Increasing awareness of aquatic contamination and environment crisis have spurred explosive research on solar energy conversion and utilization (Li et al., 2017a; Hu et al., 2018; Kayaalp et al., 2019). Wastewater is mostly generated from the spillage of a broad range of organic contaminants, which are extensively present in industrial, agricultural, and household applications (Zhang et al., 2017a; Wang et al., 2017; Shen et al., 2021). Organic reagents, closely linked with industry, are typical by-product of the rapid development of human society (Basith et al., 2018; Sordello et al., 2018). Not only effective contamination degradation but also appropriate applications of the purified water should benefit to the remediation of wastewater. However, due to their complex chemical structure, it is difficult to damage these pollutants, which are intentionally chosen to resist sunlight and oxidation by microorganisms (Zhang et al., 2017a; Saitow et al., 2018). For this reason, removal and lowering of toxicity of pollution effluents from wastewater is impending (Fan et al., 2015; Gao et al., 2018).

Organic pollutants removal through photocatalytic technology has spurred intense research in recent years, since it offers an environmentally friendly route to transfer solar energy into chemical energy for reducing the excessive pollution (Li et al., 2018a; Lu et al., 2018; Zewu et al., 2018; Liu et al., 2019a). The common photocatalyst is TiO₂ for its fascinating features and relatively high photocatalytic capacity (Liu et al., 2019b). A bare TiO₂, howbeit, is excited only under UV light, which attributes to its large bandgap (about 3.2 eV) (Zhang et al., 2016a). Thus, enormous efforts have been devoted to exploring the reaction processes and developing methods to improve the photocatalytic activity and efficiency of TiO₂ (Liu et al., 2019b; Lee et al., 2019). As expected, huge breakthroughs have emerged in the process of preparation, functionalization, and modification of TiO₂-based photocatalysts to enhance the utilization of solar spectrum (∼50% of solar light) for practical applications (Collado et al., 2018; She et al., 2019). In general, modifying TiO₂ with appropriate dopant not only suppresses the photo-induced e⁺-h^- recombination rate but also increases more visible light capacity that is absent with pure TiO₂. Various strategies have been investigated to advance activities and applications of TiO₂, including elements doping (Han et al., 2019), surface modification (Fu et al., 2017), semiconductor heterojunction (Zhen-Dong et al., 2018), and so on (Park et al., 2019).

Carbon materials are impressive candidates due to their striking optical and superior electron-transfer properties (Wang et al., 2016), including carbon dots (CDs; Zhang et al., 2016b; Chen et al., 2017; Sang et al., 2017), carbon nanotube (Li et al., 2017b), graphene (Xia et al., 2019; Xin et al., 2019), and graphene quantum dots (Zhi et al., 2019). Among them, carbon dots with size below 10 nm possess unique merits and have attracted widespread interests to improve the photocatalytic performance of catalysts (Cheng et al., 2016; Zhang et al., 2017b). The introduction of CDs in photocatalysts can extend the absorbance of sunlight and prompt electron transfer and decomposition of photoinduced electron–hole pairs, hence enhance the photocatalytic property of catalysts (Li and Zhang, 2018).

In this study, CD-TN photocatalysts were synthesized via the hydrothermal method and were applied to remove the three different pollutants (CR, RhB, and TC) under visible light irradiation. The morphology structures, chemical states, and optical properties of the prepared photocatalysts were investigated in detail. The photocatalytic performance of the as-prepared photocatalysts was tested by the decomposition of CR, RhB, and TC, which are toxic to human health under visible light irradiation. Additionally, the stability of CD-TN photocatalysts was also evaluated by five successive cycles.

The typical TEM images of prepared TNs, CDs, and 0.2CD-TNs are shown in Figure 1. As shown in Figures 1A,B, TNs present a rectangular sheet-like structure and a lattice fringe of crystal, which can possess a large specific surface area. The TEM image of CDs (Figure 1C) displays a quasi-spherical morphology. Figure 1D shows that CDs disperse on the surface of TNs for 0.2CD-TNs, suggesting the coupling of CDs on TNs.

