According to the chemical formula of the NbC, the precursors of Nb and C were weighed in equal molar ratios. Nb and NbCl₅ were used as the Nb sources, and acetylene black (C₂H₂) powder was used as the C source. The eutectic molten salt of the NaCl-KCl was used as the reaction medium, with a mass ratio of 7 : 1 to the aforementioned precursor. Thoroughly mix Nb, NbCl₅, C₂H₂ and NaCl-KCl and place them in a covered alumina crucible. Place the crucible in a tube furnace filled with argon gas and react at 900 °C for 2 h. After cooling to room temperature, take it out to obtain the primary product. Wash the primary product several times with deionized water to obtain the washed product. The cleaned product is dried in a drying oven at 120 °C for 24 h to obtain the final product (NbC). The corresponding preparation flowchart is shown in Figure 1. According to the experimental description, the following reactions (Equations 1, 2) can occur: Nb + 4 Nb V   →   5 Nb IV   in NaCl − KCl (1) 5 Nb IV + C   →   4 Nb V + NbC (2)

In reaction Equation 2, since Nb(V) is generated on the right side, it will continue to react with the remaining Nb until it is exhausted. Therefore, the entire reaction process is equivalent to reaction Equation 3. Nb + C   →  NbC (3)

The main purpose of adding molten salt is that the reaction system can react with Nb to form NbCl₅ even in the absence of NbCl₅ (Yan et al., 2019). The specific reaction (Equation 4) are as follows: Nb + 5 Cl −   →   NbCl 5 (4)

The preparation of the CeO₂ was performed by the most classic polyacrylamide gel method (Hu et al., 2018; Wang et al., 2021). There are two differences from the methods reported in the literature: First, the polymerization of acrylamide (C₄H₇NO₂) and methylene diacrylamide (C₇H₁₀N₂O₂) needs to be initiated through temperature polymerization. Second, in some methods reported in the literature, the use of methylene diacrylamide is not required. Cerium sulfate (Ce₂(SO₄)₃) was used as the Ce source, citric acid (C₆H₈O₇) as the chelating agent, and glucose (C₆H₁₂O₆) as the anti-gel collapse agent. After weighing them in the corresponding proportions, 6.0645 g Ce₂(SO₄)₃, and 4.7282 g C₆H₈O₇ were successively added to 50 mL of deionized water. Subsequently, 9.5958 g acrylamide and 1.9192 g methylene diacrylamide were added almost simultaneously. After the above reagents are completely dissolved, add ammonia water (NH₃·H₂O) to adjust the pH value of the solution to 5. After adjusting the pH value, heat up to 100 °C until a jelly-like gel is obtained. The obtained gel was placed in a drying oven and dried at 120 °C for 48 h to obtain black xerogel. The black xerogel was ground into fine powder and sintered at 900 °C in a tube furnace to obtain CeO₂ nanoparticles. The corresponding preparation process is shown in Figure 1.

The NbC prepared in Section 2.1 and the CeO₂ prepared in Section 2.2 were weighed into the corresponding powders in mass ratios of 25 wt% NbC, 50 wt% NbC and 75 wt% NbC, respectively, and then ground in a mortar for 2 h. The obtained samples were marked in sequence as CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75. The mixture ground in a mortar is sintered at 200 °C for 5 h to obtain CeO₂-NbC-x. The preparation flowchart of CeO₂-NbC-x as shown in Figure 1.

The phase structure of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75 and NbC was measured by a D8 ADVANCE X-ray powder diffractometer (XRD) with Cu Kα radiation. The surface morphology of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75 and NbC were observed by a field emission scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The optical properties of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75 and NbC were analyzed by a UV-2450 type UV-visible spectrophotometer.

To explore the photocatalytic activity of the CeO₂-NbC-x catalyst, OTC-HCl was used as the target pollutant to complete the corresponding photocatalytic experiment. Before the photocatalytic experiment was performed, to eliminate the influence of adsorption on photocatalysis, a half-hour adsorption experiment was conducted. During photocatalysis, a 300 W xenon lamp is used as the light source to simulate sunlight. The experiment was carried out in a self-made reactor. The dosage of the catalyst is 0.5–3 g/L, the pollutant concentration is 25–100 mg/L, and the pH value is 1–7. These parameters are adjusted according to the experimental requirements. The specific details are mainly described in the discussion section of the photocatalytic results. After the photocatalytic experiment began, samples were taken every 10 min until the experiment was completed 60 min later. Each time, 5 mL was sampled. After sampling, the samples were first centrifuged and then the concentrations of contaminants before and after the reaction were measured using a 721 spectrophotometer, which were respectively recorded as C₀ and C_t. The corresponding degradation percentage (D%) can be described by Equation 5. D % = C 0 − C t / C 0 × 100 (5) Where, C₀ and C_t is the initial concentration and the concentration at time t of the OTC-HCl, respectively. To explore the contribution of free radicals in photocatalytic experiments, capture experiments were performed. The EDTA-2Na, IPA and BQ were used to capture the hole, hydroxyl radical and superoxide radical, respectively. The experimental process is similar to the photocatalytic experiment, except that 2 mol/L of capture agent is added to the reaction solution.

