Molecular additive-driven control of Cu/Cu₂O nanoparticle growth on mesoporous silica for enhanced photocatalytic hydrogen production

Abstract

Size-controlled cuprous oxide-based nanoparticles (NPs) are promising materials for enhancing visible-light-driven photocatalytic hydrogen production by increasing the number of Cu⁺ surface active sites. This study investigates the role of molecular additives in the growth of Cu/Cu₂O NPs on mesoporous silica (m-SiO₂) templates. The molecular additives cetyltrimethylammonium bromide (CTAB), ascorbic acid (AA), and citric acid (CA) are analyzed for their ability to modify the zeta potential of m-SiO₂, facilitating the adsorption of Cu²⁺ ions. The modified surface effectively controls the interaction between Cu²⁺ ions and the m-SiO₂ surface through the influence of molecular additives. The CTAB system facilitates a rapid nanoparticle (NP) growth and significant aggregation, thereby promoting the adsorption of Cu species and the subsequent formation of larger NPs. In contrast, CA provides better control over the formation of NPs, preventing aggregation through Cu²⁺ chelation and stabilizing the particles within the mesoporous voids of silica. Furthermore, the intensity ratio of metallic Cu to Cu₂O has the lowest value of 0.47 in the CA system, indicating a higher Cu₂O content compared to the CTAB and AA systems. CTAB and AA favor metallic Cu formation, while CA stabilizes Cu⁺ and promotes Cu₂O phase growth. As a result, the CA system achieves a 5-fold increase in the hydrogen production rate under visible light compared to the CTAB system. These findings highlight the critical role of molecular additives in tailoring NP growth and enhancing photocatalytic performance.

---

1. Introduction

The world is moving toward renewable energy due to drastic climate changes, the depletion of fossil fuel reserves, and the insistent need for sustainable development. Therefore, the demand for sustainable and green H₂ production has increased.^1,2 The photocatalytic process is a promising technology for green H₂ production due to its ability to utilize abundant solar energy and water. Predominantly, TiO₂ has excellent H₂ production activity. However, it has limitations of poor visible-light-driven H₂ production.^3–5 Meanwhile, Cu₂O-based catalysts absorb visible radiation, have less band gap energy as compared to TiO₂-based catalysts, and have high reduction potential for visible light H₂ production. Nevertheless, the rapid recombination of photogenerated charge carriers minimized the H₂ activity. Cu₂O composites are being utilized to enhance the density of photogenerated charge carriers while offering synergistic properties that facilitate efficient hydrogen production processes.^6–9

Cu^δ+ active sites are formed in copper-based catalysts through phase composition, which distinguishes their chemical and electrical properties, enabling the exploration of the active surface for effective reduction reactions.^10–13 The Cu⁺ site is the active center during photocatalytic H₂ production that interacts with H⁺ ions to produce molecular hydrogen. However, excessive oxidation of Cu⁺ to Cu²⁺ on the surface of catalysts can reduce the availability of active sites.^14,15 This is accomplished by structural control and multifunctional involvement in catalytic reactions, enhancing the catalyst's performance at the molecular level. For example, Qi Wu et al. reported that Cu(I) sites are generated in situ by reducing Cu(II) salts using AA.¹⁶ These generated Cu(I) sites are adsorbed onto the polymer surface, where they interact with nitrogen-rich triazole groups. These interactions stabilize Cu(I) and facilitate the formation of Cu₂O nanocrystals under weakly acidic conditions. However, these active sites become unstable during photocatalytic reactions, resulting in reduced efficiency and performance over time. Instability can be induced by surface degradation or photocorrosion, which diminishes the catalyst's long-term efficiency.

Photocorrosion of Cu₂O occurs due to self-photooxidation during photocatalytic reactions.¹⁷ This critical factor greatly influences and diminishes active site density, which can significantly reduce photocatalytic hydrogen performance. Various studies, for example, constructing core–shell structures enwrapped with graphene oxide, carbon-encapsulated nanostructures, or structures supported with metal–organic frameworks (MOFs), have been conducted to minimize the photocorrosion of Cu₂O and preserve the surface active sites.^18–20 Wang et al. reported that Zn-doped Cu₂O supported by graphdiyne (GDY) provides abundant active sites and redistribution of surface charges.²¹ The formation of an S-scheme heterojunction between Zn–Cu₂O and GDY created an internal electric field that suppressed electron–hole recombination. However, stability tests conducted with eosin Y (EY) as a sensitizer revealed that the H₂ production efficiency decreased over multiple reaction cycles due to photocorrosion and the degradation of EY molecules under prolonged light exposure. The researchers examined the porous nature of MOFs, which provide a protective environment for the effective stabilization of Cu₂O NPs, with their high surface area and tunable structures. The coordination between the metal sites in the MOFs and the Cu₂O NPs creates strong interactions that are advantageous for heterojunction formation, which prevents charge carrier recombination by creating internal electric fields and minimizes photocorrosion.²² Nevertheless, binder-free supported Cu₂O NPs are likely to aggregate, which results in limited numbers of active sites and surface areas. In this context, stabilized Cu–Cu₂O NPs supported with SiO₂ reduce aggregation and enhance the dispersion of Cu₂O NPs.^23,24 The mesoporous structure of SiO₂ not only stabilizes Cu₂O-based NPs but also facilitates efficient reactant diffusion. Moreover, m-SiO₂ has a negatively charged surface that strongly interacts with Cu⁺ ions for adsorption, facilitating nucleation sites for the growth of NPs.^25,26 This interaction can be controlled using molecular additives such as CTAB or AA.^27,28 Molecular additives used during the synthesis process can significantly affect the formation and stabilization of active sites on the surface of Cu-related phases, which are supportive of SiO₂.^29,30 However, these molecular additives vary the reactive environment of the negatively charged mesoporous SiO₂ surface to control the surface interaction of Cu ions. Also, it can be beneficial for the reduction of Cu ions in copper phases. These molecular additives have been shown to modify the phase distribution by influencing the nucleation and growth of copper phases.