The light-harvesting capacity of photocatalyst is a vital factor for its photocatalytic performance (Shi et al., 2018; Valenti et al., 2016). To investigate the role of CD loading on the light absorption of TNs, the UV-vis diffuse reflectance spectra (DRS) of as-prepared samples are compared (Figure 2). For pure TNs, an absorption edge at about 400 nm is observed, exhibiting the bandgap of TNs is 3.2 eV, which is in accordance with previous reports (Li et al., 2018b). Compared to TNs, there is no obvious difference in the UV absorption after CD introduction. However, CD-TN photocatalysts reveal an enhanced visible light absorption with the increasing CD content, which directly confirms that the doping of CDs on TNs improves the visible light–harvesting capacity.

The crystal phase of synthesized samples was measured by X-ray diffraction (XRD) (Figure 3). The spectrum of CDs shows an obvious broad peak at around 29.7°, which is contributed to (200) planes of graphitic carbon (Qu et al., 2016). All the characteristic peaks in TNs and CD-TN photocatalysts are identical to the crystal planes of anatase TiO₂ (PDF No.00–021-1272). Notably, no characteristic peak for CDs is detected in CD-TN photocatalysts with the amount of CDs from 0.1 to 0.3, which may ascribe to the relatively limited CD loading and low diffraction intensity in the composites (Shi et al., 2017). The peaks of CD-TNs with different CDs are consistent with TNs without any change, indicating that there is no obvious influence on the structure of TNs by loading CDs.

The X-ray photoelectron spectroscopy (XPS) technique was conducted for CD-TNs to delve the valence band and the related elements, the results are illustrated in Figure 4. A survey scan (Figure 4A) depicts that the prepared CD-TN photocatalysts contain C, O, N, and Ti elements. Figure 4B shows the C1s spectrum, where the energies binding at 283.5, 284.9, 287.3, and 288.2 eV are attributed to C-Ti, C-C&C=C, C-O, and C=O, indicating the coexistence of CDs and TNs in the CD-TN catalysts. The Ti 2p spectrum shown in Figure 4C is deconvoluted into two signals: the bind energy observed at 457.4 eV is ascribed to Ti 2p_1/2 and the peak at 463.1 eV comes from Ti 2p_3/2. The O1s spectrum (Figure 4D) divided into three peaks at 528.6, 529.6, and 531.6 eV are corresponding to Ti-O, C-O, H-O, suggesting chemical interaction occurs between the two components.

Meanwhile, based on the Tauc plot equation, the bandgap energy of the 0.2CD-TNs is estimated. As illustrated in Figure 5, the bandgap energy of 0.2CD-TNs is 1.89 eV, implying that prepared 0.2CD-TN photocatalysts can generate active species for degrading pollutants under visible light irradiation in theory.

The FTIR spectra of samples are shown in Figure 6. Three peaks emerged at 3,426, 1,633, and 1,383 cm⁻¹ in all samples correspond to the stretching of O-H, C=C, and C=O, signifying the good hydrophilicity of CDs (Que et al., 2017). The peaks at 3,218 and 1,406 cm⁻¹ in the FTIR spectrum of CDs are assigned to N-H and C-N = bonds, suggesting the presence of carboxylic and amide groups on the surface of CDs. In the FTIR spectrum of 0.2CD-TNs, a stronger peak centered at 1,383 cm⁻¹ can be ascribed to the C=O group, and a new peak emerged at 1,257 cm⁻¹ is assigned to the C-O-C bond. Furthermore, the characteristic peaks of CDs and TNs are detected, demonstrating that CDs are successfully introduced onto TNs.

The photoresponse property of CDs is shown in Figure 7. When the excitation wavelength is 350 nm, the emission peak is located at about 440 nm. Significantly, CDs display bright fluorescence under UV-light irradiation, suggesting CDs possess the up-conversion property. CDs with the up-conversion property can absorb visible light and emit shorter wavelengths, which help CD-TNs utilize solar energy more efficiently.

For photocatalyst activity, the separation and transfer of photo-induced electron–hole in photocatalyst is considerably significant (Zhang et al., 2017c). Transient photocurrent and EIS are typical measurements used to delve the electron–hole excitation effect. As illustrated in Figure 8A, the arc radius of 0.2CD-TNs is narrower than that of TNs, indicating lower charge transfer resistance for 0.2CD-TN photocatalysts, which accelerates the charge transport at the interface of 0.2CD-TNs. Moreover, a uniform and increased photocurrent can be captured under light (Figure 8B), which verifies that CDs can effectively promote the charge transfer and migration, demonstrating a higher capacity in photocatalytic degradation of the CD-TN photocatalysts than that of TNs.