In this experiment, NbC, CeO₂, and CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75 were prepared by the molten salt method, polyacrylamide gel method, and heat treatment technology, respectively. To determine the phase structure and purity of the synthesized target products, detailed analyses were conducted on them using an X-ray diffractometer (XRD). XRD patterns of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75 and NbC as shown in Figure 2a. For CO100%, nine obvious diffraction peaks observed, respectively marked as (111), (200), (220), (311), (222), (400), (331), (420) and (422) crystal planes can be assigned to the cubic phase of CeO₂ with the standard JCPDF no. 89–8,436 and space group of Fm-3 m (225). Characteristic peaks without any impurities were observed. Similarly, six diffraction peaks were observed in the XRD pattern of NbC, corresponding to the (111), (200), (220), (311), (222), and (400) crystal planes of the cubic phase of NbC with the standard JCPDF no.89-3,690 and space group of Fm-3 m (225). With the increase of NbC content, the intensity of the diffraction peak of CeO₂ in the CeO₂-NbC-x sample weakens while the intensity of the diffraction peak of NbC increases. Compared with the diffraction peaks of CeO₂ and NbC, the positions of the diffraction peaks of CeO₂ and NbC in the CeO₂-NbC-x sample have slightly shifted towards a lower angle (Figure 2b), suggesting that a special interfacial contact has formed between CeO₂ and NbC. This conclusion requires further elemental mapping characterization for confirmation.

The microstructure, particle size distribution, elemental distribution, and elemental composition of the synthesized samples can be observed through scanning electron microscopy (SEM) and transmission electron microscopy (TEM).

Figure 3a shows the SEM image of the CeO₂. The CeO₂ sample is mainly composed of some fine particles that are approximately spherical in shape, and a small amount of adhesion and agglomeration occurred between the particles. The particles are relatively uniform, with an average particle size of approximately 60 nm (Figure 3b). Figure 3c shows the SEM image of the NbC. The NbC sample is mainly composed of some large and fine irregular particles, with the largest particle diameter reaching approximately 500 nm. When 75% of NbC is coupled with CeO₂, the fine particles slightly increase and adhere to the surface of the large particles as shown in Figure 3d. The CeO₂ and NbC cannot be distinguished from SEM images, so further analysis through TEM characterization is required. There are mainly three reasons why CeO₂ and NbC cannot be clearly distinguished: First, the uneven distribution of components or agglomeration phenomena mask the morphology of other components. The second is visual neglect caused by contrast differences. Thirdly, the locality of the imaging area leads to some particles being unable to be observed.

Figure 4a shows the TEM image of the CeO₂-NbC-75. The CeO₂-NbC-75 sample is composed of large particles and fine particles with obvious agglomeration. The diameter of the largest particle exceeds 500 nm, which is consistent with the results observed by SEM. Due to the obvious agglomeration of fine particles, the spherical outline is not very clear. Figure 4b shows the HRTEM image of the CeO₂-NbC-75. The lattice fringes between the particles are very clear. The particles with lattice spacings of 0.31 and 0.22 nm are attributed to the (111) crystal plane of CeO₂ and the (200) crystal plane of the NbC, respectively. The results confirmed that CeO₂ and NbC were present in the CeO₂-NbC-75 sample, and the particles formed interfacial contact in a special way. Figures 4c–g shows the BF-TEM and element mapping images of the CeO₂-NbC-75. All elements are uniformly distributed on the parent body of the NbC, confirming the formation of a heterojunction between CeO₂ and NbC. To further analyze the purity of the CeO₂-NbC-75 sample, Figure 4h shows the corresponding EDS spectrum of the CeO₂-NbC-75. The EDS spectrum contains elements such as C, O, Ce, Nb and Cu. Due to the use of copper mesh in the TEM test process, the sample contains Cu element, so there are no impurity elements in the sample.

Whether semiconductor materials have a high optical absorption coefficient can be used to determine their potential applications in the field of photocatalysis. To study the optical properties of the CeO₂, NbC, and CeO₂-NbC-x, they were characterized by an ultraviolet-visible spectrophotometer.

