In situ synthesis of g-C3N4/TiO2 heterojunction by a concentrated absorption process for efficient photocatalytic degradation of tetracycline hydrochloride

The construction of heterojunctions between semiconductors is a preferred route to improve overall photocatalytic activity. In this work, a facile and feasible method was innovatively developed to one-step prepare g-C3N4/TiO2 heterojunctions via an absorption-calcination process using nitrogen and titanium precursors directly. This method can effectively avoid interfacial defects and establish a tight interfacial connection between g-C3N4 and TiO2. The resultant g-C3N4/TiO2 composites exhibited prominent photodegradation efficiency for tetracycline hydrochloride (TC-HCl) under visible light and simulated-sunlight irradiation. The optimal g-C3N4/TiO2 composite (urea content of 4 g) showed the highest photocatalytic efficiency, which can degrade 90.1% TC-HCl under simulated-sunlight irradiation within 30 min, achieving 3.9 and 2 times increases compared to pure g-C3N4 and TiO2, respectively. Besides, photodegradation pathways based on the role of active species ·O2− and ·OH were identified, indicating that a direct Z-scheme heterojunction was formed over the g-C3N4/TiO2 photocatalyst. The enhanced photocatalytic performance can be attributed to the close-knit interface contact and the formation of Z-scheme heterojunction between g-C3N4 and TiO2, which can accelerate the photo-induced charge carrier separation, broaden the spectra absorption range, and retain a higher redox potential. This one-step synthesis method may provide a new strategy for the construction of Z-scheme heterojunction photocatalysts consisting of g-C3N4 and TiO2 for environmental remediation and solar energy utilization.