This study investigated the influence of different molecular additives such as CTAB, AA, and CA on the preparation of the copper phase composition on m-SiO₂. The catalytic activity of these NPs was evaluated for photocatalytic H₂ production, focusing on understanding how the additives influenced the growth and aggregation of Cu species on the m-SiO₂ surface. HR-TEM analysis was used to examine the surface morphology and the role of these molecular additives in controlling the distribution of Cu species within the mesoporous voids of SiO₂. Furthermore, XRD analysis was employed to identify the metallic Cu and Cu₂O nanocrystalline phases formed in the nanocomposite. In addition, the XRD intensity ratios were analyzed to confirm the role of molecular additives in the synthesis of Cu-related crystalline phases. The results demonstrated that molecular additives significantly affect the growth rate of Cu species, aggregation behavior, and the formation of crystalline phases on the positively charged m-SiO₂ surface. Among the additives, CA was found to be particularly effective in optimizing and controlling electrostatic interactions with the mesoporous SiO₂ surface, ensuring uniform distribution and stable formation of Cu/Cu₂O NPs.

2. Experimental

Materials

Copper nitrate trihydrate (Cu(NO₃)₂·3H₂O, ≥99.9% trace metals basis), dimethyl sulfoxide (DMSO, ACS reagent, ≥99.9%), citric acid (CA, 99.0%), ascorbic acid (AA, 99.0%), cetyltrimethylammonium bromide (CTAB, ≥98.0%), sodium sulfate (Na₂SO₄, ≥99.0%), sodium sulfide (Na₂S, ≥99%), and sodium sulfite (Na₂SO₃, ≥98.0%) were obtained from Sigma-Aldrich and used as received without further purification.

Preparation of Cu/Cu₂O@m-SiO₂

0.1 mmol (60 mg) of m-SiO₂ with 200 nm diameter and 10 mL of DMSO were loaded into a 20 mL glass vial and sonicated until fully dispersed. CA as a molecular additive and Cu(NO₃)₂·3H₂O were added to the dispersed m-SiO₂ solution. The amounts of CA and Cu(NO₃)₂·3H₂O used were 0.2 and 0.4 mmol, respectively. The glass vial with the solution sample was heated to 100 °C while stirring to partially evaporate the DMSO solvent and obtain a highly viscous paste, which was further dried under vacuum at 120 °C. The obtained dry powder mixture was ground in an agate mortar to achieve a homogeneous powder mixture. The mixture was then heated in a tube furnace at various temperatures ranging from 200 to 400 °C in 100 °C intervals under a 5% H₂ atmosphere for 5 h. The flow rate of 5% H₂ gas was maintained at 30 cc min⁻¹. Finally, the samples were cooled to room temperature naturally. Furthermore, catalysts were synthesized at different amounts of CA: 0.1, 0.2, and 0.4 mmol. All other experimental conditions were kept constant. The resulting dry powder was annealed at the optimized temperature of 300 °C. For comparison, additional catalysts were synthesized using different molecular additives such as AA and CTAB. The obtained products were used for further characterization.

Characterization

Thermogravimetric analysis (TGA) was performed using a Mettler-Toledo TGA/DSC 1 instrument under a nitrogen atmosphere at a heating rate of 5 °C min⁻¹ from room temperature to 800 °C. Transmission electron microscopy (TEM) images were obtained using a JEOL-2100F electron microscope operated at 200 keV. X-ray diffraction (XRD) patterns of the obtained powder product were recorded with a Bruker AXS D8 diffractometer using Cu-Kα radiation at λ = 1.54 Å. X-ray photoelectron spectroscopy (XPS) spectra were obtained using a Thermo Fisher Scientific/K-alpha + spectrometer equipped with a monochromatic Al Kα X-ray source (hν = 1486.6 eV). Field-emission scanning electron microscopy (FE-SEM) coupled with energy-dispersive X-ray spectroscopy (EDS) was performed using a JEOL microscope (JSM-7610F) to examine the elemental composition of the samples and the corresponding elemental mapping images. Sulfur contents before and after photocatalytic H₂ production were measured using an automatic elemental analyzer by combusting the sample at approximately 1800 °C and detecting it with a TCD. UV-vis absorption spectra of the sample dispersed in the DMSO solvent were recorded under ambient conditions using a Shimadzu UV-2600 spectrophotometer. The diffuse reflectance spectra of the produced catalysts were recorded using a UV-vis-NIR spectrophotometer (Thermo Scientific, Evolution One Plus UV-vis). EIS and Mott–Schottky curves were recorded in Na₂SO₄ using a potentiostat (Biologic VSP). The photoactivity of the synthesized catalysts was tested by recording I–t curves under chopped illumination in Na₂SO₄ electrolyte. The zeta potential of m-SiO₂ dispersed in DMSO was measured using a Zetasizer Nano ZS (Malvern). For solution preparation, 0.1 mmol (60 mg) of mSiO₂ was dispersed in 10 mL of DMSO by ultrasonication, and then surfactant was added. The resulting solution was diluted six-fold to reduce the m-SiO₂ concentration to 1 mg mL⁻¹.

