Mater. Chem. A, 2023, 11, 4473–4486.
8. C. Fu, Z. Wan, X. Yang, J. Zhang and Z. Zhang, J. Mater. Chem. A, 2024, 12, 28618–28657.
9. J. Zhang, X. Wang, X. Wang and C. Li, Acc. Chem. Res., 2025, 58, 787–798.
10. J. Liu, H. Wang, H. Wu, Y. Yang, C. Wang, Q. Wang, B. Jia and J. Zheng, J. Mater. Chem. A, 2024, 12, 20838–20867.
11. W. L. de Almeida, N. S. Ferreira, F. S. Rodembusch and V. C. de Sousa, Mater. Chem. Phys., 2021, 258, 123926.
12. W. Jiang, H. Loh, B. Q. L. Low, H. Zhu, J. Low, J. Z. X. Heng, K. Y. Tang, Z. Li, X. J. Loh, E. Ye and Y. Xiong, Appl. Catal., B, 2023, 321, 122079.
13. K. J. Archana, A. C. Preetha and K. Balasubramanian, Opt. Mater., 2022, 127, 112245.
14. I. Ahmad, S. Shukrullah, M. Y. Naz, E. Ahmed, M. Ahmad, S. U. Rehman, H. N. Bhatti and A. Ghaffar, Int. J. Hydrogen Energy, 2021, 46, 30320–30333.
15. I. Ahmad, S. Shukrullah, M. Y. Naz, M. Ahmad, E. Ahmed, M. S. Akhtar, S. U. Rehman and M. M. Makhlouf, Int. J. Hydrogen Energy, 2021, 46, 30019–30030.
16. I. Ahmad, S. Shukrullah, M. Y. Naz and H. N. Bhatti, Int. J. Hydrogen Energy, 2023, 48, 29485–29496.
17. C. Sun, X. Wu, H. Meng, X. Xu, J. Xu and X. Zhang, J. Phys. Chem. Solids, 2014, 75, 726–731.
18. S. J. Armaković, S. Armaković, A. Bilić and M. M. Savanović, Catalysts, 2025, 15, 793.
19. I. Ahmad, R. Bousbih, A. Mahal, W. Q. Khan, M. Aljohani, M. A. Amin, N. N. Jafar, M. S. Jabir, H. Majdi, A. S. Alshomrany and M. Shaban, Mater. Sci. Semicond. Process., 2024, 180, 108578.
20. Q. Yang, X. Li, L. Chen, X. Han, F. R. Wang and J. Tang, Angew. Chem., 2023, 135, e202307907.
21. Y. Sun, W. Zhang, Q. Li, H. Liu and X. Wang, Adv. Sens. Energy Mater., 2023, 2, 100069.
22. S. Liu, Y. Wang, S. Wang, M. You, S. Hong, T. S. Wu, Y. L. Soo, Z. Zhao, G. Jiang, J. Qiu and B. Wang, ACS Sustainable Chem. Eng., 2019, 7, 6813–6820.
23. A. Talebian-Kiakalaieh, M. Guo, E. M. Hashem, B. Xia, Y. Jiang, C. Chuah, Y. Tang, P. Kwong, J. Ran and S. Z. Qiao, Adv. Energy Mater., 2023, 13, 2301594.
24. Z. Li, Y. Yang, S. Wang, L. Gu and S. Shao, ACS Appl. Mater. Interfaces, 2021, 13, 46608–46619.
25. S. Wang, J. J. Zhang, M. Y. Zong, J. Xu, D. H. Wang and X. H. Bu, ACS Catal., 2022, 12, 7960–7974.
26. G. Chacktas, B. Pfund, T. Kerackian, P. Yaltseva, M. Villeneuve, D. Durand, N. Fabre, C. Fiorini-Debuisschert, J. C. Cintrat, O. S. Wenger and E. Romero, ACS Catal., 2025, 15, 13938–13947.
27. M. Solakidou, A. Zindrou, S. Smykała and Y. Deligiannakis, Ind. Eng. Chem. Res., 2024, 63, 5773–5786.
28. C. P. Chen, O. Abubakar, X. Zhang and T. W. Hamann, Inorg. Chem., 2025, 64, 13599–13603.
29. N. Deka, T. E. Jones, L. J. Falling, L. E. Sandoval-Diaz, T. Lunkenbein, J. J. Velasco-Velez, T. S. Chan, C. H. Chuang, A. Knop-Gericke and R. V. Mom, ACS Catal., 2023, 13, 7488–7498.
