Introduction

The sustainable progress of mankind has been significantly threatened since the Industrial Revolution by environmental pollution issues and energy scarcity (Wang et al. 2020a; Taifi et al. 2022; Mahdi et al. 2021; Aljeboree et al. 2021; Ganduh et al. 2021). Photocatalysis is considered the most viable solution to these issues because of its many good qualities, including environmental friendliness, affordability, and, efficiency. In 1972, titanium dioxide (TiO2) was utilized for the first time to split water photoelectrochemically to create H2 (Fujishima &Honda 1972). TiO2 has been the subject of a substantial amount of study to date. However, visible light, which accounts for 43% of the energy in the solar spectrum, cannot excite TiO2, a semiconductor with a wide bandgap; this severely restricts the vast range of applications for TiO2 (Xu et al. 2020, 2018).

Creating photocatalysts with benign visible light responses is essential for efficiently using sunshine. The fundamental objective is to synthesize photocatalysts with a visible light response to maximize solar efficiency. Several visible light-active photocatalysts, including Bi2WO6 (Wang et al. 2021), CdS (Li et al. 2019), and g-C3N4 (Zhang et al. 2021a), etc., have been developed and employed in redox processes.

Graphitic carbon nitride (g‐C3N4) with its usual 2D network has been a research hotspot due to its great qualities, for instance, strong thermal-chemical stability non-toxicity, adequate energy band, available raw materials, and facile preparation (Ren et al. 2021). g-C3N4 is commonly utilized for CO2 reduction (Huo et al. 2019), H2 evolution (Xiao et al. 2019b), bacterial inactivation, organic pollutant degradation (Wang et al. 2020b; Xia et al. 2020), and NOx elimination (Xia et al. 2020). The typical thermal polymerization approach produces g‐C3N4 with large grain boundaries due to the high-temperature process, which hinders its photocatalytic activity.

Several strategies have been used to increase its activity, including heterojunction fabrication (Zhang et al. 2021b), noble metal deposition (Xiao et al. 2019a), and nonmetal doping (Zhao et al. 2019). A heterostructure photocatalyst is ideal for efficiently separating photoinduced electrons and holes, resulting in increased photocatalytic activity (Lu et al. 2020). Wang et al. (Wang et al. 2017) used a bottom-up approach to create an ultrathin g-C3N4/Bi2WO6 composite. The ibuprofen degradation efficiency of 25% g-C3N4/Bi2WO6 was 96%. Zhang et al. (Wang et al. 2020b) manufactured 2D/2D BaNb4O15/g‐C3N4 photocatalysts using calcine technique, and 1:20‐BaNb4O15/g‐C3N4 obtained the greatest hydrogen production rate of 2.67 mmol g–1 h–1. In most cases, g-C3N4-based heterojunctions use the classic type-II transfer mechanism of photoinduced charge carriers.

While a type-II mechanism can efficiently separate charge carriers, it is important to note that trapped electrons and holes might negatively impact photocatalytic activity due to kinetics and thermodynamics (He et al. 2020). Yu et al. (Fu et al. 2019) suggested a step-scheme charge transfer technique to show the enhance function of the WO3/g-C3N4, addressing kinetic and thermodynamic issues with classic Z-scheme charge transfer. Photocatalyst I (PC I) and Photocatalyst II (PC II) are n-type semiconductors typically found in an S-scheme heterojunction (Wang et al. 2020c). The low Fermi level of Photocatalyst I makes it an oxidation-type photocatalyst. In comparison, photocatalyst II has an elevated Fermi level and is a reduction-type photocatalyst. Electrons move from PC II to PC I over an intimate interface and reach identical Fermi levels, causing a built-in electric field (Xie et al. 2020). At the contact, holes in photocatalyst II’s valence band (VB) and electrons in photocatalyst I’s conduction band (CB) recombine due to the action of the internal electric field (Luo et al. 2020). Strong redox-capable holes and electrons will be retained in PC I’s VB and photocatalyst II’s CB, respectively (Ren et al. 2020). In conclusion, S-scheme heterojunctions can store electron–hole pairs with strong redox capacity in addition to increasing the spatial charge separation (Peng et al. 2021). S-scheme TiO2/CsPbBr3 has recently been shown by Xu et al. to display outstanding activity in photocatalytic CO2 reduction (Xu et al. 2020). Therefore, it is necessary to choose another photocatalyst with an appropriate energy band shape in order to properly create the S-scheme heterojunction.