Photocatalytic activity

The photocatalytic H₂ generation activity of the catalyst was investigated under visible light with a 400 nm cut-off filter. 5 mg of the catalyst was ultrasonically dispersed in 10 ml of Na₂S and Na₂SO₃ as sacrificial agents. Afterwards, Ar gas was bubbled through the system for 30 min. A 300 W Xe lamp was used to illuminate the reaction vial. The amount of H₂ was evaluated using gas chromatography (Youngin YL 6500) with Ar as the carrier gas.

3. Results and discussion

A schematic representation of the synthesis of Cu/Cu₂O nanocomposites supported by Cu₂S on porous m-SiO₂ NPs is shown in Fig. 1. This process involves the use of CTAB, AA, and CA molecular additives to prepare the active sites of Cu(I) catalysts in composite NPs for photocatalytic H₂ production. Nonetheless, the surface interaction of m-SiO₂ dispersed in DMSO with molecular additives and Cu ions plays a crucial role in the distribution of NPs on the surface of mesoporous SiO₂. This surface interaction is vital for controlling the adsorption of Cu species by forming a molecular bridge on the surface of the SiO₂. These molecular additives alter the surface charges, and the resulting surface charges were measured by evaluating the zeta potential. The zeta potential values for m-SiO₂ (1 mg mL⁻¹) dispersed in DMSO with different additives (CTAB, AA, and CA) are presented in Table 1. The m-SiO₂ NPs inherently possess a negatively charged surface. This negative charge results in a highly negative zeta potential of −44.70 mV, which is subsequently modified by the addition of molecular additives. The zeta potential values for CTAB, AA, and CA are −38.20, −33.97, and −28.40 mV, respectively. This reduction in the zeta potential is attributed to interactions between the molecular additives and the negatively charged m-SiO₂ surface. Although both CA and AA contain carboxyl groups, CA reduced the surface charge more effectively with a zeta potential value of −28.40 mV, likely due to the formation of multiple hydrogen bonds via its three carboxyl groups. In contrast to AA, which directly reduces Cu²⁺ species, CA coordinates with both Cu²⁺ and the SiO₂ surface, thus preventing aggregation of the NPs. TGA revealed that the carboxyl groups in CA and AA decompose near 300 °C (Fig. S1 and S2†). This observation is consistent with the report by Venkatesha et al., who found that the Cu–citric acid complex contains coordinated carboxyl and hydroxyl groups.³¹ These groups decompose at 300 °C under an inert atmosphere, leading to the formation of Cu₂O NPs. Therefore, the synthesized Cu precursor supported with the additive complex decomposed at 300 °C.

Fig. 1 Schematic representation of the preparation of Cu/Cu₂O NPs on mesoporous SiO₂ using various molecular additives.

Table 1 Zeta potentials of m-SiO₂ in DMSO with various molecular additives

Sample	Solvent	Molecular additive	Zeta potential (mV)
m-SiO2	DMSO	None	−44.70 ± 0.15
		CTAB	−38.20 ± 1.40
		AA	−33.97 ± 0.21
		CA	−28.40 ± 0.22

---

HR-TEM images were used to examine the loading behavior of the Cu/Cu₂O NPs on the surface of m-SiO₂ using different molecular additives (Fig. 2a–c). The TEM images reveal significant differences in aggregation behaviour among the samples. Notably, the aggregation of Cu species on the m-SiO₂ NP surface is most pronounced in the CTAB-assisted system compared to the CA-assisted and additive-free system. This increased aggregation of NPs in the system is attributed to the strong surface adsorption of Cu ions, which facilitates the formation of nucleation sites for Cu²⁺ species in solution. Because the zeta potential values are below the isoelectric point, the CTAB on the negatively charged m-SiO₂ results in a less negative zeta potential value compared to that on bare m-SiO₂, but the surface is not fully neutralized.³² This modified surface facilitates control of the adsorption and aggregation of Cu species on the SiO₂ surface. Furthermore, during hydrogenation treatments, decomposition of CTAB occurs, leading to the formation of aggregated Cu/Cu₂O NPs. The CA-assisted system shows reduced NP aggregation due to more controlled Cu²⁺ adsorption. As a weak molecular acid, CA interacts with m-SiO₂ through its three carboxylate groups, lowering the zeta potential from −38.20 to −28.40 mV. Compared to AA and CTAB, CA achieves higher surface coverage and better charge neutralization. Furthermore, CA chelates Cu²⁺ ions, enabling uniform dispersion and limiting nucleation sites. N. Balighieh et al. reported that chelation involves coordinating metal ions with silicon to form coordination complexes.³³ Consequently, a metallic citric complex formed at the atomic scale.³⁴ According to P. Sukpanish et al., the presence of carboxyl and hydroxyl groups on the surface of this compound facilitates the formation of metallic–citrate complexes through coordination with metal ions.³⁵ This suggests that the CA molecule can act as a bridge surrounding the SiO₂ surface via Cu²⁺ ions and control the extensive aggregation of NPs. This behavior promotes more sites for homogeneous nucleation and controlled growth of Cu species, preventing excessive aggregation on the m-SiO₂ surface. In the absence of molecular additives, the negatively charged m-SiO₂ surface induces electrostatic attraction with Cu²⁺ ions, leading to rapid nucleation and subsequent growth of Cu species. However, this process lacks the controlled distribution observed in the CA-assisted system. HR-TEM images of the AA-assisted system are shown in Fig. S3.† Therefore, it is inferred that CA has the optimum reaction rate for growing Cu species on m-SiO₂ and prevents rapid aggregation of Cu species on the surface of m-SiO₂. HAADF-STEM analysis also supports these findings, demonstrating a similar trend in the aggregation and distribution of Cu species (Fig. 2d–f). Elemental mapping images further confirm the uniformity and distribution of Cu elements in the CA-assisted system compared to the others (Fig. 2j–l).