Figure 5a shows the diffuse reflection spectrum of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75, and NbC. For all samples, the reflectance drops sharply in the range of 190–250 nm, and increases with the increase in wavelength in the range of 250–350 nm. The difference is that after a wavelength of 350 nm, the reflectance of the CeO₂ sample increases sharply in the range of 350–450 nm and slowly in the range of 460–900 nm. The reflectance of the CeO₂-NbC-25 and NbC samples increased sharply in the range of 350–420 nm, slightly decreased in the range of 420–800 nm, and increased sharply after exceeding 800 nm. Excluding the influence of the test steps, the reflectance of the CeO₂-NbC-50 and CeO₂-NbC-75 samples remained almost constant in the 350–800 nm range and it increased sharply after 800 nm. It is worth noting that after a wavelength of 350 nm, the CeO₂ sample has the highest reflectance, while the reflectance of the CeO₂-NbC-25 and NbC samples is similar. The reflectance of the CeO₂-NbC-50 and CeO₂-NbC-75 samples is also similar. Within the wavelength range of 500–800 nm, the reflectance of the CeO₂-NbC-50 sample is the smallest. The reflectance of CeO₂ and NbC is greater than that of the CeO₂-NbC-50 and CeO₂-NbC-75, indicating that the reflectance of the CeO₂-NbC-x after the coupling of CeO₂ and NbC is not a simple mechanical superposition of the two.

Based on Kubelka-Munk (K-M) Equation 6 (Singh et al., 2026) and the diffuse reflection spectrum of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75, and NbC, the absorption spectra of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75, and NbC obtained, as shown in Figure 5b. F R = α S = 1 − R ∞ 2 2 R (6)

Where, R, α and S is the reflectivity, the absorption coefficient and the scattering coefficient of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75, and NbC, respectively. The absorption coefficients and reflectance of all samples show an opposite trend. The CeO₂ sample has a high ultraviolet optical absorption coefficient, and the NbC sample has a high optical absorption coefficient in both the ultraviolet and visible light wavelength ranges. In the range of 200–350 nm, the absorption coefficient of the CeO₂-NbC-x decreases with the increase of NbC content. In the range of 350–900 nm, the absorption coefficient of the CeO₂-NbC-x does not show a significant linear correlation. The absorption band at 200–350 nm can be assigned to the Ce-O centers in CeO₂ (Zheng et al., 2021). Similar characteristic absorption peaks were also observed in NbC, mainly due to the possible presence of a small amount of C in the sample (Gupta and Pandey, 2019). The absorption observed at 350–900 nm can be associated with the presence of disordered Nb-C centers in NbC (Gupta et al., 2018). Absorption spectrum analysis indicates that the CeO₂-NbC-x has a high ultraviolet-visible optical absorption coefficient, revealing that it can respond to ultraviolet-visible light.

According to the Tauc relationship (Equation 7) (Escobedo-Morales et al., 2026), the absorption spectra of the CeO₂, CeO₂-NbC-25, CeO₂-NbC-50, CeO₂-NbC-75, and NbC can be converted into a curve relationship of (αhν)² ∼ hν or (αhν)^1/2 ∼ hν, as shown in Figures 5c,d. α h ν n = A h ν − E g (7)

Where, h is the Planck constant, ν is the frequency, E.g., is the optical band gap value, and A is a constant. n = 2 and 1/2 is a direct bandgap semiconductor and an indirect bandgap semiconductor, respectively. The CeO₂ and NbC is a direct bandgap semiconductor (Isik et al., 2023) and an indirect bandgap semiconductor (Sivasankeerthana et al., 2024), respectively. Figure 5c shows the, E.g., value of the CeO₂. The intersection point value of the slope at the steepest part of the (αhν)² ∼ hν curve and the abscissa is the, E.g., value. According to the calculation, the, E.g., value of the CeO₂ is approximately 3.09 eV. The, E.g., value of the NbC as shown in Figure 5d. The, E.g., value of the NbC is 1.38 eV, slightly lower than the results reported in the literature (Gurung, 2012). When CeO₂ and NbC are coupled, the, E.g., value of the host lattice does not change. When the NbC content is low, the, E.g., value of the CeO₂-NbC-25 catalyst is 3.09 eV, as shown in Figure 5e. When the NbC content exceeds 25%, the, E.g., value of the CeO₂-NbC-x catalyst is 1.38 eV, as shown in Figure 5f.

