Surface texture dependency of photocatalytic behavior of facile synthesized mesoporous ZnS-ZnO heterostructure under LED illumination

Simultaneous thermal analysis (STA) was employed to investigate the decomposition temperature and phase transformation of the synthesized ZnS. Figure 2-a demonstrates the thermal behavior of ZnS upon heating up to 1000 °C in an air atmosphere. Analysis of the thermogravimetric (TG) curve reveals approximately a 10% weight loss below 200 °C, which is associated with water elimination. This observation is supported by a gentle endothermic peak in the differential thermal analysis (DTA) curve^32,33. Additionally, a weight change of about 16% is expected during the complete conversion of ZnS to ZnO, as shown in the TG curve in the temperature range of 550–700 °C³⁴. The oxidation mechanism involves the release of sulfur species-typically in the form of SO₂-and the concurrent formation of ZnO as follows:

Initially, oxygen molecules adsorb onto the surface of ZnS particles. At elevated temperatures, these O₂ molecules dissociate into reactive atomic oxygen species (O•), which can attack the Zn-S bonds. Mechanistically, intermediate species such as surface Zn-O-S or Zn-SOₓ compounds may form before complete desulfurization³⁵. Consistent with this mechanism, the DTA curve reveals an exothermic peak around 640 °C, associated with the oxidation of ZnS. A subsequent decline between 640 and 680 °C indicates the complete conversion of ZnS to ZnO^34,36. Beyond 700 °C, no significant weight loss is observed in the TG curve. The DTA peak observed around this temperature may be attributed to the release of residual sulfur species³⁷. According to the thermal analysis result in Fig. 2-a, the oxidation temperature of ZnS is determined to be about 550 °C. Despite the fact that this oxidation is thermodynamically possible at lower temperatures, the kinetic barrier delays the reaction until higher temperatures are reached. Therefore, annealing treatments were performed at temperatures both below and above 550 °C, and the X-ray diffraction patterns of the obtained particles are presented in Fig. 2-b³⁸. As illustrated in Fig. 2-b, the formation of ZnS and ZnO phases is confirmed by the characteristic diffraction peaks. Specifically, the (111), (220), and (311) crystal planes of cubic ZnS (JCPDS 36-1450) are observed at 28.59°, 47.62°, and 56.47°, respectively. Additionally, peaks at 31.83°, 34.48°, 36.32°, 47.62°, 56.76°, 62.93°, and 67.86° correspond to the (100), (002), (101), (102), (110), (103), and (112) planes of hexagonal wurtzite ZnO (JCPDS No. 36-1451)^39,40,41,42.

According to Fig. 2-b, only ZnS peaks are observed up to 400 °C, confirming that no oxidation occurs at this stage-consistent with STA data. Although the conversion of ZnS to ZnO is thermodynamically feasible at lower temperatures, it is kinetically restricted. The increasing sharpness of ZnO peaks with increasing temperature indicates improved crystallinity^37,43, with the most intense peak corresponding to the (101) plane, suggesting preferential growth along this orientation⁴⁴. ZnO diffraction peaks gradually appear from 500 °C onward, while ZnS peak intensities diminish, indicating that oxidation begins at this point and completes around 650 °C. This observation also agrees with the STA data. At intermediate temperatures, the simultaneous presence of ZnS and ZnO diffraction patterns suggests the formation of ZnS-ZnO heterostructures.

Furthermore, The crystallite size of the synthesized specimens was calculated using the Debye-Scherrer equation⁴¹:

Where:

• D represents the crystallite size in nm.

• K is the shape factor, typically assumed to be 0.9.

• λ denotes the wavelength of the Cu-K_α radiation, which is 1.54 Å.

• β represents the full width at half maximum intensity.

• θ is the Bragg angle.

The calculated crystallite sizes are presented in Table S1 (Supplementary Materials). As the annealing temperature increases, the crystallite size of ZnO and ZnS increases, which can impact the photocatalytic performance^45,46,47.