Fig. 2 (a–c) TEM images and (d–f) HAADF-STEM images of Cu/Cu₂O, Cu/Cu₂O–CTAB, and Cu/Cu₂O–CA. Elemental mapping images of the (g–i) O elements and (j–l) Cu elements in Cu/Cu₂O, Cu/Cu₂O–CTAB, and Cu/Cu₂O–CA.

Fig. 3 shows the XRD patterns of the photocatalysts prepared with different molecular additives. All samples exhibit characteristic peaks corresponding to cubic metallic Cu and Cu₂O phases, which clearly match JCPDS cards 003-1005 and 005-0667, respectively. Variations in the diffraction intensities of these phases were observed. These exposed phases depend on the molecular additive used in the synthesis of NPs. The intensity ratio of the major diffraction peak of metallic Cu to that of Cu₂O was calculated to determine the dominant Cu phase in each sample (Fig. 3b). The intensity ratios for Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, Cu/Cu₂O, and Cu/Cu₂O–CA were 4.52, 4.22, 0.48, and 0.47, respectively. These results suggest that molecular additives such as CTAB and AA favor the formation of metallic Cu, whereas the CA system significantly suppresses metallic Cu and enhances the Cu₂O phase. This suggests that CA supports the stabilization of oxidized Cu₂O phases through its chelation capability, which suppresses complete reduction to metallic Cu. In contrast, AA primarily acts as a reducing agent, promoting rapid conversion of Cu²⁺ to Cu⁺ and Cu⁰. Thus, CA treatment yields a higher population of Cu⁺ active sites than AA, as shown in Fig. 3a. This is beneficial for photocatalytic H₂ generation. Further insights were provided by FT-IR analysis, as shown in Fig. 3c, which confirmed the presence of functional groups associated with the Cu₂O structure. The spectra exhibited Cu–O stretching vibrations at around 616 and 666 cm⁻¹, consistent with the Cu₂O lattice structure.³⁶ Interestingly, minor peaks corresponding to Cu₂S were also observed in the XRD patterns. These peaks are attributed to the synthesis conditions, particularly the use of DMSO as a solvent. At 100 °C, the highly viscous nature of the reaction mixture due to DMSO prevented complete solvent evaporation. This concentrated environment, enriched with molecular additives and Cu²⁺ complexes, facilitated the formation of Cu–DMSO complexes.³⁷ The thermal decomposition of these complexes resulted in crystalline Cu₂S.

Fig. 3 (a) XRD patterns and (b) intensity ratios of the major diffraction planes of metallic Cu to Cu₂O. (c) FT-IR spectra of Cu/Cu₂O, Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA.

The surface oxidation states and bonding characteristics of the Cu species were analyzed through deconvoluted XPS spectra of Cu 2p, as shown in Fig. 4a. Peaks at 932.22 and 934.00 eV are attributed to the Cu⁰/Cu⁺ and Cu²⁺ oxidation states, respectively.³⁸ The presence of strong satellite peaks further confirms the Cu²⁺ oxidation state. These results indicate that the synthesized NPs consist of a mixture of Cu₂O, metallic Cu, and CuO, suggesting that the molecular additives used during synthesis facilitated the formation of mixed oxidation states (Fig. 4a). These findings suggest that molecular additives promote the formation of mixed oxidation states in cuprous oxide-based NPs.

Fig. 4 (a) Cu 2p and (b) O 1s XPS spectra of Cu/Cu₂O, Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA. (c) TRPL spectrum of Cu/Cu₂O–CA. (d) Average lifetimes of Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA.

The influence of molecular additives on the surface chemical environment is evident from the shifts in XPS peak positions. The spin–orbit splitting between the Cu 2p_3/2 and Cu 2p_1/2 peaks was approximately 19.8 eV. However, variations in binding energy shifts were observed, reflecting differences in oxidation states and surface environments. For Cu/Cu₂O–AA, lower shifts in binding energy indicate a higher proportion of metallic Cu, consistent with the reducing nature of AA. Conversely, a positive binding energy shift was observed in Cu/Cu₂O–CA, which promotes the stabilization of the Cu₂O phase. Compared to the AA and CTAB systems, the CA-treated catalysts and those prepared without reducing agents exhibited a higher Cu₂O content. In addition, the peak at 935.85 eV in the deconvoluted XPS spectra was attributed to the hydroxyl groups bonded to Cu species. The bonding of oxygen to Cu in the lattice was confirmed through O 1s XPS peaks (Fig. 4b). The peak at 531.1 eV corresponds to oxygen atoms integrated into the crystal lattice of the oxide. In comparison, the strong peak at 532.7 eV is associated with the m-SiO₂-support material. Further insights were obtained from the XPS spectra of Si 2p (Fig. S4†) and C 1s (Fig. S5†). The C 1s peaks likely originate from molecular additives or residual carbon left after incomplete decomposition during synthesis. These carbon functional groups were also confirmed in the FT-IR spectra (Fig. 3c), with a band at 1650 cm⁻¹ indicating residual molecular compounds in the catalyst.