Copper manganese oxide (CuMn2O4) is a fascinating compound belonging to the family of spinel oxides, which have garnered significant attention in materials science and condensed matter physics due to their diverse properties and potential applications (Sobhani-Nasab et al. 2020). CuMn2O4, specifically, exhibits a rich array of electronic, magnetic, and catalytic behaviors, making it an intriguing subject of study across various scientific disciplines. At its core, CuMn2O4 possesses a crystal structure characterized by a cubic spinel framework, where copper (Cu), manganese (Mn), and oxygen (O) ions are arranged in a well-defined lattice (Afriani et al. 2018). This spinel structure consists of alternating layers of metal cations and oxygen ions, with manganese ions occupying both tetrahedral and octahedral sites within the lattice, while copper ions primarily occupy octahedral sites (Gao et al. 2021). This unique arrangement of ions gives rise to a myriad of interesting physical phenomena, including magnetic ordering, charge transport, and catalytic activity. One of the most compelling aspects of CuMn2O4 is its magnetic properties. Depending on the oxidation state and arrangement of the transition metal ions, CuMn2O4 can exhibit various magnetic behaviors, ranging from antiferromagnetic to ferrimagnetic or even ferromagnetic ordering (Parida et al. 2016). This tunable magnetism makes CuMn2O4 an attractive candidate for applications in spintronics, magnetic storage devices, and magneto-optical devices (Afriani et al. 2018). Furthermore, CuMn2O4 has garnered attention for its potential as a multifunctional material with applications in energy conversion and storage. Its unique electronic structure and catalytic activity make it promising for use as an electrode material in rechargeable batteries, electrocatalysts for water splitting, and photocatalysts for solar fuel production. Additionally, CuMn2O4-based materials have shown promise in environmental remediation applications, including the removal of organic contaminants in wastewater and air purification processes (Pramothkumar et al. 2019; Sobhani 2022).

CuMn2O4 and g-C3N4 are promising materials photocatalysis due to their unique properties and elevated efficiency in degrading organic pollutants. CuMn2O4 is a spinel oxide with a wide bandgap, excellent stability, and good photocatalytic activity (Zhao et al. 2013). On the other hand, g-C3N4 is a metal-free semiconductor with high chemical stability and a suitable bandgap for the absorption of visible light. When combined as composites, CuMn2O4/g-C3N4 exhibit significant effects that enhance their photocatalytic performance for dye degradation (Venkatesh et al. 2022). The mixture of CuMn2O4 and g-C3N4 can enhance light absorption, charge separation, and surface area, leading to increased efficiency in degrading organic dyes.

Considering the above-mentioned ideas, an innovative CuMn2O4/g-C3N4 S-scheme heterojunction was effectively created using an ultrasonic-assisted co-precipitation approach. These composites have been used to degrade various toxic pollutants, for example, methyl violet (MV), erythrosine (ER), and eriochrome black T (ECBT), under visible irradiation. The improved photocatalytic movement of CuMn2O4/g-C3N4 makes them a potential candidate for wastewater treatment and environmental remediation. The S-scheme charge transfer mechanism was also thoroughly studied using a variety of characterizations, including band edge position estimates and active species trapping studies.

Materials and methods

Materials

Melamine 99.9% (C3H6N6), manganese nitrate tetrahydrate 98.0% (Mn(NO3)2.4H2O), copper (II) nitrate trihydrate, 99% (Cu(NO3)2.3H2O), 1,4‒benzoquinone (BQ), lactose, ethylenediaminetetraacetic acid (EDTA), eriochrome black‒T (ECBT), benzoic acid (BA), erythrosine (ER), and methyl violet (MV), were conveyed from Sigma-Aldrich and used as-received.

Preparation of CuMn 2 O 4

1 mmol of Cu(NO3)2 and 2 mmol Mn(NO3)2 were separately dissolved in 10 mL of deionized water. The solution containing Cu(NO3)2 was added to Mn(NO3)2 solution and mixed for 25 min. Next, the mixture above was mixed with 30 mL of 1.0 M lactose solution added dropwise. Then, the temperature of solution was raised up to 110 ℃ until the gel was formed. The gel was calcined at 900 °C for 5 h after being dried at 65 °C for 12 h. The following reactions can occur:

$${\text{Cu}}\left( {{\text{NO}}_{{3}} } \right)_{{2}} \, \to \,{\text{CuO}}\, + \,{\text{2NO}}_{{2}} + {\text{O}}_{{2}}$$
(1)
$${\text{Mn}}\left( {{\text{NO}}_{{3}} } \right)_{{2}} \, \to \,{\text{MnO}}\, + \,{\text{2NO}}_{{2}} + {\text{O}}_{{2}}$$
(2)
$${\text{MnO }} + {\text{O}}_{{2}} \, \to \,{\text{Mn}}_{{2}} {\text{O}}_{{3}}$$
(3)
$${\text{Mn}}_{{2}} {\text{O}}_{{3}} + {\text{O}}_{{2}} \, \to \,{\text{2MnO}}_{{2}}$$
(4)
$${\text{2MnO}}_{{2}} \, + \,{\text{CuO}}\, \to \,{\text{CuMn}}_{{2}} {\text{O}}_{{4}} + {\text{O}}_{{2}}$$
(5)

Preparation of CuMn2O4/g-C3N4

The first step included creating g-C3N4 nanosheets, which we discussed in our earlier study (Ghanbari &Salavati-Niasari 2021). Following the instructions, a 550 °C direct heat treatment was applied to 12.0 g of melamine for 4 h. We collected the yellow g-C3N4 powder for further analysis. Consequently, 25 mL of water was subjected to an equal distribution of 100 mg of g-C3N4 using ultrasonic waves for 15 min. After that, the g-C3N4 dispersion was stirred for a day and five successive additions of 10, 20, 40, 80, and 100 mg of CuMn2O4 were made. Centrifugation and drying at 65 °C for a day were employed for the deposition. The following reactions are suggested:

$${\text{C}}_{{3}} {\text{H}}_{{6}} {\text{N}}_{{6}} \, \to \,{\text{C}}_{{3}} {\text{H}}_{{3}} {\text{N}}_{{5}} \, + \,{\text{NH}}_{{3}}$$
(6)
$${\text{C}}_{{3}} {\text{H}}_{{3}} {\text{N}}_{{5}} \, \to \,{\text{C}}_{{3}} {\text{N}}_{{4}} \, + \,{\text{NH}}_{{3}}$$
(7)
$${\text{CuMn}}{\text{O}}\, \left( {{\text{particles}}} \right)\, + \,{\text{g}} - {\text{C}}{\text{N}}\, \left( {{\text{surface}}} \right)\, \to \,{\text{CuMn}}{\text{O}}/{\text{g}} - {\text{C}}{\text{N}}\, \left( {{\text{composite}}} \right)$$
( 8)

Photocatalytic process

Pure CuMn2O4 and its various contents, in combination with g-C3N4, were used as photocatalysts because these compounds have a high potential for degrading ECBT, MV, and ER when subjected to visible radiation. The photocatalytic operation employed a 400-W Osram lamp as a radiation source. This lamp emits around 400 to 700 nm. Without a catalyst and light, hardly any dye had been degraded after 90 min. 50 mg of various materials (for example, CuMn2O4 and CuMn2O4/g-C3N4 at differing CuMn2O4 concentrations) were mixed with 50 mL of 10 ppm ER solutions in each test experiment. It is essential to note that 30 min before radiation, ER and catalyst were mixed in the dark. Particles are removed from the suspension every 15 min by taking a 3 mL sample, which is then centrifuged for 2 min at 15,000 rpm. The absorbance of the floats was determined using a UV–VIS spectrophotometer. The following is the computation of the degradation percentage (%D). The aqueous solution absorption at 0 and t minute is donated by A0 and At, respectively (Hosseini et al. 2024b).

$$D(\% ) = \frac{{A_{0} - A_{t} }}{{A_{0} }} \times 100$$
(9)

Characterization of materials

To confirm the kind of structure and purity of the as-synthesized nanoparticles, X-ray diffraction (XRD) patterns were collected using a Philips diffractometer equipped with X’PertPro monochromatized Cu K radiation (λ = 1.54 Å). The FTIR spectra of the material were acquired in the 400–4000 cm‒1 range using the Shimadzu FTIR-8300 E spectrophotometer. Field emission scanning electron microscopy (FESEM MIRA3 TESCAN) fitted with energy dispersive spectrometry (EDS) was used to analyze the distribution and form of the nanoparticles. Utilizing a JEM-2100 transmission electron microscope, the nanoparticles were examined. The UV–visible spectrum was acquired using a Jasco V-670 spectrophotometer, and the bandgap was determined using a diffuse reflectance spectrophotometer (Model: HO-SP-DRS100). N2 adsorption at –196 ˚C was used to define the surface areas (BET) using an automated gas adsorption analysis system (Tristar 3000, Micromeritics).