FESEM analysis was performed to investigate the morphology of the synthesized compounds. In Fig. 3-a, ZnS particles are observed to be clustered, exhibiting a lumpy appearance. The high surface-to-volume ratio of nanoparticles enhances their surface reactivity, contributing to this morphology⁴⁸. The particles are predominantly spherical, with an average size of 13 nm, as indicated by the particle size distribution histogram. Figure 3-b shows the surface morphology of the ZnS-ZnO nanocomposite, where the particals are unevenly agglomerated with an average size of 43 nm. Finally, Fig. 3-c displays ZnO nanoparticles with a relatively smooth surface. These particles appear larger than those of ZnS, which is attributed to the reduction of sulfur and the complete conversion of ZnS into ZnO⁴⁹. Additionally, each sample underwent energy dispersive X-ray spectroscopy (EDS) analysis to confirm their elemental composition, which are consistent with the corresponding XRD results. In Fig. 3-a, distinct peaks for zinc (Zn) and sulfur (S) confirm the presence of ZnS. In Fig. 3-b, the appearance of an oxygen (O) peak verifies the formation of ZnO in the nanocomposite. Furthermore, the decreasing intensity of sulfur in Fig. 3-c further supports the successful synthesis of ZnO.

HRTEM results depicted in Fig. 4, provide further insight into the structure of ZnS-ZnO nanocomposite. The visible contrast between bright and dark regions at the interface signifies the formation of two distinct phases, thereby confirming the successful creation of a composite material⁵⁰. ZnO nanoparticles are observed to be formed near the ZnS nanoparticles, with instances of structural aggregation in some regions. Analysis of the selected area electron diffraction pattern (SEDP) for ZnS-ZnO sample reveals a polycrystalline structure. The size of a single ZnS crystal is estimated to be approximately 6.5 nm, which is in good agreement with the crystallite size determined from XRD results. Each bright point and ring in the SEDP corresponds to the interplanar distances of ZnS and ZnO, indicating the crystalline structure of both materials. Notably, the lattice of ZnS and ZnO exhibits distinct characteristics, confirming their crystallization⁵¹. Moreover, Further analysis of the fast Fourier transform (FFT) diffraction pattern shows specific d-spacings values: 0.31 nm corresponding to the (111) plane of ZnS and 0.24 nm for the (101) plane of ZnO⁴³. These observations confirm the successful formation of ZnS-ZnO heterojunctions.

FTIR spectra provide valuable insights into the chemical structure and any changes due to molecular interactions, particularly in composite materials⁵². In Fig. 5, the characteristic functional groups of ZnS, ZnO, and the ZnS-ZnO composite are identified. The peaks observed at 656 cm⁻¹ and 1005 cm⁻¹ in the spectra corresponds to sulfides and the Zn-S bond, respectively, which are characteristic of ZnS. In contrast, the peak at 621 cm⁻¹ in the ZnO spectrum is linked to Zn-O stretching vibrations^53,54. In the ZnS-ZnO composite spectrum, peaks appearing in the range of 550–900 cm⁻¹ indicate the presence of both Zn-O and Zn-S bonds. Additionally, the bond associated with Zn-S in the 1050–1150 cm⁻¹ range confirms that ZnS remains in the composite^55,56. Importantly, the absence of Zn-S peaks in the pure ZnO spectrum helps to the distinguish between ZnS and ZnO phases. A peak near 1122 cm⁻¹, observed in both the ZnS and ZnO spectra, undergoes a shift in the ZnS-ZnO spectrum, suggesting changes in the bonding environment due to composite formation. This band is associated with C-O stretching vibrations^57,58. A weak band near 1385 cm⁻¹ is attributed to N-O vibrations, likely arising from residual NaNO₃ by-products, indicating the persistence of trace secondary phases despite washing steps. Additionally, peaks at 2853 cm⁻¹ and 2922 cm⁻¹, corresponding to C-H stretching modes, are consistently present across all ZnS, ZnO, and ZnS-ZnO samples^52,54. Moreover, ZnS, ZnO, and ZnS-ZnO spectra showed peak levels at 1618 cm⁻¹, 1620 cm⁻¹, and 1616 cm⁻¹,respectively, are attributed to surface-adsorbed water (H_₂O bending vibration). The reduction in intensity of these bands after heating confirms the partial removal of physically adsorbed water^56,59. Finally, the broad absorption bands in the 3000–3600 cm⁻¹ range correspond to O-H stretching vibrations. Strong peaks are observed at 3405 cm⁻¹ for ZnS, 3422 cm⁻¹ for ZnO, and 3417 cm⁻¹ for the ZnS-ZnO composite, further supporting the presence of hydroxyl groups in all samples^51,60.