Fig. 4c and S6 and S7† show the TRPL decay profiles of the Cu/Cu₂O–CA, Cu/Cu₂O–CTAB, and Cu/Cu₂O–AA catalysts. The decay profiles were analyzed using a tri-exponential decay fitting model,³⁹ providing lifetime components of τ₁, τ₂, and τ₃, which correspond to fast, intermediate, and slow recombination processes, respectively. These components are summarized in Table S1.† The τ₁ is attributed to non-radiative recombination involving surface trap states, while τ₂ and τ₃ reflect charge carriers trapped by shallow trap states caused by structural defects and radiative recombination of charge carriers.^40,41 Among the systems, Cu/Cu₂O–CA exhibits a higher τ₂ value than Cu/Cu₂O–AA but lower than Cu/Cu₂O–CTAB, suggesting efficient interfacial charge transfer in the CA-assisted composite NPs. This trend is consistent with the mean lifetime (τ_avg), as shown in Fig. 4d. The τ_avg values for Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA are 4.14, 3.30, and 3.43 ns, respectively. The higher τ_avg value in the CTAB system may be attributed to the formation of a higher proportion of the Cu₂S phase, which can influence recombination processes. However, the slightly increased τ_avg value in the CA system (1.04 times higher than that of Cu/Cu₂O–AA) indicates improved charge transfer efficiency.

Fig. 5a–c show the Mott–Schottky (MS) plots of Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA. All samples exhibit a negative slope in the linear region of the MS plots, confirming their p-type conductivity. Notably, this conductivity remains unaffected by the use of different molecular additives. The flat band potential (V_fb) was determined by extrapolating the linear portion of the MS plots at 1/C² = 0. The V_fb values for Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA were found to be 1.48, 1.54, and 1.33 V vs. RHE, respectively. The observed variation in the V_fb value is attributed to the presence of mixed Cu-related crystalline phases. The acceptor concentration (N_A) was calculated using the formula N_A = 2/(q·ε·ε₀ × slope). The N_A values were 5.80 × 10²⁰ cm⁻³, 8.86 × 10²⁰ cm⁻³, and 1.15 × 10²¹ cm⁻³ for Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA, respectively (Fig. 5d). Cu/Cu₂O–CA demonstrates the highest N_A value, indicating improved charge carrier mobility and charge carrier separation. This was confirmed by the photoelectrochemical (PEC) response (Fig. 5e). The superior performance of the CA system is likely due to the abundance of surface-active sites on m-SiO₂, which effectively support NP loading. In contrast, the CTAB-assisted system exhibited significant NP aggregation, minimizing the exposed surface area and reducing the accessibility for active sites. Furthermore, residual CTAB within the NPs, retained during synthesis, decomposes between 200 and 300 °C.⁴² The retained residual of CTAB obstructed the surface-active sites, hindering their accessibility for catalytic reactions. This limitation can adversely affect the efficiency of surface-related reactions by reducing the number of available active sites and diminishing the overall performance. The charge transport behavior of Cu/Cu₂O–CA under light and dark conditions was analyzed through electrochemical impedance spectroscopy (EIS), as shown in Fig. 5f. The charge transfer resistance (R_ct) was determined by fitting the EIS data to an equivalent circuit model.³⁹ For Cu/Cu₂O–CA, the R_ct value decreased from 164 × 10³ Ω under dark conditions to 114 × 10³ Ω under light irradiation. This significant reduction suggests efficient utilization of photogenerated charges for the reduction of H⁺ ions, driving visible light-induced hydrogen production.

Fig. 5 (a–c) Mott–Schottky plots recorded at a constant frequency of 1 kHz. (d) Accepter density values and (e) I–t curves recorded under chopped conditions of Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA. (f) EIS curves of Cu/Cu₂O–CA recorded under dark and light conditions.

The H₂ production activity of the photocatalysts synthesized at different temperatures (200, 300, and 400 °C) was evaluated under visible light irradiation. As shown in Fig. S8,† the sample prepared at 300 °C exhibited the highest H₂ evolution rate. This temperature optimally balances the decomposition of organic ligands and the stabilization of catalytically active Cu⁺ species on the m-SiO₂ surface. Fig. 6a and b show the H₂ production activities of catalysts prepared using CA, AA, and CTAB. An increase in H₂ production was observed after one hour of illumination. This may be attributed to the activation of the photocatalyst surface through the formation of a secondary phase. These results are consistent with previous studies reporting metal sulfide phase formation during H₂ production.⁴³ The H₂ production rates for Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA were determined to be 38.33, 86.49, and 135.81 μmol g⁻¹ h⁻¹, respectively. These results clearly indicate that Cu/Cu₂O–CA exhibits superior photocatalytic activity compared to the other systems. The enhanced performance of Cu/Cu₂O–CA can be attributed to the stabilization of nanosized Cu species within the voids of the m-SiO₂ surface, which prevents NP aggregation and increases the active surface area for the H₂ production reaction. Moreover, the higher availability of Cu⁺ sites in the CA system further promotes efficient charge separation and facilitates proton reduction to hydrogen, thereby enhancing the overall photocatalytic activity.

Fig. 6 (a) Photocatalytic H₂ production and (b) H₂ production rates for commercial Cu₂O, Cu/Cu₂O–CTAB, Cu/Cu₂O–AA, and Cu/Cu₂O–CA. (c) Photocatalytic H₂ production and (d) H₂ production rates for the Cu/Cu₂O–CA catalyst with various scavengers. (e) Stability of the Cu/Cu₂O–CA catalyst over ten consecutive H₂ production cycles.