Results and discussion

Material description

The XRD technique is a successful approach for studying the structural properties of the products. The cubic crystal structure of the copper manganese oxide is confirmed by its presence in pure, in accordance with JCPDS No. 01–074-2422 (Fig. 1a). Because of special properties of cubic structure, this crystal structure can be used in a variety of electrical and energy-storage applications. The (111), (220), (311), (222), (400), (422), (511), (440), and (622) planes of CuMn2O4 are responsible for the diffraction peaks located at 2θ = 18.5°, 30.4°, 35.8°, 37.5°, 43.6°, 54.1°, 57.6°, 63.3°, and 71.8°, respectively. Figure 1b shows the graphitic carbon nitride XRD pattern. This pattern matches the hexagonal structures of 01–087-1526 C3N4 quite well. The two distinct peaks located at 13.1° and 27°, respectively, fit the (001) and (002) planes. Further details on the crystalline phases and composition of the CuMn2O4/g-C3N4 nanocomposite (Fig. 1c–g) are given by the XRD patterns. The effective synthesis of the nanocomposites is shown by the appearance of peaks corresponding to both CuMn2O4 (01‒074‒2422) and g-C3N4 (01‒087‒1526). Notably, when the amount of CuMn2O4 in the composition increases, there is a noticeable decrease in the intensity peaks of and g-C3N4.

Fig. 1
figure 1

XRD patterns of as-prepared samples

Full size image

Table 1 indicates crystallite size of the samples calculating by Scherrer and Williamson–Hall methods.

$$D = \frac{K\lambda }{{\beta \cos \theta }}$$
(10)
Table 1 Microstrain and crystallite size of all samples
Full size table

In Eq. (2), λ represents the X-ray wavelength (1.54 Å), “β” the FWHM value, and K is the Bragg constant (0.9). Also, crystallite size was determined using Williamson–Hall plots to investigate microstructural characteristics. By graphing βcosθ vs sinθ, we can determine the crystallite size (Kλ/D = intercept) and strain component from the slope (Samimi &Ghiyasiyan-Arani 2024). The results in Table 1 show that for g-C3N4 and CuMn2O4/g-C3N4 (1:1), nonzero residual tension causes the crystallite size obtained by the Scherer equation to be less than the crystallite size indicated by the Williamson–Hall plots. Figure 2 depicts the practically zero slope of other samples. This validates the Williamson–Hall formulas and the Scherrer equation’s results with the same crystallite size.

Fig. 2
figure 2

Williamson–Hall plot for synthesized samples

Full size image

As seen in Fig. 3, a Fourier transform infrared spectrometer verifies that metal‒oxygen (M‒O) bonds have formed in CuMn2O4 composition. The wavelength range that is being studied for CuMn2O4 is 400 to 4000 cm‒1. The M‒O bond’s low-frequency oscillations can be seen below 1000 cm‒1. Figure 3a shows the Cu–O and Mn–O stretching vibrational modes as two firm peaks at locations 700 and 528 cm‒1, respectively. These are the main indications that CuMn2O4 is formed (Parida et al. 2016). The stretching vibration of –OH is identified as a band that emerged in the range of 1630 and 3520 cm‒1 (Zhang et al. 2021c). The band at 3156 cm−1 is related to free amino groups (N–H) for g-C3N4 (Fig. 3b). The C–N stretching vibration is associated with the peak at 1645 cm−1, whereas the C–N heterocycle stretching vibration modes are linked to the ones at 1563 and 1410 cm−1 (Zhu et al. 2015). It is further confirmed that the bands at 1331 and 1245 cm−1 coordinate with the vibration of linked units of C − NH − C by the wide band at 3156 cm−1 for hydrogen bonding interactions. Additionally, the tri-s-triazine rings are described by the strong signal at 812 cm−1 (Yang et al. 2016). The FTIR spectra of the CuMn2O4/g-C3N4, which resemble the g-C3N4 spectrum, are shown in Figs. 3c-3g. This implies that the distinctive peaks of g-C3N4 are unaffected by the addition of CuMn2O4 to the g-C3N4 substrate. It also means that the CuMn2O4 nanoparticles have dispersed uniformly throughout the g-C3N4 matrix without causing any structural changes to the structure. This is an important finding since it demonstrates that the structural integrity of g-C3N4 nanosheets is not negatively impacted by the addition of CuMn2O4.

