Introduction

The overconsumption of fossil fuels has caused a significant rise in atmospheric CO2 levels, triggering major environmental issues such as global warming, glacier melting, and rising sea levels1,2,3,4. Various strategies have been explored to reduce the ecological impact of CO2. One strategy is converting CO2 into value-added products or fuels through chemical processes, providing a bifold advantage in mitigating the ongoing energy crisis5. In addition to CO₂ emissions, organic pollutants—including widely used pigments and dyes—contaminate wastewater. Many of these compounds are carcinogenic and nonbiodegradable, posing significant environmental and health risks6. Triphenylmethane dyes, such as crystal violet (CV), have essential uses in several industries, including cosmetics and food production. However, the toxic and carcinogenic nature of these dyes poses severe threats to living organisms7. These compounds also affect aquatic ecosystems by hindering light penetration and disrupting photosynthesis in marine plants8.

Heterogeneous photocatalysis has gained significant attention as a potential solution to these environmental challenges. It offers low-cost and efficient water treatment solutions by using solar energy to breakdown organic contaminants into CO2 and other compounds, which can be converted into valuable products such as methane and methanol9,10. Rare earth oxides (REOs) have emerged as key materials in heterogeneous photocatalysis and various other applications, including fuel cells, rechargeable batteries, biosensors, and computer devices11. Their unique properties enhance their catalytic performance in diverse applications, such as automotive emission control, energy conversion, and environmental remediation. REOs possess strong selective adsorption capabilities and high thermal stability, attributed to the 4f electronic structure of rare earth elements (4fn6s2 or 4fn- [15d16s2), which provides abundant energy levels and multielectron configurations12. The partially filled 4f orbital and nearly empty 5d orbital promote catalytic activity and improve the kinetics of the separation of the photoinduced charge carriers involved in the catalytic activity13.

Graphitic carbon nitride (GCN, g-C3N4) also possesses advantageous properties for photocatalysis, including stability, nontoxicity, visible-light absorption, a narrow bandgap, favorable electrical characteristics, and ease of synthesis14,15,16. GCN can be prepared simply through the thermal condensation of widely available nitrogen compounds, such as melamine and urea17. In photocatalysis, GCN-based materials are widely studied for applications such as CO2 reduction, water splitting, and organic pollutant degradation18,19,20,21,22,23,24,25,26,27,28,29,30. Recently, composites such as Bi12SiO20/GCN, ZnGa1.01Te2.13/GCN, BiOxBry/BiOmIn/GCN, PbBiO2I/GCN, and Sb2MoO6/GCN have demonstrated improved photocatalytic performance, mainly due to the ability of GCN to promote better separation of photogenerated charge carriers8,31,32,33,34.

We present herein the synthesis of a novel YTO/GCN heterojunction (Yb6Te5O19.2/g-C3N4) through a straightforward hydrothermal approach. The synthesized composites were thoroughly characterized via XPS, XRD, DRS, SEM‒EDS, HR‒TEM, and BET techniques and demonstrated high efficiency in photocatalytic applications. Compared to pure YTO and GCN, the YTO/GCN heterojunction exhibited enhanced performance in terms of CO₂ reduction and CV dye degradation. The YTO/GCN photocatalysts displayed excellent chemical stability and durability over multiple photocatalytic cycles. Scavenger, UPS, and ESR experiments were conducted to identify the active species and elucidate the plausible photocatalytic mechanisms. These findings highlight the potential of YTO/GCN photocatalysts for water treatment and CO₂ emission mitigation.

Experimental

Materials

All reagents were utilized without any additional purification as supplied: ytterbium (III) nitrate pentahydrate from Strem Chemicals, Inc.; tellurium tetrachloride, melamine, p-benzoquinone and crystal violet dye from Alfa Assar; DMPO (5,5-dimethyl-1-pyrroline N-oxide) and sodium azide from Sigma‒Aldrich; ammonium oxalate and sodium hydroxide from Shimakyu Chemical; and ammonium acetate, reagent-grade HNO3, HPLC-grade methanol and isopropanol from Merck. Deionized water was used in all the experiments and was obtained via a Milli-Q water ion exchange system (resistivity = 1.8 × 107 Ω-cm).