In contrast, the Cu/Cu₂O–CTAB system displayed rapid aggregation of NPs, significantly reducing the number of active sites available for catalysis and, consequently, lowering H₂ production. To further evaluate the performance of Cu/Cu₂O–CA, its photocatalytic H₂ production rate was compared with that of commercial Cu₂O NPs. Remarkably, Cu/Cu₂O–CA exhibited a 9.1-fold higher H₂ production rate than the commercial Cu₂O NPs, underscoring the effectiveness of the synthesized catalyst (Fig. 6b). Moreover, the effect of varying CA concentrations on H₂ production activity was investigated (Fig. S9†). The results revealed that the catalytic performance improved with increasing CA concentration, reaching an optimal point. This enhancement can be attributed to the dual functionality of CA as both a structure-directing agent and a metal-complexing ligand, which promotes uniform dispersion of copper species and better control over NP morphology. Furthermore, the H₂ production activity of optimized Cu/Cu₂O–AA was tested in the presence of different sacrificial agents, including 10 v/v% TEOA, methanol, and glycerin, as shown in Fig. 6c. The results demonstrated a linear, time-dependent increase in H₂ production activity. However, the rates were lower compared to those obtained using a Na₂S/Na₂SO₃ scavenger. These findings suggest that the reaction of the catalyst surface with S²⁻ ions in the Na₂S/Na₂SO₃ system contributes to the formation of a secondary phase, enhancing H₂ production efficiency. The order of H₂ production rates (Fig. 6d) for various scavengers with Cu/Cu₂O–AA was as follows: Na₂S/Na₂SO₃ > TEOA > methanol > glycerin. Notably, the Na₂S/Na₂SO₃ scavenger exhibited the highest H₂ production rate, 12.8-, 15.5-, and 458.8-fold higher than those of TEOA, methanol, and glycerin, respectively. The stability and reusability of the Cu/Cu₂O–CA catalyst were evaluated through ten consecutive photocatalytic H₂ production cycles (Fig. 6e). Interestingly, the activity in the second cycle increased by 1.26-fold compared to the first run, as shown in Fig. 6e, possibly due to surface reconstruction or activation. A slight decline in activity was observed in later cycles, which correlates with structural and compositional changes in the catalyst. XRD analysis (Fig. S10†) revealed a phase transformation from Cu₂O to Cu₂S during the reaction. The XPS spectra (Fig. S11†) exhibited a shift of Cu 2p peaks toward lower binding energies, which may be attributed to surface chemical changes induced by sulfur species with copper. This interaction is responsible for the observed increase in sulfur species on the catalyst surface during H₂ production, as clearly shown in the EDS mapping images (Fig. S12 and S13†). The elemental maps reveal a denser and more homogeneous distribution of sulfur after the reaction. Quantitative EDS data, as summarized in Table S2,† indicate a significant increase in the atomic percentage of sulfur. Elemental analysis (Table S3†) confirms this trend, revealing a threefold increase in sulfur content after the photocatalytic reaction. These findings strongly suggest the in situ formation of Cu₂S species during H₂ evolution.

The charge transfer mechanism in the Cu/Cu₂O/Cu₂S heterojunction for photocatalytic H₂ production is schematically illustrated in Fig. 7. The proposed mechanism is based on the energy band positions and their role in the transfer of photogenerated electron–hole pairs. Metallic Cu has a work function of 4.20–4.65 eV.⁴⁴ Cu₂O, a p-type semiconductor, has its Fermi level above the valence band,⁴⁵ with an electron affinity of approximately 3.2 eV (ref. 46) and a work function of 4.80–5.00 eV.^47,48 The difference in the work functions of Cu and Cu₂O creates a barrier height at the Cu₂O/Cu interface,⁴⁹ forming an electron-rich region. This induces electron transfer between the materials until equilibrium is reached. This leads to the formation of band bending due to dissimilar work functions, which creates a built-in electric field.⁵⁰ The built-in electric field within the Cu/Cu₂O/Cu₂S heterojunction plays a crucial role in enhancing charge separation and suppressing charge recombination. The photogenerated electrons and holes possess thermodynamically sufficient potentials to drive water splitting.⁵¹ In particular, Cu⁺ species act as more active surface sites for redox reactions. The use of CA as a molecular additive stabilizes a higher proportion of Cu⁺ species, as evidenced by XRD analysis, where the Cu₂O phase was the dominant component in the CA-prepared catalyst. In contrast, the CTAB and AA systems showed higher metallic Cu content. These compositional variations between metallic Cu and Cu₂O can alter the electronic properties of the material. They affect the alignment of band positions, the Fermi level, and the density of states at the interface.⁵² Therefore, the CA additive promotes the formation and retention of Cu₂O, which has a conduction band edge sufficiently negative to reduce protons. Suppaso et al. have demonstrated that Cu-based heterojunctions benefit from such favorable energy band alignment, which facilitates photogenerated electron transfer to active sites, enabling efficient visible-light-driven water splitting.¹² As a result, favorable energy band alignment directly influences the charge carrier separation efficiency and the overall photocatalytic performance of the material. Upon photon absorption by Cu₂O, electrons are excited from its valence band (VB) to its conduction band (CB). Simultaneously, the built-in electric field at the heterojunction interface facilitates the migration of holes toward the Cu₂S side. The photogenerated electrons in the CB of Cu₂O are subsequently transferred to the Cu layer and react with H⁺ ions, resulting in molecular hydrogen.⁵³ During H₂ production, the partial transformation of Cu₂O into Cu₂S was confirmed by elemental analysis, which further supports the formation of a heterostructure that aids in charge separation. Such multi-phase heterojunctions highlight the synergistic effects of compositional tuning, interface engineering, and additive-assisted phase control in enhancing the photocatalytic H₂ production performance of the Cu/Cu₂O–CA system, as shown in Fig. 6a.

Fig. 7 Charge transfer and separation mechanism through a built-in electric field over the surface of Cu/Cu₂O–CA supported with Cu₂S for H₂ production.