Fig. 3
figure 3

FTIR spectra of as-prepared samples

Full size image

Figure 4a shows the surface morphology of CuMn2O4, which is formed of microstructures and, in certain places, nanoparticles. It is evident that the majority of the CuMn2O4 particles aggregate together and exhibit irregular shapes. Moreover, the CuMn2O4 particles have an excellent capacity for dye adsorption because they are fused and highly linked at the same time. Figure 4b clearly shows that exfoliated g-C3N4 forms lamellar-shaped 2D nanosheets with a smooth surface. Figure 4(c-g) depicts the morphology of CuMn2O4/g-C3N4 with different contents of CuMn2O4 and shows that even after CuMn2O4 are loaded, g-C3N4 maintains its lamellar sheet-like structure. Over the g-C3N4 nanosheets, CuMn2O4 is observed as tiny, aggregated formations. This suggests that g-C3N4 and CuMn2O4 create a 2D nanocomposite. In other words, CuMn2O4 nanoparticles are loaded on the outer layer of g-C3N4 following their hybridization with g-C3N4. This good heterojunction is formed, which is advantageous for the transmission of photoinduced electron–hole pairs and increases photocatalytic performance.

Fig. 4
figure 4

FESEM images of as-prepared samples

Full size image

The elemental mapping and EDX analysis of CuMn2O4, g-C3N4, and CuMn2O4/g-C3N4 are depicted in Fig. 5, which validates the formation of the CuMn2O4/g-C3N4 composite and the presence of the appropriate elements (Cu, C, O, Mn, and N) in the spectra. Additionally, the elemental mapping of CuMn2O4/g-C3N4 (0.2:1) demonstrates the homogeneity of the Cu, C, O, Mn, and N element distribution, confirming the presence of CuMn2O4 with carbon nitride catalyst.

Fig. 5
figure 5

Elemental mapping and EDX analysis and of CuMn2O4, g-C3N4, and CuMn2O4/g-C3N4 (0.2:1)

Full size image

The HRTEM images of CuMn2O4/g-C3N4 (0.2:1) nanocomposite are depicted in Fig. 6. The parallel lines that divided the crystal surfaces in Fig. 6b demonstrated the remarkable crystallinity of CuMn2O4. The cubic CuMn2O4 crystal planes (311), which have an interplanar spacing of 2.50 Å, are well suited.

Fig. 6
figure 6

HRTEM images of CuMn2O4/g-C3N4 (0.2:1)

Full size image

The N2 isotherms and pore size distribution curves of CuMn2O4/g-C3N4 (0.1:1) are shown in Fig. 7a and 7b. The isotherm agrees with H3 hysteresis and the IV-type isotherm. At increasing pressures, the adsorption capacity of this isotherm gradually increases after a plateau zone and a sharp initial slope. The adsorbate molecules are quickly occupying the available surface area of the adsorbent material, as evidenced by the steep initial slope. When adsorbate molecules reach a saturation point, they can no longer form new monolayers on the surface of the adsorbent material, as shown by the plateau area. Multilayer adsorption is evident at higher pressures due to the progressive rise in adsorption capacity. In contrast, a particular kind of hysteresis loop that is seen in IV-type isotherms during BET analysis is referred to as H3 hysteresis. When the desorption and adsorption curves deviate significantly at high relative pressures, a hysteresis loop takes place. In particular, the related adsorption curve is moved to lower relative pressures than the desorption curve. H3 hysteresis may be caused by capillary condensation effects or pore obstruction in the adsorbent material’s porous structure. At high relative pressures, these factors may cause desorption to be inhibited or delayed. The BET data provides the average pore diameters at 9.2169 nm, the total pore volumes (p/p0 = 0.921) at 0.042748 cm3 g‒1, aBET = 18.552 m2 g‒1. The BJH plot (Fig. 7b) shows that the nanoparticle size distribution is within the range of 1 to 10 nm.