Adsorption and desorption of N₂ analyses were conducted to evaluate the porous nature and determine the specific surface area of the synthesized particles. The results are shown in Fig. 6 and were analyzed using the BET (Brunauer-Emmett-Teller) and BJH (Barrett-Joyner-Halenda) methods. The specific surface areas were determined to be 165 m² g⁻¹, 35 m² g⁻¹, and 10 m² g⁻¹ for ZnS, ZnS-ZnO, and ZnO, respectively. Notably, ZnS exhibits the highest specific surface area, indicating the presence of mesoporous structures⁶¹. This significant surface area is particularly remarkable compared to typical hydrothermal synthesis conditions. Given that the molar volume of ZnS (24.48 cm³ mol⁻¹) is greater than that of ZnO (14.5 cm³ mol⁻¹), an increase in porosity was anticipated upon ZnS oxidation. However, ZnO particles are larger than ZnS particles, potentially contributing to the blocking or collapse of pores and the subsequent decrease in the specific surface area⁶². Furthermore, the conversion of micropores into mesopores also contributes to the decline in surface area. From Fig. 6-a, ZnS exhibits type IV isotherms with an H₃ hysteresis loop, while ZnS-ZnO and ZnO display H₁ hysteresis loops. These loop types provide insight into the dominant pore geometries: ZnS primarily features narrow parallel-plate pores, ZnS-ZnO presents cylindrical pores open at both ends, and ZnO demonstrates a mixture of non-parallel and cylindrical planar pores^63,64. Furthermore, the BJH diagram (Fig. 6-b) indicates that the pore size distribution in ZnS-ZnO and ZnO is broader than that of ZnS. The average pore diameter for ZnS-ZnO is 23.5 nm, which is larger than those of ZnS (5.5 nm) and ZnO (13.6 nm). A broader hysteresis loop indicates an extensive range of pore sizes⁶⁴. However, ZnS retains highest pore volume, indicating that oxidation leads to the development of larger pores while maintaining some of the smaller ones inherent to ZnS. Micropore volumes, as represented in the MP diagram and detailed in Table 1, support these observations. ZnS demonstrates a higher micropore volume compared to ZnS-ZnO and ZnO. In comparing the distribution of porosity, it is determined that ZnO predominantly contains mesopores, with a limited volume of micropores. This structural characteristic contributes to its relatively lower specific surface area when compared to ZnS. These findings emphasize the importance of both micro and mesoporous structures in controlling photocatalytic performance. The coexistence of mesopores (which improve molecular diffusion and light scattering) and micropores (which increase surface area and adsorption potential) enables more efficient interaction between dye molecules and active sites. Therefore, the observed degradation behavior is closely linked to the hierarchical porosity of the synthesized materials, rather than surface area alone¹⁵.

The use of a non-solvent environment (ethanol) during the synthesis of ZnS contributes to its high specific surface area. However, the presence of the NaNO₃ by-product, which is insoluble in ethanol, requires washing with water to remove it. The removal of NaNO₃ is the main mechanism through which porosity is introduced into the samples. The XRD pattern of ZnS-NaNO₃ is provided in Figure S1 (Supplementary Materials). After washing, only ZnS peaks are observed, as shown in the XRD results of Fig. 2. The specific surface area of this sample is determined to be 9.26 m² g⁻¹ based on Figure S1-b (Supplementary Materials). The elimination of NaNO₃ from the ZnS-NaNO₃ induces porosity in the remained ZnS particles. A summary of all results obtained for the synthesized structures is presented in Table 1.

According to the diffuse reflectance spectroscopy (DRS) analysis in Fig. 7, the samples exhibit strong absorption mainly in the ultraviolet (UV) range (200–400 nm) and relatively weak absorption in the visible region (400–700 nm). The absorption edge for ZnS appears around 370 nm, while ZnS-ZnO and ZnO both exhibit absorption edges around 450 nm. Notably, ZnS shows higher absorption in the UV region, whereas the opposite trend is observed in the visible light region.