4. Conclusions

Molecular additives are demonstrated in this study as bridges that promote controlled Cu⁺ ion adsorption onto negatively charged surfaces and control the growth of NPs. A molecular layer of CTAB, AA, and CA formed on m-SiO₂ systematically reduced the highly negative zeta potential of m-SiO₂ (−44.70 mV) to −38.20, −33.97, and −28.40 mV. This demonstrates that the CTAB system exhibited a faster NP growth rate and higher aggregation compared to the CA system. The CTAB system enhanced the interaction of Cu⁺ ions with the negatively charged silica surface. However, it has limited capping ability, resulting in uncontrolled particle growth and clustering. In contrast, CA offers greater control over NP formation by functioning as both a chelating agent and a stabilizer, promoting gradual growth while preventing aggregation. The XRD results revealed that CTAB and AA molecular additives are more favorable for the formation of the metallic copper phase, while the CA molecular additive stabilizes the Cu₂O phase. This was confirmed by the major peak XRD intensity ratio of metallic Cu to Cu₂O. In addition, the acceptor density of the CA system used to prepare the catalyst is 1.15 × 10²¹ cm⁻¹, which is higher than that of the CTAB and AA-prepared system. The results show that Cu/Cu₂O–CA exhibits superior photocatalytic activity for H₂ production under visible light compared to Cu/Cu₂O–CTAB and Cu/Cu₂O–AA. The built-in electric field between metallic Cu and Cu₂O is significantly influenced by the separation of charge carriers, which depends on the electronic properties of the materials and can be considerably affected by phase variations between Cu and Cu₂O. In addition, the progressive transformation of Cu₂O into copper sulfide during photocatalysis led to an enhanced photocatalytic performance. Elemental analysis revealed a 3.727-fold increase in sulfur species after H₂ production. These findings highlight the importance of selecting appropriate molecular additives to achieve the desired NP properties.

Author contributions

Santosh V. Mohite: conceptualization, data curation, and writing the original draft. Artavazd Kirakosyan: experimentation, conceptualization, and investigation. Kwangchan An: validation and EIS analysis. Yeongseok Shim: formal analysis. Jihoon Choi: investigation, data curation, and supervision. Yeonho Kim: conceptualization, investigation, supervision, data curation, and project administration. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting this article have been included as part of the ESI.†

Acknowledgements

This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education (RS-2021-NR060128) and the Korean government (MSIT) (No. RS-2025-00554653).