Fig. 7
figure 7

a N2 adsorption–desorption isotherm, b BJH plot of CuMn2O4/g-C3N4 (0.2:1), c DRS spectra and d) Tauc plots of CuMn2O4, g-C3N4, and CuMn2O4/g-C3N4 (0.2:1)

Full size image

The electron transition from the VB to the CB is frequently linked to optical absorption. Consequently, the optical bandgap was computed by DRS data. The best way to achieve this is via the Kubelka–Munk (K-M) function and Tauc plots. In a standard K-M formula, the following relation must be used to alter R-values (in%) to K-values, which are measures of the transformed R (Panahi et al. 2023).

$${\text{K}}\, = \,\left( {{1} - {\text{R}}} \right)^{{2}} /{\text{2R}}$$
( 11)

The sample’s bandgap can then be computed using the Tauc model:

$$\left( {\alpha {\text{h}}\upsilon } \right)\, = \,\beta \left( {{\text{hv}} - {\text{Eg}}} \right)^{{\text{n}}}$$
( 12)

The Planck constant is represented by h = 6.626 × 10−34 J s, and the light frequency is denoted by ʋ in Eq. (3). The absorption coefficient, α = ([2.303 × A]/d), depends on absorbance (A) and the sample thickness (d). Beta (β) is related to the absorbance. Depending on the kind of electrical transition, the value of “n” varies. For permitted and prohibited direct transitions, it uses 1/2 and 3/2; for permitted and prohibited indirect transitions, it uses values of 2 and 3 (Panahi et al. 2023). As seen in Fig. 7c, the bandgap value of pure CuMn2O4 and g-C3N4 was predicted to be 1.6 and 2.72 eV. Bandgap values for both CN and CuMn2O4 are shown at 2.4 and 2.7 eV in CuMn2O4/g-C3N4 (0.2:1) nanocomposites (Fig. 7d). The increase in the bandgap of CuMn2O4 when combined with g-C3N4 can be allocated to constructing an interface between the two materials. The dissimilarity in Fermi levels between two materials causes an integrated electric field to form at the interface when they come into contact. In this case, g-C3N4 has a higher bandgap than CuMn2O4. When they are combined, electrons from CuMn2O4 will flow to g-C3N4 until equilibrium is reached, resulting in a depletion region at the interface. This depletion region effectively increases the bandgap of CuMn2O4 to 2.4 eV. Additionally, the heterojunction can also lead to improved charge separation and transfer efficiency, which can enhance the overall performance of the material in applications such as photocatalysis or photovoltaics.

Photocatalytic efficiency

Figure 8a shows the breakdown of contaminants using various photocatalysts. The degradation efficiencies for erythrosine were 37.4% and 46.0%, respectively, when utilizing pristine CuMn2O4 and g-C3N4 photocatalysts. The CuMn2O4/g-C3N4 composites with mass ratios of 0.1:1 and 0.2:1 showed degradation efficiency of 63.9 and 87.1%, respectively. CuMn2O4/g- C3N4 composites (0.4:1, 0.8:1, and 1:1) degraded at 31.5, 1.5, and 22.2% after 90 min of visible light exposure, respectively. CuMn2O4/g-C3N4 (0.2:1) demonstrated the strongest photocatalytic activity. Furthermore, increasing the quantity of CuMn2O4 in nanocomposites reduces photocatalytic activity. Similarly, effective separation of photoinduced electron–hole pairs within the photocatalyst contributes significantly to the photocatalytic destruction of contaminants. Figure 8b reveals the effect of three different dyes, including erythrosine, eriochrome black T, and methyl violet over CuMn2O4/g-C3N4 (0.2:1) nanocomposite. As observed, CuMn2O4/g-C3N4 (0.2:1) degraded 87.1, 46.1 and 32.9% of the degradation of erythrosine, ECBT, and MV, respectively. Several concentrations of erythrosine (10, 15, and 20 ppm) were studied over CuMn2O4/g-C3N4 (0.2:1). Figure 8c illustrates the decline in 10, 15, and 20 ppm erythrosine, which was 87.1, 43.7, and 34.5%, respectively. The results show that efficiency decreases with increasing erythrosine concentration. Boosting ER concentration can prevent dye degradation by intercepting light availability, encouraging recombination of electron holes, colonizing active spots on catalyst surfaces, and producing a shaded effect. The effect of catalyst content affects the lowest and highest amounts of catalyst needed for degradation. Three concentrations, including 30, 50, and 70 mg of CuMn2O4/g-C3N4 (0.2:1) catalyst were added to an aqueous solution of erythrosine (50 ml at 10 ppm). The results demonstrate that the performance of erythrosine degradation enhances as the catalyst content rises from 30 to 50 mg (Fig. 8d). The improvement in degradation is maximum when the catalyst content is 50 mg, and then the more increase in catalyst content (70 mg) results in efficiency reduction. This occurs as the opacity rises and the dye solution scatters. The effect of pH of the erythrosine solution was studied and the results are observed in Fig. 8e. The results showed that the degradation efficiencies were 87.1, 91.0, and 44.4% at pH = 7, pH = 4, and pH = 10, respectively. The increased efficiency of degradation at acidic and neutral pH values can be attributed to the fact that erythrosine is more stable and reactive in these conditions. At lower pH values, the erythrosine molecule is protonated, which increases its reactivity and makes it more susceptible to degradation. Additionally, acidic conditions can lead to the production of reactive oxygen species (ROS), like hydroxyl radicals, which can further enhance the degradation process. Furthermore, at neutral pH values, the stability of erythrosine is maintained, allowing for efficient degradation without compromising the integrity of the molecule. In contrast, at higher pH values (pH = 10), the stability of erythrosine is compromised due to deprotonation of functional groups on the molecule, leading to decreased reactivity and lower degradation efficiency.