30. M. Sohail, W. Lv and Z. Mei, ACS Sustainable Chem. Eng., 2023, 11, 17564–17594.
31. J. Yu, Q. He, G. Yang, W. Zhou, Z. Shao and M. Ni, ACS Catal., 2019, 9, 9973–10011.
32. C. A. Wang, H. R. Cui, Y. R. Wang, F. Zhao, Y. F. Han and F. S. Liu, ACS Appl. Polym. Mater., 2025, 7, 8199–8212.
33. Y. Jiang, Q. Liu, B. Xie and J. Zhang, Ind. Eng. Chem. Res., 2024, 63, 10532–10543.
34. J. H. Ko, Y. J. Choi, Y. J. Jin, C. H. Park, Y. J. Jang, Y. P. Son and C. M. Chung, ACS Appl. Nano Mater., 2025, 8, 15472–15484.
35. M. E. Usteri, G. Giannakakis, A. Bugaev, J. Pérez-Ramírez and S. Mitchell, ACS Catal., 2024, 14, 12635–12646.
36. Y. Shu, D. X. Xiang, M. Y. Li, X. R. Xie, Q. Y. Chen, Y. T. Lin, Z. Y. Yang, Y. J. Ouyang and G. Chen, Appl. Surf. Sci., 2025, 690, 162637.
37. Q. Xiao, S. Sarina, E. Jaatinen, J. Jia, D. P. Arnold, H. Liu and H. Zhu, Green Chem., 2014, 16, 4272–4285.
38. X. Jia, J. Zhao, Y. Lv, X. Fu, Y. Jian, W. Zhang, Y. Wang, H. Sun, X. Wang, J. Long and P. Yang, Appl. Catal., B, 2022, 300, 120756.
39. H. Q. Yang, Q. Q. Chen, F. Liu, R. Shi and Y. Chen, Chin. Chem. Lett., 2021, 32, 676–680.
40. L. Sun, G. Wang, S. Bi, H. Li, H. Zhan and W. Liu, Appl. Organomet. Chem., 2024, 38, e7353.
41. Z. J. Wang, S. Ghasimi, K. Landfester and K. A. Zhang, Chem. Mater., 2015, 27, 1921–1924.
42. C. Han, D. E. Gómez, Q. Xiao and J. Xu, J. Catal., 2021, 397, 205–211.
43. T. J. Bora, N. Hazarika, N. K. Gour, S. Lee, Y. B. Park, S. Biswas, A. Devi and K. K. Bania, ACS Appl. Nano Mater., 2023, 6, 17972–17985.
44. F. F. Alharbi, S. Manzoor, T. Munawar, M. U. Nisa, A. G. Abid, F. Iqbal, S. Aman, M. F. Ehsan, M. Najam-Ul-Haq and M. N. Ashiq, Surf. Interfaces, 2022, 31, 101991.
45. C. F. Holder and R. E. Schaak, ACS Nano, 2019, 13, 7359–7365.
46. M. A. Behrens, A. Franzén, S. Carlert, U. Skantze, L. Lindfors and U. Olsson, Soft Matter, 2025, 21, 2349–2354.
47. M. Anandan, S. Dinesh, B. Christopher, N. Krishnakumar, B. Krishnamurthy and M. Ayyar, Phys. B, 2024, 674, 415597.
48. P. K. Jha, C. Pokhum, P. Soison, K. A. Techato and C. Chawengkijwanich, Water Sci. Technol., 2023, 88, 1564–1577.
49. K. R. Aranganayagam, S. Senthilkumaar, N. Ganapathi Subramaniam and T. W. Kang, Int. J. Nanosci., 2013, 12, 1350009.
50. G. Nagaraju, S. A. Prashanth, M. Shastri, K. V. Yathish, C. Anupama and D. Rangappa, Mater. Res. Bull., 2017, 94, 54–63.
51. K. R. Aranganayagam, S. Senthilkumaar, N. Ganapathi Subramaniam and T. W. Kang, Int. J. Nanosci., 2013, 12, 1350009.
52. A. Mahapatra, R. S. Ajimsha, D. Deepak and P. Misra, Sci. Rep., 2024, 14, 11871.
53. M. T. Dieng, B. D. Ngom, P. D. Tall and M. Maaza, Am. J. Nanomater., 2019, 7, 1.
54. J. Zhang, F. Jiang, Y. Yang and C. Sui, Mater. Lett., 2017, 195, 136–138.
55. T. Kumari, R. Gopal, A. Goyal and J. Joshi, J. Inorg. Organomet. Polym. Mater., 2019, 29, 316–325.
56. R. Bardhan, H. F. Zarick, A. Schwartzberg and C. L. Pint, J. Phys. Chem. C, 2013, 117, 21558–21568.
57. R. Madhu, A. Karmakar, S. Kumaravel, S. S. Sankar, K. Bera, S. Nagappan, H. N. Dhandapani and S. Kundu, ACS Appl. Mater. Interfaces, 2021, 14, 1077–1091.