Apparatus and instruments

A MAC Science MXP18 instrument with Cu-Kα radiation (80 mA, 40 kV) was used for X-ray powder diffraction (XRD) analysis. A JEOL JSM-7401 F instrument (15 kV accelerating voltage) was utilized for field-emission scanning electron microscopy coupled with electron dispersive X-ray spectroscopy (FE-SEM‒EDS). A JEOL-2010 instrument (accelerating voltage of 200 kV) was used for high-resolution transmission electron microscopy (HRTEM) images, energy-dispersive X-ray spectroscopy (EDS), and selected area electron diffraction (SAED). A Thermo Nicolet 5700 Fourier Transform Infrared (FT-IR) spectrometer equipped with a Greasby-Specac advanced overhead (Specaflow) ATR (attenuated total reflection) system P/N 1401 series was used for the ATR and transmission FT-IR experiments. Photoluminescence (PL) experiments were performed on a Hitachi F-7000 instrument, ultraviolet photoelectron spectroscopy (UPS) was performed via ULVAC-PHI XPS, PHI Quantera SXM, and high-resolution X-ray photoelectron spectroscopy (HRXPS) was performed via a ULVAC-PHI instrument. Brunauer–Emmett–Teller (BET) specific surface areas were measured by a Micrometrics Gemini automated system using N2 adsorbate gas at 77 K. Bruker EMX A300-10/12 was used to measure the electron spin resonance (ESR) signals of •OH and •O2− radicals trapped with DMPO. The system was equipped with a microwave bridge (microwave and modulation frequencies = 9.85 GHz and 100 kHz, respectively; modulation amplitude = 1 G; and microwave power = 22.8 mW).

Synthesis of YTO/GCN

YTO was prepared by mixing 1.0 mmol of ytterbium(III) nitrate pentahydrate (Yb(NO3)3∙5H2O) and 1.0 mmol of tellurium tetrachloride (TeCl4) in deionized water (5 mL). The mixture was stirred for 30 min, after which its pH value was adjusted to 12 by dropwise addition of NaOH solution (2 M). The obtained slurry was heated at 200 °C for 48 h in a 23-mL Teflon-lined stainless-steel autoclave. The mixture was allowed to cool to room temperature (rt). The YTO precipitate was filtered, thoroughly washed with absolute ethanol and deionized water, and dried overnight at 60 °C.

GCN was synthesized by calcining melamine in a muffle furnace under atmospheric conditions. Melamine (5 g) was heated in an alumina crucible to 540 °C in 4 h at a rate of 10 °C/min. The mixture was cooled to room temperature, and the graphitized carbon nitride was obtained by grinding the powder in an agate mortar. The general procedure for preparing YTO/GCN composites with varying GCN weight ratios (5%, 10%, 25%, 50%, 75%, and 90%) was as follows: a predetermined amount of GCN was dispersed in 10 mL of ethane-1,2-diol by ultrasonication, followed by the addition of YTO powder and 30 min of stirring at RT. The mixture was heated to 100 °C in a 23-mL Teflon-lined autoclave for 4 h. The precipitate was filtered, washed with deionized water, and dried overnight at 60 °C. The composites prepared in this study are designated YTO-x%GCN, where x indicates the % weight content of GCN.

Photocatalytic experiments

Photocatalytic degradation of CV

The photocatalytic activity of the YTO/GCN composites was evaluated through CV degradation under visible-light irradiation. A 15 W fluorescent lamp, emitting primarily in the 400–700 nm range, was employed as the source of light with the Pyrex reaction vessel at a distance of 30 cm. Each catalytic run included 50 mg of the photocatalyst and 10 ppm CV in 100 mL of water. Adsorption/desorption equilibrium was established before irradiation by stirring the aqueous suspension for approximately 30 min in the dark. Aliquots of 5 mL were taken at designated time intervals. The aliquot was centrifuged to separate the photocatalyst, and the supernatant was analyzed via UV-PDA. The concentration of CV dye was measured at a wavelength of 580 nm.

The scavenging experiments consisted of the addition of a series of 1.0 mM scavengers (1 mL) to a typical photocatalytic run as described above. Isopropanol was used to selectively quench hydroxyl radicals (•OH), benzoquinone for superoxide radicals (•O2−), sodium azide for singlet oxygen species1O2), and ammonium oxalate for holes (h+)35.