Fig. 8
figure 8

Photocatalytic performance of a different samples of CuMn2O4, g-C3N4 and their nanocomposites, b effect of different organic dyes, c effect of different erythrosine concentrations in the presence of 50 mg CuMn2O4/g-C3N4 (0.2:1), d effect of different dosages of CuMn2O4/g-C3N4 (0.2:1) over 10 ppm erythrosine, e effect of various pH solution, and f effect of different scavengers

Full size image

To investigate the exact mechanisms of pollutants degradation and to determine the primary active species throughout photocatalysis, multiple scavengers were applied to remove the essential active species. Scavengers include 1,4‒benzoquinone for O2•− radicals, ethylenediaminetetraacetic acid for holes (h+), and benzoic acid for HO• radicals (Hosseini et al. 2024a). As demonstrated in Fig. 8f, benzoquinone greatly prevented erythrosine degradation over CuMn2O4/g-C3N4 (0.2:1), but EDTA and benzoic acid influenced erythrosine degradation to various degrees. The results show that O2•− is the primary active species in the elimination of contaminants. However, h+ and •OH also contribute to erythrosine degradation throughout photocatalysis (Sakthivel et al. 2003). The following formula was used to obtain the rate constants (k) for each catalyst (Belghiti et al. 2022):

$$\ln \left( {\frac{{C_{0} }}{C}} \right) = kt$$
(13)

The dye concentrations in this case are denoted by Co and C, respectively, at t = 0 and t (min). Figure 9 provides k for the destruction procedure utilizing various catalysts. For the different photocatalysts, the pollutants degradation rate constants are as follows: CuMn2O4/g-C3N4 (0.2:1) > CuMn2O4/g-C3N4 (0.1:1) > g-C3N4 > CuMn2O4 > CuMn2O4/g-C3N4 (0.4:1) > CuMn2O4/g-C3N4 (1:1) > CuMn2O4/g-C3N4 (0.8:1). The results indicate that the bigger rate constant (k = 0.0257 min‒1) belongs to the maximum photocatalytic efficiency (91.0%).

Fig. 9
figure 9

Kinetics study of a different samples of CuMn2O4, g-C3N4 and their nanocomposites, b effect of different organic dyes, c effect of different erythrosine concentrations in the presence of 50 mg CuMn2O4/g-C3N4 (0.2:1), d effect of different dosages of CuMn2O4/g-C3N4 (0.2:1) over 10 ppm erythrosine, e effect of various pH solution,

Full size image

It is essential to evaluate the photocatalyst’s photocatalytic stability over several degradation cycles in the experiment. To explore its stability in this regard, five runs of repeated cycles of tests were carried out to remove erythrosine using CuMn2O4/g-C3N4 (0.2:1) photocatalyst. Following every cycle of the experiment, CuMn2O4/g-C3N4 (0.2:1) was removed by an ultracentrifuge, rinsed with water, and dried at 70 °C. As demonstrated in Fig. 10a, after five iterations of deterioration, no discernible loss happens. Nevertheless, a partial catalyst deactivation through multiple reuses and photocatalyst loss during recovery may account for the observed small drop in the percentage of deterioration (9.2%). These findings demonstrated that even after several cycles of wastewater treatment, CuMn2O4/g-C3N4 (0.2:1) was still recyclable and effective for the breakdown of contaminants. Figure 10b indicates the XRD pattern of CuMn2O4/g-C3N4 (0.2:1) after five cycle photodegradation. This pattern is formed of cubic structure of CuMn2O4 with JCPDS No. 01–074-2422 and hexagonal structure of g-C3N4 with reference code of 01–087-1526. Table 2 presents a comparison of the photodegradation of various semiconductor compounds. The results indicate that CuMn2O4/g-C3N4 (0.2:1) is competitive with other semiconductor photocatalysts, suggesting that it can be proposed as a new catalyst for wastewater purifying procedure.