58. J. Sahu, S. Kumar, V. S. Vats, P. A. Alvi, B. Dalela, S. Kumar and S. Dalela, J. Lumin., 2022, 243, 118673.
59. M. T. Uddin, Y. Nicolas, C. Olivier, L. Servant, T. Toupance, S. Li, A. Klein and W. Jaegermann, Phys. Chem. Chem. Phys., 2015, 17, 5090–5102.
60. H. W. Kunert, J. Barnas, D. J. Brink and J. Malherbe, J. Phys. IV, 2006, 132, 329–336.
61. Y. Deligiannakis, V. Tsikourkitoudi, P. Stathi, K. Wegner, J. Papavasiliou and M. Louloudi, Energy Fuels, 2020, 34, 15026–15038.
62. X. Jiang, J. Wang, T. Huang, G. Fu, Y. Tang, X. Qiu, J. Zhou and J. M. Lee, J. Mater. Chem. A, 2019, 7, 26243–26249.
63. J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben, Handbook of X-Ray Photoelectron Spectroscopy, Physical Electronics, Inc., Eden Prairie, MN, USA, 1995.
64. W. Xi, L. Jin, A. Mahmood, W. Zhang, Y. Li, H. Li, P. An, J. Zhang, T. Ma, S. Liu and J. Yan, Adv. Energy Mater., 2023, 13, 2302668.
65. Y. Dang, T. Wu, H. Tan, J. Wang, C. Cui, P. Kerns, W. Zhao, L. Posada, L. Wen and S. L. Suib, Energy Environ. Sci., 2021, 14, 5433–5443.
66. National Institute of Standards and Technology, NIST X-ray Photoelectron Spectroscopy Database, Version 3.0, Gaithersburg, MD, USA, http://srdata.nist.gov/XPS (accessed February 2026).
67. X. Zhang, Z. Su, D. Xiang, W. Xu, Q. Guo, Y. Fan, X. Kang, Y. Sheng, F. Zheng and W. Chen, Adv. Funct. Mater., 2024, 34, 2409306.
68. M. Ouyang, S. Cao, S. Yang, M. Li and M. Flytzani-Stephanopoulos, Ind. Eng. Chem. Res., 2020, 59, 2648–2656.
69. V. Platonov, A. Nasriddinov and M. Rumyantseva, Polymers, 2022, 14, 3481.
70. S. M. Majhi, H. J. Lee, H. N. Choi, H. Y. Cho, J. S. Kim, C. R. Lee and Y. T. Yu, CrystEngComm, 2019, 21, 5084–5094.
71. J. Low, J. Yu, M. Jaroniec, S. Wageh and A. A. Al-Ghamdi, Adv. Mater., 2017, 29, 1601694.
72. S. Wan, W. Wang, B. Cheng, G. Luo, Q. Shen, J. Yu, J. Zhang, S. Cao and L. Zhang, Nat. Commun., 2024, 15, 9612.
73. A. H. Naggar, A. S. Ahmed, T. A. S. El-Nasr, N. F. Alotaibi, K. F. Chong and G. A. Ali, Inorganics, 2023, 11, 484.
74. A. K. Ganguli, G. B. Kunde, W. Raza, S. Kumar and P. Yadav, Molecules, 2022, 27, 7778.
75. D. Hassen, M. A. Shenashen, A. R. El-Safty, A. Elmarakbi and S. A. El-Safty, Sci. Rep., 2018, 8, 3740.
76. J. Long, S. Wang, H. Chang, B. Zhao, B. Liu, Y. Zhou, W. Wei, X. Wang, L. Huang and W. Huang, Small, 2014, 10, 2791–2795.
77. F. Pellegrino, L. Pellutiè, F. Sordello, C. Minero, E. Ortel, V. D. Hodoroaba and V. Maurino, Appl. Catal., B, 2017, 216, 80–87.