To calculate the bandgap energies of the samples, Tauc’s theory was applied using the following equation:

Here, α is the absorption coefficient, hν is the photon energy, E_g​ is the bandgap energy, A is a material-dependent constant, and n depends on the nature of the electronic transition. The Beer-Lambert law was used to determine the absorption coefficient⁶⁵:

Where T is the value of the transmitted beam, and t is the thickness of the used sample. By extrapolating the plots of (αhυ)^0.5-hυ, the bandgap energies of the ZnS, ZnS-ZnO, and ZnO were found to be 3.36 eV, 2.88 eV, and 2.96 eV, respectively.

To further investigate the electrochemical properties and energy band levels of the prepared semiconductors, the Mott-Schottky technique was used. The Mott-Schottky equation is as follows⁶⁶:

where C is the depletion layer capacitance, ε_r is the dielectric constant of the semiconductor, ε₀ ​is the vacuum permittivity, e is the initial electric charge, N_D is the donor density, E is the applied potential, E_FB ​is the flat band potential, k is Boltzmann’s constant, and T is the temperature.

The Mott-Schottky plots in Fig. 7(b-e) reveal n-type behavior for all three materials-ZnO, ZnS, and ZnS-ZnO-evidenced by the positive slopes in their respective plots. By extrapolating the linear portion of the curves, the flat band potentials were estimated and then converted to the normal hydrogen electrode (NHE) scale using the following equation:

It is important to note that the conduction band (CB) level of an n-type semiconductor is typically 0.1 V more negative than the flat band potential⁶⁶. Based on the conduction band energy and the previous calculated band gaps, the valance band (VB) energy levels were determined as shown in Table S2 (Supplementary Materials). Figure 7-e illustrates the energy levels of the semiconductors. Additionally, the charge carrier densities were calculated to be 9.21 × 10¹⁵ cm⁻³ for ZnS, 1.07 × 10¹⁶ cm⁻³ for ZnS-ZnO, and 2.58 × 10¹⁵ cm⁻³ for ZnO.

Photoluminescence (PL) analysis of the synthesized compounds provides valuable insights into charge generation and the behavior of electrons and holes. Figure 7(f) and (g) display the PL spectra of the three samples at excitation wavelengths of 300 nm and 350 nm, respectively. Emission in the 380–550 nm range is attributed to various defect states, including Zn²⁺ vacancies, S²⁻ vacancies, dislocations, and interstitials. the intensity of emission associated with Zn²⁺ and S²⁻ vacancies decreases upon oxidation, suggesting that charge carriers are trapped at defect sites⁴¹. An emission peak around 360 nm could be assigned to exciton recombination. The weak blue emission observed at approximately 450 nm may be due to zinc vacancies or surface states. For ZnS, the emission peak observed in the visible region (~ 480–520 nm) can be ascribed to sulfur vacancies and zinc interstitials⁶⁷. In ZnO, a sharp emission at ~ 380 nm is due to the corresponding recombination, while a broad visible emission centered around ~ 500–550 nm is attributed to deep-level or trap-state emissions, commonly linked to intrinsic defects such as oxygen vacancies, zinc interstitials, and oxygen interstitials. Following oxidation, an increase in ultraviolet emission is observed, which may be due to ZnS-ZnO facilitating the tunneling or transportation of charge carriers from ZnS to ZnO. This process leads to a higher number of photogenerated electrons and holes being confined inside the ZnO⁶⁸. Furthermore, The green emission observed at around 500 nm in ZnO is attributed to profound levels like oxygen vacancies or surface states⁶⁹.