Fig. 10
figure 10

Recycle test in the presence of CuMn2O4/g-C3N4 (0.2:1) and b XRD pattern of CuMn2O4/g-C3N4 (0.2:1) after five cycles

Full size image
Table 2 Photocatalytic performance of different compounds
Full size table

Photocatalytic mechanism

Using the following equations, the Mulliken electronegativity theory was employed to compute the valence band (VB) and conduction band (CB) edge potentials of the CuMn2O4 and g-C3N4 in order to show matching of energy levels. The primary reason for the improvement of the photocatalytic performance in CuMn2O4/g-C3N4 nanocomposites is the combined action of CuMn2O4 and g-C3N4, which effectively generate and separate electron–hole pairs through the photocatalysis (Singh et al. 2019).

$${\text{E}}_{{{\text{VB}}}} \, = \,0.{\text{5E}}_{{\text{g}}} \, + \,{\text{X}} - {\text{E}}_{{\text{c}}}$$
( 14)
$${\text{E}}_{{{\text{CB}}}} \, = \,{\text{E}}_{{{\text{VB}}}} - {\text{E}}_{{\text{g}}}$$
( 15)

where Eg and X represent the bandgap and absolute electronegativity, respectively. Ec represents the free electron energy on the hydrogen scale (about 4.5 eV). Scheme 1 illustrates the estimated CB and VB edge potentials of g-C3N4 as −1.13 and + 1.59 eV, and the computed values for CuMn2O4 as -0.25 and + 1.35 eV, respectively. CuMn2O4 and g-C3N are stimulated to form electron–hole pairs when subjected to visible radiation due to their bandgaps. Because the CB position of g-C3N4 is more negative than that of CuMn2O4, photogenerated electrons on its CB readily transfer to CuMn2O4. Nevertheless, the photogenerated holes on g-C3N4’s VB go to CuMn2O4’s VB. This leads to the accumulation of holes and electrons on the VB and CB of CuMn2O4 and g-C3N4, respectively, and their separation from one another. This slowed down the pace at which electron–hole pairs recombined, which increased the CuMn2O4/g-C3N4 heterostructure’s photocatalytic activity. Consequently, Scheme 1 schematically represents a likely S-scheme photocatalytic process of separation of carriers and migration. Both g-C3N4 and CuMn2O4 photogenerated electron–hole pairs separate when exposed to light on the CuMn2O4/g-C3N4 catalyst. The photoinduced holes in the VB of g-C3N4 merge with the electrons in the CB of CuMn2O4. As a result, there will be no more hole and electron recombination. Consequently, the electrons that are still present in the g-C3N4 conduction band efficiently combine with O2 to generate O2•− that oxidizes the dyes directly. Furthermore, O2•− reacts with H2O, reducing further to produce hydroxyl radicals (HO•). Conversely, HO• radicals are produced when a water molecule and HO− combine with the photoinduced h+ in the VB of CuMn2O4 (Chen et al. 2016, Kumar &Rao 2017). Strong oxidative species like HO• and O2•− radicals effectively break down dye molecules into safer and innocuous byproducts like CO2 and H2O. In addition, the photocatalyst’s surface area, electron–hole pair separation, and capacity to absorb light all affect its photocatalytic efficacy.

Scheme 1
scheme 1

Schematic diagram of S-Scheme mechanism of photoinduced electron–hole separation processes

Full size image

Conclusions

A new and effective S-scheme CuMn2O4/g-C3N4 heterojunction with various mass ratios was produced using an ultrasonic-assisted co-precipitation technique. The CuMn2O4/g-C3N4 (0.2:1) composite had the highest activity (91.0% for erythrosine) under illumination. On the other hand, the CuMn2O4/g-C3N4 (0.2:1) rate constant reaches its maximum at 0.0257 min–1. Heterojunctions are highly active due to S-scheme transfer route of photogenerated electron–hole pairs and larger surface area (18.552 m2 g‒1). This method separates photoinduced electron–hole pairs while retaining useful electrons and holes for the reaction. This paper presents a realistic technique for creating unique step-like heterojunctions for environmental protection and energy generation. Furthermore, the photocatalysts demonstrated remarkable recyclability and stability after five different cycles of deterioration.