References

1. P. Shen, B. Zhou, Z. Chen, W. Xiao, Y. Fu, J. Wan, Z. Wu and L. Wang, Appl. Catal., B, 2023, 325, 122305.
2. R. Kumar, R. Singh and S. Dutta, Energy Fuels, 2024, 38, 2601–2629.
3. Y. Chen, L. Soler, C. Cazorla, J. Oliveras, N. G. Bastús, V. F. Puntes and J. Llorca, Nat. Commun., 2023, 14, 6165.
4. C. Xing, L. Yang, M. C. Spadaro, Y. Zhang, P. Guardia, J. Arbiol, T. Liu, X. Fan, M. Fernández-García, J. Llorca and A. Cabot, ACS Appl. Electron. Mater., 2024, 6, 5833–5841.
5. J. Hou, L. Gan, N. Bao, C. Zhang, K. Liu and Q. Wang, J. Water Process Eng., 2024, 68, 106364.
6. M. Muscetta, R. Andreozzi, L. Clarizia, I. Di Somma and R. Marotta, Int. J. Hydrogen Energy, 2020, 45, 28531–28552.
7. X. Zhao, M. Lei, Z. Chen and Z. Jin, J. Environ. Chem. Eng., 2024, 12, 114181.
8. M. R. Dustgeer, A. Jilani, M. O. Ansari, M. B. Shakoor, S. Ali, A. Imtiaz, H. S. Zakria and M. H. D. Othman, J. Water Process Eng., 2024, 59, 105053.
9. H. Wang, M. Xiao, Z. Wang, X. Chen, W. Dai and X. Fu, Chem. Eng. J., 2023, 459, 141616.
10. J. Lee, J. H. Seo, C. Nguyen-Huy, E. Yang, J. G. Lee, H. Lee, E. J. Jang, J. H. Kwak, J. H. Lee, H. Lee and K. An, Appl. Catal., B, 2021, 282, 119576.
11. H. Son, J.-H. Lee, P. Uthirakumar, D. Van Dao, A. Soon and I.-H. Lee, J. Alloys Compd., 2023, 942, 169094.
12. C. Suppaso, T. Kanazawa, S. Nozawa, K. Maeda and N. Khaorapapong, ChemCatChem, 2024, 16, e202300858.
13. S. V. Mohite, C. Lee, K. An and Y. Kim, J. Alloys Compd., 2024, 1008, 176601.
14. K. Sekar, C. Chuaicham, B. Vellaichamy, W. Li, W. Zhuang, X. Lu, B. Ohtani and K. Sasaki, Appl. Catal., B, 2021, 294, 120221.
15. S.-T. Guo, Z.-Y. Tang, Y.-W. Du, T. Liu, T. Ouyang and Z.-Q. Liu, Appl. Catal., B, 2023, 321, 122035.
16. Q. Wu, R. Ma, Y. Wang, J. Li, H. Xue, Z. Zhang, L. Xu, H. Fu, L.-Z. Wu, X.-B. Li and H. Wang, Catal. Sci. Technol., 2024, 14, 2574–2579.
17. C. Y. Toe, Z. Zheng, H. Wu, J. Scott, R. Amal and Y. H. Ng, Angew. Chem., Int. Ed., 2018, 57, 13613–13617.
18. Y. Liu, H. Tan, Y. Wei, M. Liu, J. Hong, W. Gao, S. Zhao, S. Zhang and S. Guo, ACS Nano, 2023, 17, 5994–6001.
19. N. Li, W. Yan, Y. Niu, S. Qu, P. Zuo, H. Bai and N. Zhao, ACS Appl. Mater. Interfaces, 2021, 13, 9838–9845.
20. N. Li, W. Yan, W. Zhang, Z. Wang and J. Chen, J. Mater. Chem. A, 2019, 7, 19324–19331.
21. M. Wang, K. Wang and Z. Jin, J. Environ. Chem. Eng., 2024, 12, 113530.
22. J.-P. Chang, C.-Y. Wang, Y.-J. Hsu and C.-Y. Wang, Appl. Catal., A, 2023, 650, 119005.
23. P. Zhu, Y. Li, Bilal, Y. Ma, X. Ruan and Q. Zhang, Ceram. Int., 2023, 49, 12518–12528.
24. R. Huang, L. Chen, C. Yu, X. Zhang, H. Peng, W. Sun and Q. Du, J. Environ. Chem. Eng., 2023, 11, 110326.
25. J. A. Greathouse, T. J. Duncan, A. G. Ilgen, J. A. Harvey, L. J. Criscenti and A. W. Knight, Environ. Sci.: Nano, 2021, 8, 1992–2005.
26. Y. Niu, W. Yu, S. Yang and Q. Wan, Sci. Rep., 2024, 14, 13521.
27. B. Fouconnier, A. Román-Guerrero and E. J. Vernon-Carter, Colloids Surf., A, 2012, 400, 10–17.
28. D. Videira-Quintela, F. Guillén, G. Montalvo and O. Martin, Colloids Surf., B, 2020, 195, 111216.
29. S. Sasaki, E. Kurniawan, K. Sato, K. Matsusaka, T. Kojima, T. Hara, Y. Yamada and S. Sato, Appl. Catal., A, 2024, 671, 119561.
30. S. Hosaka, E. Kurniawan, Y. Yamada and S. Sato, Appl. Catal., A, 2023, 653, 119079.
31. N. J. Venkatesha and S. Ramesh, Ind. Eng. Chem. Res., 2018, 57, 1506–1515.
32. M. Khademi, W. Wang, W. Reitinger and D. P. J. Barz, Langmuir, 2017, 33, 10473–10482.
33. N. Balighieh, M. R. Zamani, S. F. Kashani-Bozorg, M. Kheradmandfard, M. R. Barati and E. Mansoori, J. Mater. Res. Technol., 2024, 30, 2211–2222.
34. D. S. Todorovsky, M. M. Getsova, I. Wawer, P. Stefanov and V. Enchev, Mater. Lett., 2004, 58, 3559–3563.
35. P. Sukpanish, B. Lertpanyapornchai, T. Yokoi and C. Ngamcharussrivichai, Mater. Chem. Phys., 2016, 181, 422–431.
36. S. Yadav, M. Chauhan, M. Jacob and P. Malhotra, New J. Chem., 2022, 46, 9685–9694.
37. A. R. Uhl, J. K. Katahara and H. W. Hillhouse, Energy Environ. Sci., 2016, 9, 130–134.
38. S. V. Mohite, S. Kim, C. Lee, J. Bae and Y. Kim, J. Alloys Compd., 2022, 928, 167168.
39. S. R. Jadhav, S. V. Mohite, K. C. An, D. H. Jang, K. Erande, Y. Kim and K. Y. Rajpure, Ceram. Int., 2024, 50, 22783–22794.
40. M. K. Singh and M. S. Mehata, Ceram. Int., 2021, 47, 32534–32544.
41. E. Mitchell, A. Law and R. Godin, J. Photochem. Photobiol., C, 2021, 49, 100453.
42. W. L. S. Soares, L. F. Feitosa, C. R. Moreira, F. Bertella, C. W. Lopes, A. M. D. de Farias and M. A. Fraga, ACS Appl. Mater. Interfaces, 2025, 17, 13146–13161.
43. S. R. Jadhav, S. V. Mohite, C. Lee, J. Bae, R. S. Pedanekar, Y. Kim and K. Y. Rajpure, Sustainable Mater. Technol., 2023, 38, e00731.
44. B. Shahzad Khan, T. Mehmood, A. Mukhtar and M. Tan, Mater. Lett., 2014, 137, 13–16.
45. C.-S. Tan and M. H. Huang, Inorg. Chem. Front., 2021, 8, 4200–4208.
46. S.-C. Wu, C.-S. Tan and M. H. Huang, Adv. Funct. Mater., 2017, 27, 1604635.
47. G.-Z. Yuan, C.-F. Hsia, Z.-W. Lin, C. Chiang, Y.-W. Chiang and M. H. Huang, Chem. – Eur. J., 2016, 22, 12548–12556.
48. C.-S. Tan, S.-C. Hsu, W.-H. Ke, L.-J. Chen and M. H. Huang, Nano Lett., 2015, 15, 2155–2160.
49. M. G. Rahimi, A. Wang, G. Ma, N. Han and Y. Chen, RSC Adv., 2020, 10, 40916–40922.
50. Z. Wei, Q. Mu, X. Li, X. Yuan, Y. Su, Z. Deng, Y. Peng and M. Shen, Adv. Energy Sustainability Res., 2022, 3, 2100134.
51. M. N. I. Salehmin, S. K. Tiong, H. Mohamed, D. A. Umar, K. L. Yu, H. C. Ong, S. Nomanbhay and S. S. Lim, J. Energy Chem., 2024, 99, 223–252.
52. J. Ma, W. Zhong, L. You, Y. Pei, C. Lu, Z. Xiao, Z. Shen, X. Jiang, N. Qian, X. Liu and S. Zhang, Appl. Surf. Sci., 2022, 605, 154563.
53. A. Krishnan, M. Yoosuf, K. Archana, A.S. Arsha and A. Viswam, J. Energy Chem., 2023, 80, 562–583.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5nr02191j

‡ These authors contributed equally to this work.

---