Photocatalytic reduction of CO2

Helium gas (pressure = 2 bar) was used to purge the reactor to check for leaks before the reaction. A CO2-compressed steel cylinder (99.99%) was used to deliver CO2 at a 20 mL/min delivery rate into the photoreactor comprising 0.1 g of the photocatalyst in an aqueous NaOH solution (1 N). Samples were withdrawn at determined time intervals of UV light irradiation from the reaction mixture via a syringe and injected into a gas chromatograph (GC). A Thermo Trace 1300 GC with a flame ionization detector (FID) and a thermal conductivity detector (TCD) was used. A TG-BOND-Q column (30 m length, 10 μm film, 0.32 mm ID) was connected to the FID detector to separate the obtained reduction products. Light gases such as N2, O2, CO, and CO2 were separated by connecting a TG-BOND MSieve 5 A column (30 m length, 30 μm film, 0.32 mm ID) to the TCD detector. High-purity He (99.999%) or N2 (99.9995%) were employed as carrier gases, whereas approximately 5000 ppm of various mixed high-purity standard gases (CO, CO2, CH4, etc.) were used to establish a calibration curve for GC analysis.

Results and discussion

Characterization of YTO/GCN composites

Phase structure

Figure 1 presents the XRD patterns of the as-prepared pristine YTO, GCN, and YTO/GCN composites with various mass ratios. The characteristic peak in the GCN sample at 27.4° arises from the stacking of the aromatic groups between the layers, and the peak at 13.1° corresponds to in-plane structural packing36. The diffraction peaks in the pure YTO sample correspond to the Yb6Te5O19.2 phase (JCPDS-00-037-1390). The observed 2θ values were 28.6°, 33.1°, 47.6°, 56.5°, and 59.2°, attributed to the (1 1 1), (2 0 0), (2 2 0), (3 1 1), and (2 2 2) planes of the Yb6Te5O19.2 cubic structure, respectively. The composite samples displayed patterns of both components, YTO and GCN. The GCN diffraction peaks weakened as the YTO content increased, indicating composite formation. The stability and purity of the YTO/GCN composites are further confirmed by the absence of any other impurity peak.

Fig. 1
figure 1

XRD patterns of the as-prepared YTO/GCN samples with different mass ratios of GCN.

Full size image

Figure 2 presents the FT-IR spectra of the samples. The peak observed at 2380 cm− 1 is ascribed to CO2 adsorbed on the sample surface. The YTO spectrum displayed strong absorption between 600 and 800 cm−1, attributed to Yb–OH and Te–O vibrations37,38. For the pure GCN sample, the peak at approximately 810 cm− 1 is attributed to the s-triazine breathing mode, that is, from 1200 to 1650 cm− 1 to C − N heterocycle stretching, and that in the 2800–3400 cm− 1 range corresponds to the N-H bonds of the aromatic groups39. The spectra of the composites show bands from both GCN and YTO, indicating heterojunction formation and aligning with the XRD results.

Fig. 2
figure 2

FT-IR spectra of the as-prepared YTO/GCN samples with different mass ratios of GCN.

Full size image

Morphological structure and composition

FE-SEM was used to examine the morphologies of the pure samples and the YTO-90%GCN composite. The GCN sample (Fig. 3a) has a stacked multilayered structure, in accordance with the literature40. The SEM image in Fig. 3b of the pristine YTO sample clearly shows that the YTO particles have spherical morphologies with diameters of approximately 0.2–0.5 μm, which are composed of numerous small nanocrystallites. The YTO particles, with their distinct morphology, are also visible in the SEM image of the YTO/GCN composites (Fig. 3c). The image illustrates the deposition of spherical-like YTO on the GCN surface, confirming the formation of the composite material. SEM‒EDS analysis further verified the coexistence of YTO and GCN, confirming the presence of ytterbium, tellurium, oxygen, carbon, and nitrogen in the composites.

Fig. 3
figure 3

FE-SEM and EDS of the as-prepared (a) GCN, (b) YTO, and (c) YTO-90%GCN.