The photocatalytic potential was evaluated by monitoring the decomposition efficiency of methylene blue, an organic contaminant with a maximum absorption peak at 663 nm. As illustrated in Fig. 8(a-c), the absorption spectrum of methylene blue displays a distinct peak that decreases over time, indicating photocatalytic degradation. Figure 8-d displays the time-dependent changes in methylene blue concentration, reflecting the ability to absorb and the level of photocatalytic performance. To further quantify the degradation behavior, pseudo-first-order kinetic analysis was performed. The linear relationship of ln (C₀/C) versus time provided rate constants (k) and correlation coefficients (R²) for each sample. Among the three photocatalysts, ZnS showed the highest rate constant (k = 0.005 min⁻¹, R² = 0.66), confirming its superior photocatalytic activity under visible light. ZnO and ZnS-ZnO also followed pseudo-first-order kinetics with a strong linear correlation. This kinetic evaluation supports the observed trends in dye removal efficiency. The decolorization efficiency (η) can be calculated using the following equation:

where C₀ represents the initial concentration and C denotes the concentration following irradiation.

Figure 8-e illustrates the efficiency of color removal, with ZnS demonstrating the highest efficiency. ZnS-ZnO follows, achieving 55% decomposition, while ZnO exhibits the lowest efficiency, removing only 43% of the organic pollutants. Furthermore, Fig. 8-f presents the results of the absorption percentage of methylene blue in the dark. In dark status, ZnS shows higher absorption rate compared to the other samples within 30 min.

A collective interpretation of various analyses, including XRD, UV-Vis, FESEM, and. confirms the successful formation of crystalline, nanoscale semiconductors with a defined morphology and optical activity. While our synthesis does not involve phytochemical agents, the obtained structural and optical features are comparable to those reported in biogenic synthesis routes^70,71.

Process mechanism

The degradation mechanism of methylene blue is schematically demonstrated in Fig. 9. As shown in Table 2, the superior photocatalytic activity of ZnS compared to ZnS-ZnO can be attributed to several factors, despite ZnS having a larger bandgap, fewer charge carriers, and a higher potential for electron-hole recombination. It is important to note that the photocatalytic tests were conducted under a cool-white LED light source (50 W, Brux, China), which emits primarily in the visible light range. One significant factor handling the present result is the high porosity of ZnS, which enhances light scattering and increases the degradation of pollutants. The presence of pores in ZnS particles provides additional active sites for adsorption and catalytic reactions, thereby enhancing overall methylene blue removal activity⁷². Additionally, porosity in ZnS can modulate its electronic properties by including its band structure and energy levels. These surface states can facilitate charge separation and reduce recombination, thus positively influencing photocatalytic activity.

The interconnected pores in the ZnS structure form continuous channels that facilitate the migration of photogenerated charge carriers (electrons and holes), reducing the diffusion length to active sites and increasing their lifetime. This structural feature enhances the probability of charge carriers reaching the catalytic interface before recombination, a phenomenon referred to as dynamic photocatalytic degradation. Moreover, ZnS exhibits a higher adsorption capacity under dark conditions compared to ZnS-ZnO and ZnO, due to its larger specific surface area (as shown in Fig. 8-b). When the lamp is turned on during the photocatalytic treatment, both the production of charge carriers and the adsorption of pollutants continue. This dual mechanism is schematically illustrated in Fig. 9 (a and b). In the initial stage (darkness), methylene blue is adsorbed onto the high surface area of ZnS particles, leading to a reduction in pollutant concentration. Subsequently, when the lamp is turned on, photons reach the ZnS surface, initiating the degradation (decomposition) of methylene blue most likely to CO₂ and H₂O^73,74. This process refreshes the surface for the re-adsorption of pollutants, effectively enhancing the photocatalytic efficiency of ZnS.

The inter-band energy levels in the ZnS structure can greatly influence its photocatalytic efficiency. In the ZnS lattice, the existence of defects can create sub-bands energy states located below the conduction band, thereby broadening the light absorption range and facilitating more effective photoexcitation. These additional energy states increase the number of excited electrons and holes available to participate in redox reactions, particularly in the oxidation of organic pollutants, thereby improving the overall photocatalytic activity. Defects within the ZnS lattice, such as S vacancies (V_S), Zn vacancies (V_Zn), and interstitial atoms of Zn and S (I_Zn and I_S), play a crucial role in this process. As illustrated in Fig. 9-c, the schematic energy level diagram shows how these four types of defects contribute to the electronic structure of ZnS. Sulfur vacancies and interstitial Zn atoms act as donor-like charge states, while Zinc vacancies and interstitial Sulfur atoms act as acceptor states.