Photocatalytic experiments have confirmed that the CeO₂-NbC-x catalyst has high photocatalytic activity for the degradation of OTC-HCl. XRD and elemental mapping characterations confirmed the formation of a special interfacial contact between CeO₂ and NbC. The capture experiment confirmed that the holes, hydroxyl radicals and superoxide radicals are the main active species for the degradation of OTC-HCl by the CeO₂-NbC-x catalyst. According to reference (Bindumadhavan et al., 2013), heterojunctions constructed by grinding and low-temperature sintering are prone to introducing defects, which contribute to enhancing the photocatalytic activity of the system. In this experiment, the formation of a CeO₂-NBC-X heterojunction by grinding CeO₂ and NbC helps to construct a special defect structure, which can promote the photocatalytic degradation of OTC-HCl by the CeO₂-NBC-X catalyst. Therefore, the photocatalytic mechanism of the CeO₂-NbC-x catalyst needs to be further plotted in combination with the band arrangement theory. The conduction band potential (E_CB) and valence band potential (E_VB) of CeO₂ and NbC catalysts were calculated through Formulas 11,12, (Anjitha et al., 2026). E CB = X − E e − 0.5 E g (11) E VB = E CB + E g (12)

Where, X is the absolute electronegativity of CeO₂ and NbC catalysts. E^e = 4.5 eV. The X of the CeO₂ and NbC catalysts can be calculated through Formulas 13,14 X Ce O 2 = X C e ∗ X 2 O 3 (13)

Where, X(Ce) = 2.19 eV, and X(O) = 7.54 eV. X Nb C = X N b ∗ X C 2 (14)

Where, X(C) = 6.27 eV, and X(Nb) = 4.0 eV. The X of the CeO₂ and NbC catalysts is 4.99 and 5.01 V, respectively. According to the calculation, the conduction band potentials of CeO₂ and NbC are −1.055 and −0.180 V, respectively. The valence band potentials of CeO₂ and NbC are 2.035 and 1.200 V, respectively. It can be known from the calculation results that after CeO₂ and NbC couple to form the CeO₂-NbC-x catalyst, they follow the type I band arrangement. Type I band arrangement helps the recombination of charge carriers, thereby reducing the catalytic activity of CeO₂-NbC-x catalysts. However, due to the fact that the interface between CeO₂ and NbC is prone to form an energy barrier (Zhu et al., 2024), it is difficult for electrons to relax from the conduction band of CeO₂ to that of NbC. Instead, they react with the dissolved oxygen in the reaction solution at the conduction band of CeO₂ to generate superoxide radicals. The valence band holes of CeO₂ are prone to transition to the valence band of NbC, but the valence band potential of NbC is only 1.200 V, which is not sufficient for it to react with H₂O/OH⁻ to form hydroxyl radicals. From this perspective, it is contradictory to the results obtained from the capture experiment. Therefore, it is inappropriate to explain the photocatalytic mechanism of CeO₂-NbC-x catalysts through type I band arrangement theory. To further explore the charge carrier transfer and separation efficiency of each catalyst, photoluminescence experiments were performed as shown in Figure 9a. Under the light excitation of 300 nm, CeO₂ can observe two fluorescence emission peaks at 410 and 455 nm. The former is attributed to the transition of electrons between 4f⁰ and 4f¹, while the latter is attributed to the electronic relaxation of F^0* excited State to F⁰ state (Wang et al., 2021). For NbC, only one peak at 455 nm can be observed due to the defect. For CeO₂-NbC-x catalyst, the fluorescence emission peak is almost quenched. The results show that the CeO₂-NbC-x catalyst has a high charge carrier transfer and separation efficiency, which is beneficial for the photocatalytic degradation of pollutants. In addition, after CeO₂ and NbC form the CeO₂-NBC-X catalyst, a special interface defect is formed at their interface. This interface defect is conducive to the directional transfer and separation of charge carriers. According to the literature (Rahim and Rodriguez, 2013; Zhang et al., 2025), NbC possesses metallic properties, which will enable it to serve as a carrier for charge carrier transport when combined with CeO₂, promoting charge transfer and separation within CeO₂ as shown in Figure 9b. The photoluminescence experiment confirmed that this hypothesis was reasonable. When simulated sunlight is shone on the surface of the CeO₂-NbC-x catalyst, the electrons in the CeO₂ valence band will be excited and accelerate their transition to the conduction band of the CeO₂ under the action of NbC, promoting the separation of electrons and holes in CeO₂. Because the conduction band potential of CeO₂ is lower than −0.13 V and the valence band potential is greater than 1.89 V, the conduction band electrons are prone to react with oxygen in the solution to form superoxide radicals, and the valence band holes will react with H₂O/OH⁻ to form hydroxyl radicals. Both the generated superoxide radicals and hydroxyl radicals will undergo degradation reactions with OTC-HCl, generating non-toxic small molecules. In addition, the valence band holes will also directly react with OTC-HCl to form non-toxic small molecules. The specific chemical reactions (Equations 15–20) are as follows: CeO 2 − NbC − x + h ν   →   CeO 2 − NbC − x   e CB − + h VB + (15) e CB − + O 2 → · O 2 − (16) h VB + + OH −   →   · OH (17) · OH + drug  →  Non − toxic small molecules (18) · O 2 − + drug  →  Non − toxic small molecules (19) h VB + + drug  →  Non − toxic small molecules (20)