Full size image

The FE-TEM-EDS images of the YTO-90%GCN samples (Fig. 4) reveal the interfacial interaction between YTO and GCN. As shown, the lattice spacing of 0.1868 nm relative to the (2 2 0) Yb6Te5O19.2 crystal plane is in close contact with the 0.326 nm spacing associated with the (0 0 2) crystal plane of GCN41. The EDS spectrum reveals the same chemical elements as those in the SEM‒EDS results. The elemental mapping analysis of YTO-90%GCN demonstrated the spatial correlation of the Yb, Te, O, C, and N signals and the close interfacial contacts between YTO and GCN. These results confirm the XRD results and the coexistence of YTO and GCN in the synthesized composites.

Fig. 4
figure 4figure 4

(a) FE-TEM image, (b) HR-TEM, (c) SAED, (d) EDS, and (e) mapping of YTO-90%GCN structures.

Full size image

X-ray photoelectron spectroscopy

Figure 5 shows the obtained XPS spectra of the YTO, GCN, and YTO-90%GCN composites. A consistent chemical composition is clearly demonstrated in Fig. 5a (survey scan XPS spectra), which shows the GCN C-1 s and N-1 s peaks and the Yb-4d, Te-3d, and O-1 s peaks of YTO along with all the corresponding peaks for the composite sample. The survey also reveals the presence of oxygen in pristine GCN, which may result from adsorbed water molecules42. The 4d spectra of Yb were fitted via the Gauss–Lorentz (G–L) function to 7 peaks, as shown in Fig. 5b. The ytterbium compounds have two valence states. The filled 4f shell of divalent ytterbium (Yb2+) in the 4d spectrum is a doublet, whereas the partially filled 4f shell of trivalent ytterbium (Yb3+) is a multiplet43,44. The peaks at approximately 180.9 eV and 188.0 eV are attributed to Yb2+ 4d5/2 and Yb2+ 4d3/2, respectively, and the peaks at approximately 184.5, 186.1, 192.1, 199.0, and 205.2 eV are attributed to Yb3+ because of its high susceptibility to Coulomb interactions43. The Te 3d spectra (Fig. 5c) display peaks at 575.3 and 585.7 eV attributed to 3d5/2 and 3d3/2 transitions from the Te-O phase45. The main peak in the O 1 s spectrum (Fig. 5d) was located at 529.5 eV, corresponding to Yb–O–Yb, with other high-energy peaks corresponding to Te-oxide or H2O adsorbed on the catalyst surface43. The C-1 s spectrum (Fig. 5e) exhibited peaks at 287.6 eV associated with the sp2-carbon in GCN (N–C = N) and 284.3 eV from an external hydrocarbon in the instrument46,47. The N 1 s spectra (Fig. 5f) displayed peaks at 398.1 eV from sp2 N in triazines, 399.1 eV from tertiary amine moieties, 400.2 eV from amines, and 403.8 eV due to charging effects in the heterocycles or positive charge localization48,49. Taken together, these XPS findings confirm the existence of GCN and the successful production of the composite materials.

Fig. 5
figure 5figure 5

XPS spectra of as-prepared YTO, GCN, and YTO-90%GCN samples. (a) Total survey; (b) Yb 4d; (c) Te 3d; (d) O 1s; (e) C 1s; (f) N 1s.

Full size image

Optical absorption properties

Figure 6 shows the UV‒vis diffuse reflectance spectra of the YTO/GCN composites and their colors. The DRS spectra show the absorption edges of YTO and GCN at approximately 330 and 450 nm, respectively, in accordance with their white and yellow colors (Fig. 6a, b). The YTO/GCN composites displayed a slight change in color (white to pale yellow) with increasing GCN content, which led to enhanced absorption in the visible light range, as shown in the spectra. (αhν)n = k(hν − Eg) plots were used to estimate the bandgap energies (Fig. 6c, d)50. α and hν denote the absorption coefficient and the photonic energy, respectively; n is 0.5 or 2 for indirect and direct bandgaps, k is a constant, and Eg is the absorption bandgap energy. Accordingly, the Eg values of the YTO/GCN composites were in the 2.71–2.75 eV range versus 3.77 eV for the non-GCN sample, making the materials appropriate for visible-light activation.

Fig. 6
figure 6

UV-vis absorption spectra of the as-prepared YTO, GCN, and YTO/GCN samples.