Photoluminescence analysis supports this interpretation, with emission peaks in the blue (430–470 nm) and green (510–550 nm) regions. The blue emission is associated with sulfur vacancies, interstitial Zn atoms, and surface states, while the green emission originates from electron transitions between sulfur vacancy levels and intermediate sulfur states, followed by recombination. These defects not only modify the electronic properties of ZnS but also act as active sites that enhance the adsorption of reactant molecules and improve the separation of photogenerated electron-hole pairs. This suppression of charge carrier recombination extends their lifetimes and increases their participation in surface redox reactions, which is crucial for efficient photocatalysis. In particular, sulfur vacancies are known to reduce the energy barrier for interfacial charge transfer, thereby facilitating more efficient electron transport from the ZnS surface to adsorbed target molecules. These vacancies serve as electron “bridges,” effectively lowering the activation energy required for photocatalytic degradation reactions and enhancing overall efficiency^69,75,76.

Finally, it is important to note that the light source-an LED lamp in the present study-also significantly influences the photodegradation performance of these three semiconductors. Selection of the light source the spectra of which can excite electrons from valence to conduction band, which means the energy of the photons are equal or greater than the band gap of the synthesized particles, results in substantial difference of their photodegradation performance⁷⁷.

In addition to the surface- and defect-related mechanisms discussed earlier, it is worth mentioning the possible contribution of a dye-sensitization mechanism under visible light irradiation. ZnS, due to its wide band gap, does not intrinsically absorb visible light efficiently. However, methylene blue, which strongly absorbs in the visible region, may act as a photosensitizer. Upon excitation by visible light, methylene blue molecules can inject electrons into the conduction band of ZnS, initiating photocatalytic degradation via reactive oxygen species formation. This sensitization mechanism has been previously reported in similar systems where dye molecules extend the photoresponse of wide-band-gap semiconductors into the visible range through interfacial charge transfer⁷⁸. While this effect was not the focus of this study, it may partially contribute to the unexpectedly high degradation efficiency observed for porous ZnS under visible LED illumination.

In the ZnS-ZnO structure illustrated in Fig. 10, photoexcitation by incident photons promotes electrons from the valence band of ZnS to its conduction band, leaving behind holes. A similar excitation process occurs in ZnO. Due to the lower conduction band potential of ZnS relative to ZnO, electrons tend to migrate from ZnS to ZnO, establishing Fermi level equilibrium and generating an internal electric field across the heterojunction. This electric field facilitates the transfer of holes from the valence band of ZnS to that of ZnO, thereby enhancing charge separation, particularly on the ZnS side. However, the transfer of electrons from ZnS to ZnO is limited by energy barriers due to the differences in conduction and valence band energies between these two compounds. Thus, this heterojunction primarily facilitates the spatial separation of photogenerated charge carriers on the ZnS side, while contributing only modestly to charge separation in ZnO. Consequently, electron-hole pairs are not fully separated across both semiconductors, and recombination processes-particularly in ZnO-still occur.

Contrary to many previous studies that reported superior photocatalytic activity for ZnS-ZnO heterostructures compared to the individual components (Table 3), our experimental results show that pure porous ZnS outperforms the ZnS-ZnO composite in degrading methylene blue. While ZnS-ZnO structures are generally favored due to efficient heterojunction-mediated charge separation, our findings suggest that specific surface area is a more critical factor than previously assumed. In our study, electrons from the conduction bands of both ZnO and ZnS, as well as holes in the valence band of ZnO, contribute to the photocatalytic reaction. However, due to the lower specific surface area of ZnS-ZnO, fewer pores are exposed to methylene blue adsorption compared to ZnS, resulting in the decreased photocatalytic efficiency. This deviation between literature and our research can also be related to the interface mismatch between ZnS and ZnO, in addition to the specific surface area discussed above. This interface mismatch may hinder carrier movement and instead act as a site for electron-hole recombination, thereby reducing efficiency. Additionally, most charge recombination occurs in ZnO, further contributing to the lower removal efficiency compared to the other two compounds^79,80,81.