The photocatalytic performance of prepared samples was tested via degradation of CR, RhB, and TC under simulated visible light irradiation (Figures 9A–C). It can be seen that in the presence of TNs, the concentration of CR, RhB, and TC changes negligibly under visible light. In contrast, CR, RhB, and TC are degraded remarkably in the same irradiation time in the presence of CD-TN photocatalysts. Additionally, the results reveal that the photocatalytic activity of CD-TNs greatly depends on the CD loading. The removal rate of CR is increased from 7.3 to 94.6% with the amount of CDs from 0 to 0.2 M. However, further increasing the amount of CDs from 0.2 to 0.3 M decreases the rate from 94.6 to 25.3%. Figures 9B,C depict the effect of CD-TNs on the removal of RhB and TC. The RhB degradation efficiency is measured to be around 97.2% under 150 min of irradiation in the presence of 0.2CD-TNs. Similarly, the TC degradation efficiency is achieved to be 96.1% for 0.2CD-TNs after 60 min of visible light irradiation. The experimental results also revealed that TC degrades fast because of the unstable fragments of TC (Supplementary Table S1, Supporting information). Among CD-TN catalysts, 0.2CD-TNs have the highest degradation efficiency, which probably contributes to that the high CD contents which will decrease catalytic active sites and enhance the charge carriers’ recombination rate (Giovannetti et al., 2015; Xie et al., 2018).

The kinetic fitting plots of ln (C/C₀) against reaction time t for all photocatalysts are shown in Figures 10A–C. The results show that the degradation reactions of CR, RhB, and TC by CD-TN photocatalysts fit to pseudo first order kinetics. The formula can be expressed as follows: ln ( C t / C 0 ) = − k t , (1) where k is the kinetic rate constant, C₀ is the initial concentration of pollutants, and C_t is the concentration of pollutants at different degradation time t. The parameters of photocatalytic degradation of CR, RhB, and TC by different photocatalysts are summarized in Table 1. For CR degradation, the degradation activity of 0.2CD-TN photocatalysts with k = 0.01409 min⁻¹ displays 26.09 times higher than that of pure TNs. Furthermore, the k value of 0.2CD-TNs is much larger than other CD-TN photocatalysts under visible light, illustrating its outstanding capacity in pollutants removal. Moreover, similar trend can be observed for RhB and TC, and the degradation capacity for the three pollutant follows the sequence of 0.2CD-TNs >0.1 CD-TNs >0.15CD-TNs >0.25CD-TNs >0.3CD-TNs. It is apparent that 0.2CD-TNs show the highest reaction rate constants of 0.01542 and 0.02546 min⁻¹ for RhB and TC, respectively.

Besides, the BET surface area and total pore volume of 0.2CD-TNs are summarized. As shown in Table 2, the BET specific surface area and total pore volume of 0.2CD-TNs are 129.6286 m²/g and 0.244819 cm³/g, respectively, which can be attributed to the large surface area of TNs.

The prospective applicability of CD-TNs for the practical stability was measured by degrading CR on the basis of the recycling experiments. As shown in Figure 11, after five recycling measurements, the CR degradation rate still remains at about 93.3%, and the photocatalytic activity of 0.2CD-TNs has no loss during the degradation process, indicating that the 0.2CD-TNs exhibit an excellent structural stability. Moreover, the XRD patterns of samples after reaction are shown in Supplementary Figure S1. It can be seen that there are no noticeable changes in the crystal structure for CD-TNs before and after degrading, which further verifies the robust stability of CD-TNs.

In Table 3, the photocatalytic performance of CD-TNs is compared with other catalysts reported previously. Obviously, taking into account the different experimental parameters, CD-TNs exhibit outstanding photodegradation activity among these catalysts.

Practical photocatalysts (CD-TNs) were synthesized via the hydrothermal method. The results show that all the CD-TN photocatalysts exhibit higher photocatalytic efficiency for pollutants degradation under visible light irradiation than that of TNs. For these samples, 0.2CD-TN photocatalysts present the highest reaction rate and more than 90% pollutants are removed after visible light irradiation. This excellent photocatalytic activity in CD-TN photocatalysts can be contributed to the merits of CDs that not only extend the light absorption but also impede the recombination and promote the charge transfer. Consequently, the results demonstrate that the CD-TN photocatalytic process is a promising technology for wastewater remediation and purification.