Full size image

BET and adsorption‒desorption isotherms

The N2 adsorption‒desorption isotherms and pore size distributions of the samples are presented in Fig. 7. All three samples were mesoporous, showing type IV isotherms with H3-type hysteresis loops36. The Barrett–Joyner–Halenda (BJH) method, estimated from the adsorption branch of the isotherm, further confirmed the mesoporous nature and pore diameter distribution. Table 1 summarizes the average adsorption/desorption pore sizes, pore volumes, and specific surface areas (SBET). The average pore sizes of YTO, GCN, and YTO-90%GCN were ca. 29.01 nm, 33.50 nm, and 37.06 nm, respectively. The SBET values were 2.7634 m2/g, 11.7758 m2/g, and 11.7855 m2/g. YTO-90%GCN has a much larger surface area than pure YTO because of the role of GCN, which restricts YTO particle aggregation in the composite39. The porous structure with a larger SBET enables better contact of the composite with organic molecules, improving their photodegradation.

Fig. 7
figure 7

(a) Nitrogen adsorption-desorption isotherms and (b) the corresponding pore size distribution curve of YTO, GCN, and YTO-90%GCN samples.

Full size image
Table 1 Physical and chemical properties of as-prepared samples.
Full size table

Photocatalytic activity of YTO/GCN composites

This section summarizes the investigation of the photocatalytic activities of YTO/GCN composites in CV degradation (Fig. 8). The highest activity was observed with YTO-90%GCN, which degraded 97.3% of the CV under visible-light irradiation over 72 h (Fig. 8a). The degradation kinetics were examined via ln(Co/C) plots as a function of irradiation time (Fig. 8b), with the rate constants of the composites provided in Table 2. All photocatalysts exhibited pseudo-first-order kinetics for CV degradation51. The rate constant for YTO-90%GCN (0.043 h− 1) was 1.11 and 2.97 times greater than the rate constants of pristine GCN and YTO, respectively. Table 3 compares the photocatalytic activity of YTO-90%GCN for CV degradation with that of recently reported photocatalysts8,33,41,52,53,54,55,56. Both treatment effectiveness and cost must be considered when evaluating water treatment technologies for practical applications. Earlier efficient photocatalysts typically require higher light intensities and larger catalyst loadings. This study utilized low-wattage light sources (15 W), in contrast to the high-wattage sources (hundreds of watts) used in many prior studies, which favor a more cost-effective treatment and demonstrate the efficiency of YTO/GCN materials under low-intensity visible-light irradiation.

Fig. 8
figure 8

(a) Photodegradation of CV dye as a function of irradiation time over YTO/GCN photocatalysts; (b) Photodegradation kinetics of CV dye over YTO/GCN photocatalysts.

Full size image
Table 2 Kinetic parameters (rate constants and linear regression coefficients R2) for photocatalytic degradation of CV at various YTO/GCN samples.
Full size table
Table 3 Comparison of photocatalytic activity of present work with other graphitic carbon nitride composites for crystal Violet dye degradation.
Full size table

The photocatalytic activity of the materials was further explored for CO2 reduction under UV irradiation (Fig. 9). Control experiments conducted in the absence of CO2 or the photocatalyst confirm that the products are exclusively generated through the interaction of CO2, the catalyst, and irradiation. The gas chromatograms of CO2 reduction over the YTO-90%GCN photocatalyst are displayed in Fig. 9(a). The chromatographic peaks were identified by comparison with the mass spectra of standard gases. Methane was the major product, with lower quantities of larger hydrocarbon reduction products. The methane concentration increased with increasing irradiation time (Fig. 9(b–d)). The methane yields over GCN, YTO, and YTO-90%GCN were 0.0459, 0.4189, and 0.4490 µmol g−1, respectively, under irradiation for 216 h, demonstrating the high efficiency of YTO/GCN composites in CO2 conversion.

Fig. 9
figure 9figure 9

(a) Gas chromatograms of CO2 reduction samples, and (b–d) Photocatalytic reduction of CO2 as a function of irradiation time over YTO, YTO-90%GCN, and GCN samples.