78. K. Chen, J. Scott, F. Qu, W. Dong, D. C. Tsang and W. Li, J. Build. Eng., 2025, 112, 113816.
79. Y. Lu, Y. Lin, D. Wang, L. Wang, T. Xie and T. Jiang, Nano Res., 2011, 4, 1144–1152.
80. Z. Wu, X. Chen, X. Liu, X. Yang and Y. Yang, Nanoscale Res. Lett., 2019, 14, 147.
81. Y. Li, D. Chen, X. Xu, X. Wang, R. Kang, M. Fu, Y. Guo, P. Chen, Y. Li and D. Ye, Environ. Sci. Technol., 2023, 57, 3467–3485.
82. M. Sahooli, M. R. Rahimpour and M. Khorram, Catalysts, 2017, 7, 343.
83. I. Ahmad, M. Q. Alfaifi, Y. A. Alassmy, M. M. Abduljawad, M. A. Khasawneh, M. D. Ali, M. R. Karim and I. M. Ashraf, Int. J. Hydrogen Energy, 2025, 98, 1052–1066.
84. T. S. Moss, Proc. Phys. Soc., London, Sect. B, 1954, 67, 775–782.
85. E. Burstein, Phys. Rev., 1954, 93, 632.
86. K. Sakowski, Ł. Borowik, N. Rochat, P. Kempisty, P. Strak, N. Majewska, S. Mahlik, K. Koroński, T. Sochacki, J. Piechota and M. Fouchier, J. Lumin., 2024, 269, 120473.
87. S. Phanichphant, A. Nakaruk, K. Chansaenpak and D. Channei, Sci. Rep., 2019, 9, 16091.
88. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson and H. Nakatsujiet al., Gaussian 16, Revision C.01, Gaussian, Inc., Wallingford, CT, 2016.
89. P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
90. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305.
91. T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
92. T. J. Bora, N. K. Gour, L. Saikia, U. Bora, N. Hazarika, D. Kim, Y. B. Park, R. A. Saha, A. Devi, G. V. Karunakar, M. E. Zaki and K. K. Bania, Appl. Catal., A, 2025, 691, 120053.
93. Z.-Y. Wang, C.-Q. Han, J.-X. Guo, S. Sun and X.-Y. Liu, Chem. Sci., 2025, 16, 18390–18397.
94. E. Finkelstein, G. M. Rosen and E. J. Rauckman, Arch. Biochem. Biophys., 1980, 200, 1–16.
95. P. J. Wright and A. M. English, J. Am. Chem. Soc., 2003, 125, 8655–8665.
96. P. Veerakumar, A. Sangili, K. Saranya, A. Pandikumar and K. C. Lin, Chem. Eng. J., 2021, 410, 128434.
97. R. Gogoi, H. M. Dohling, A. Singh, K. Sharma, P. S. Sagara and P. F. Siril, J. Catal., 2022, 414, 109–124.
98. Z. F. Jiao, Y. M. Tian, B. Zhang, C. H. Hao, Y. Qiao, Y. X. Wang, Y. Qin, U. Radius, H. Braunschweig, T. B. Marder and X. N. Guo, J. Catal., 2020, 389, 517–524.
99. C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165–195.
100. C. Engelin, T. Jensen, S. Rodriguez-Rodriguez and P. Fristrup, ACS Catal., 2013, 3, 294–302.
101. S. Bera, L. M. Kabadwal and D. Banerjee, Chem. Soc. Rev., 2024, 53, 4607–4647.
102. L. A. Gallego-Villada, Sustainable Chem. Pharm., 2025, 46, 102062.
103. O. Altan and E. Kalay, Appl. Surf. Sci., 2022, 580, 152287.
104. C. Han, R. Qi, R. Sun, K. Fan, B. Johannessen, D. C. Qi, S. Cao and J. Xu, Appl. Catal., B, 2023, 320, 121954.
105. H. Q. Yang, Q. Q. Chen, F. Liu, R. Shi and Y. Chen, Chin. Chem. Lett., 2021, 32, 676–680.
106. B. Liu, Z. Yan, T. Xu, C. Li, R. Gao, H. Hao and J. Bai, Chem. Eng. J., 2022, 442, 136226.
107. N. Uğur, H. I. Önal, C. A. Ava, M. Aydemir and F. Durap, Mol. Catal., 2025, 585, 115389.
108. M. Mohammadikish and N. Mosleh, Appl. Organomet. Chem., 2023, 37, e7177.
109. P. G. Banda and R. Mucherla, ACS Omega, 2022, 7, 29356–29368.
110. G. Azadi, F. Kazemi and E. Firouzeh, ChemistrySelect, 2021, 6, 630–639.

---