Full size image

The stability of YTO-90%GCN was investigated by collecting the composite via centrifugation after photocatalytic degradation of CV and recycling it in subsequent catalytic runs. As demonstrated in Fig. 10(a), YTO-90%GCN has high stability and durability for efficient reuse. In the fifth run, its photocatalytic activity slightly decreased to 85.8% because of the small loss of the solid catalyst associated with its recycling and the aliquot sampling for analysis. The photostability of YTO-90%GCN was also confirmed by the XRD patterns in Fig. 10(b), with no detectable changes between the original and reused YTO-90%GCN photocatalysts.

Photoluminescence emission spectra were recorded for GCN, YTO, and YTO-90%GCN (Fig. 11), as emission results from electron‒hole recombination. A low emission intensity generally indicates a slower recombination rate and higher photocatalytic efficiency40,51. The pure GCN sample displayed a strong emission peak at approximately 458 nm, corresponding to a bandgap of 2.7 eV due to carrier recombination48. Compared with pure GCN, the presence of YTO significantly reduced the PL emission intensity, suggesting better charge carrier separation, slower recombination rates, and thus higher photocatalytic activity in the GCN-YTO heterojunction. The electrochemical impedance spectra (EIS) of the samples were recorded (Fig. S1). The arc radius of the EIS curve indicates the interface layer impedance at the electrode surface. A smaller arc radius signifies a more efficient charge transfer process31. The YTO-90%GCN had a smaller arc as compared with pristine YTO, reflecting faster migration of charge carriers in the YTO/GCN composites. These findings account for the higher photocatalytic activity of the YTO-90%GCN heterojunction than that of pure GCN and YTO under visible-light irradiation.

Fig. 10
figure 10

(a) Cycling runs in the photocatalytic degradation of CV in the presence of YTO-90%GCN, (b) XRD of the sample powder before and after the degradation reaction. The fourth run was repeated in duplicate and mean value was quoted as result.

Full size image
Fig. 11
figure 11

PL spectra of the as-prepared YTO, GCN, and YTO-90%GCN.

Full size image

Pathways of CV photodegradation and CO2 photoreduction

Photocatalytic experiments were conducted with and without various scavengers to identify the main active species in CV degradation over YTO-90%GCN. The results shown in Fig. 12a indicate that benzoquinone (BQ) and isopropanol (IPA), which selectively scavenge •O2− and •OH radicals, respectively, significantly inhibited the degradation process. In contrast, adding selective scavengers for h+ and1O2 (ammonium oxalate (AO) and sodium azide (SA)) caused negligible changes in degradation efficiency57. These findings suggest that the radicals (•O2− and •OH) are the major active species in the present CV decomposition, along with the minor roles of h+ and1O2. ESR spin-trapping with DMPO was used to identify the reactive radicals (Fig. 12(b, c)). In the dark, no ESR signals were observed. Under irradiation, six characteristic peaks of DMPO–•O2− were detected, along with the DMPO–•OH 1:2:2:1 quartet peak58. The hyperfine splitting of these species was determined to be aN = 1.363 mT, aHβ = 0.992 mT and aHγ = 0.383 mT for DMPO-•O2− and aN = 1.500 mT and aH = 1.492 mT for DMPO-•OH. The ESR results corroborate the scavenging experiments, confirming the generation of reactive oxygen radicals under visible-light activation of the YTO-90%GCN photocatalyst.

Fig. 12
figure 12

(a) Photocatalytic degradation of CV dye with YTO-90%GCN catalyst in the absence and presence of scavengers (AO, BQ, IPA and SA) under visible light irradiation; (b) and (c) ESR spectra of DMPO-•OH and DMPO-•O2− using YTO-90%GCN catalyst under visible light irradiation.

Full size image
Fig. 13
figure 13

Mechanism of CV degradation by the S-scheme YTO/GCN heterojunctions under visible light irradiation.

Full size image

Next, we present the plausible mechanism for the YTO/GCN photocatalytic activity in Fig. 13. The edge potentials of the valence and conduction bands (EVB and ECB) were estimated via ultraviolet photoelectron spectroscopy and the following equations:

$$\begin{aligned} E_{{{\text{VB}}}} & = \,\chi - E^{\theta } \, + \,{\text{1}}/{\text{2}}E_{{\text{g}}} \\ E_{{{\text{CB}}}} &= E_{{{\text{VB}}}} - E_{{\text{g}}} \end{aligned}$$

where χ denotes the geometric average of atom electronegativity, Eθ is ~ 4.5 eV, denoting the energy of free electrons on the hydrogen scale, and Eg is the bandgap. The ultraviolet photoelectron spectra (UPS) of YTO and GCN are presented in Fig. S2. Based on UV–vis DRS and valence band spectra, the CB and VB potentials of YTO were determined to be -3.63 eV and 0.14 eV, respectively, while those of GCN were − 1.13 eV and 1.61 eV. This staggered band structure enabled the formation of an S-scheme heterojunction between YTO and GCN (Fig. 13). In the absence of light, electrons naturally transfer from YTO to GCN due to their Fermi level difference, creating an internal electric field at the interface. This field points from YTO to GCN, arising from electron accumulation on the GCN side and depletion on the YTO side. Upon illumination, this built-in field drives photogenerated electrons toward YTO and holes toward GCN. As a result, electrons of the GCN with weak reduction ability and holes of the YTO with weak oxidation ability recombine. Eventually, holes with strong oxidation ability and electrons with strong reduction ability remain on their respective material surfaces. These retained charge carriers possess strong redox potentials, facilitating redox reactions and generating reactive species like superoxide (•O₂⁻) and hydroxyl radicals (•OH). The S-scheme heterojunction thus enhances charge separation and redox capacity, significantly boosting photocatalytic efficiency59,60,61,62,63,64.

As shown in the proposed mechanism, photosensitization and photocatalytic routes can proceed simultaneously, resulting in the photodegradation of CV molecules by YTO/GCN. Both routes can produce electrons that react with oxygen on the semiconductor surface, forming superoxide radicals (•O2−). The latter produce hydroxyl radicals upon reacting with H+, H2O, and electrons. Hydroxyl radicals can be simultaneously formed from the reaction of photogenerated holes with water or OH− ions. These active species degrade CV molecules after several photooxidation cycles, as shown in the equation below65.

$${\text{CV }} + ^{ \bullet } {\text{OH}}/^{ \bullet } {\text{O}}_{{\text{2}}} ^{ - } \to {\text{ degraded compounds}}$$

The plausible mechanism of the photocatalytic reduction of CO2 is illustrated in Fig. 1466. Initially, equilibrium states among CO2, H2CO3, HCO3−, and CO32− adducts are established in aqueous solutions of CO2 according to the pH. Any established state/content can initiate the reduction process, producing adsorbed −C ≡ O on the surface of the catalyst, as shown. Continuous hydrogenation of the oxygen or carbon atoms can form ≡ C−OH and −CH = O, respectively. The latter −CH = O can undergo different rearrangements, yielding several oxygenated final products, such as alcohols and carboxylic acids. Alternatively, the −C(= O)−C(= O)− intermediate can be formed via the coupling of two neighboring −C ≡ O. Hydrocarbon products are formed via dehydration, which removes O atoms after the hydrogenation step, followed by −CH formation, and subsequently −CH2 and −CH3, leading to the major methane product. Other hydrocarbons, such as ethane, ethylene, or acetylene, can be generated by the coupling of any two adjacent −CHx intermediates.

Fig. 14
figure 14

Proposed mechanisms of photoreduction CO2 with YTO/GCN photocatalyst.

Full size image

Conclusions

This study focused on synthesizing YTO/GCN composites for efficient dye photodegradation and photocatalytic reduction of carbon dioxide. The YTO/GCN heterojunction demonstrated enhanced CV photodegradation efficiency, with the optimal rate constant for YTO-90%GCN being 2.97 and 1.11 times greater than those of pure YTO and GCN, respectively. Additionally, YTO-90%GCN achieved a photocatalytic CO2 reduction yield to methane of 0.449 µmol/g in 216 h, surpassing YTO and GCN by 1.07 and 9.78 times, respectively. The enhanced photocatalytic activities are attributed to the formation of an S-scheme heterojunction between YTO and GCN, which reduces the recombination of the photogenerated charge carriers. The main active species in CV degradation were found to be •OH and •O2− radicals according to ESR and scavenger experiments, and the photocatalytic mechanisms were consequently proposed. Overall, YTO/GCN heterojunctions have emerged as effective photocatalysts with promising applications in green CO2 reduction technologies and the mitigation of environmental pollution through the photodegradation of organic contaminants.