Introduction
For the past few decades, water pollution has arisen as one of the most significant threats to human health and the ecological system around the world (Devi and Kavitha 2013;Di Paola et al. 2012). Tetracycline hydrochloride (TC-HCl) is widely applied as a typical antibiotic for the prevention of bacterial diseases. However, a large proportion of TC-HCl with mutagenic and teratogenic effects is difficult to be metabolized and cannot be effectively removed by traditional sewage treatment, resulting in a drawn-out negative impact on the sustainable development of the ecosystem (Hong et al. 2016;Hou et al. 2021;Li et al. 2017;Sheng et al. 2019;Xu et al. 2019). In this regard, photocatalytic technology based on semiconductors came into being owing to its high mineralization efficiency and thorough detoxification of a wide range of pollutants, direct utilization of solar energy, eco-friendly, and low-cost reaction process (Hoffmann et al. 1995;Kubacka et al. 2012).
TiO 2 has been recognized as one of the most prospective candidates for photodegradation due to the unique virtues of strong oxidation ability, good chemical stability, nontoxicity, low cost, etc. (Chen 2009;Fattakhova-Rohlfing et al. 2014), whereas the practical industrial applications of TiO 2 are subject to the broad band-gap, which means that it can be excited by only 4% of the ultraviolet (UV) light in sunlight. Besides, the high recombination rate of electron-hole pairs further affects photocatalysis efficiency (Perera et al. 2012). Enormous efforts have been carried out to break through these inherent drawbacks of TiO 2 by employing metal/nonmetal doping, cocatalyst loading, dye sensitization, morphological regulation, coupling of composite semiconductors, and so forth (Cao et al. 2011;Sun et al. 2008Sun et al. , 2011. Among these, the fabrication of heterojunctions by compounding with narrow bandgap semiconductors has been reported as the preferred strategy for improving charge separation efficiency, solar light utilization, and overall photocatalytic activity of TiO 2 (Wu et al. 2022b). Especially, the Z-scheme heterojunction reserved the higher redox potential displays considerable photocatalytic performance (Chen et al. 2015;Zhang et al. 2020b). Recently, a typical narrow band-gap semiconductor, graphitic carbon nitride (g-C 3 N 4 ), has attracted tremendous interest because of its myriads of advantages like nontoxicity, easy availability, facile fabrication, excellent thermal and physicochemical stability, and unique electronic structure. And most of all, due to its moderate band gap of 2.7 eV, it can make full use of the visible light region (Wang et al. 2015(Wang et al. , 2009, although it also suffers the drawback of rapid electron-hole recombination rate and low quantum efficiency (Mamba and Mishra 2016). In the above, based on the properties of TiO 2 and g-C 3 N 4 and the suitable energy-level matching between them, coupling g-C 3 N 4 with TiO 2 is a preferable route towards extending the light absorption range, enhancing the carriers separation efficiency, and thus remarkably improving the photocatalytic performance Ji et al. 2020;Wang et al. 2021).
Up to now, a large number of research papers on designing g-C 3 N 4 /TiO 2 heterojunction catalysts have been published. Three common strategies were mainly adopted, including the mechanically mixing ready-made g-C 3 N 4 and TiO 2 , in situ growing TiO 2 on as-acquired g-C 3 N 4 , and loading g-C 3 N 4 on TiO 2 . Zhu et al. (Zhou et al. 2015) displayed that mixing as-obtained TiO 2 and C 3 N 4 powders using a simple ball-milling method can improve the dispersion and hybrid conjugation of C 3 N 4 molecule on the surface of TiO 2 and thus enhance the photocatalytic performance. Although the TiO 2 /C 3 N 4 heterojunctions can be obtained directly after physical mixture methods such as grinding and solvent evaporation (Miranda et al. 2013), it is difficult to achieve uniform mixing and sufficient interface contact between two phases because of the impurity contamination or solvent residue. Yan et al. (Yan et al. 2016) developed a hydrothermal process to prepare TiO 2 /g-C 3 N 4 heterojunctions. They first prepared g-C 3 N 4 by direct calcination of urea and then took it into the hydrolysis reaction of TiCl 4 precursors, making the in situ growth of TiO 2 on the surface of g-C 3 N 4 . Nevertheless, the controlled dispersion, desired morphology, and size of TiO 2 nanocrystals grown on the surface of g-C 3 N 4 were confronted with significant challenges due to the inevitable rapid hydrolysis process. Li et al. (Li et al. 2019) obtained the g-C 3 N 4 @C-TiO 2 heterojunctions via the dispersion of g-C 3 N 4 precursors on an as-synthesized carbon-doped TiO 2 surface and then heat treated. However, the polymerization methods of g-C 3 N 4 such as high-temperature sintering tend to lead to agglomeration of TiO 2 nanoparticles, resulting in a rapid deterioration in photocatalytic performance (Zhou et al. 2016). In addition, the complicated and multiple preparation process is always unavoidable and drive up costs. As far as we know, few related comprehensive studies have yet been conducted for directly using nitrogen and titanium precursor to a-pot prepare g-C 3 N 4 /TiO 2 heterostructured composites. From the perspective of large-scale industrial application, questing for a simplified one-step preparation method to address the above issues and acquire high-performance g-C 3 N 4 /TiO 2 heterojunction visible photocatalysts is imperatively needed.
In this work, a new method was developed to in-situ synthesize g-C 3 N 4 /TiO 2 heterojunction via an absorption process using nitrogen and titanium precursors directly. This g-C 3 N 4 /TiO 2 heterojunction preparation process is briefly described as follows: the nitrogen precursors were homogeneously dispersed on amorphous hydrolysis products of titanium precursors (titanic acid or metatitanic acid) driven by the concentration gradient, followed by a facile calcination process, affording nanocomposites of tightly connected g-C 3 N 4 sheets and TiO 2 particles. The as-synthesized g-C 3 N 4 /TiO 2 nanocomposite exhibited outstanding photocatalytic performance in TC-HCl and MB dye degradation, much superior to that of commercial TiO 2 and bare g-C 3 N 4 , which was ascribed to the fabrication of direct Z-scheme heterostructure with wide spectral response range, efficient carrier separation, and high redox capacity. Such a one-step synthesis method of g-C 3 N 4 /TiO 2 through a facile concentrated absorption process will be of interest to the designing of high-performance heterogeneous photocatalysts and the large-scale, environmentally sustainable application for the photodegradation of antibiotic pollutants.

One-step preparation of g-C 3 N 4 /TiO 2 nanoheterojunction
The one-step preparation process is illustrated by the flowchart given in Fig. 1. 5 mL TTIP was firstly added into 75 mL absolute ethyl alcohol, labeled as solution A. Different mass (2, 4 and 8 g) urea were weighed into 10 mL deionized water (DI), labeled as solution B. Thereafter, solution B was added to solution A to form white sol, and the resulting sol was stirred continuously for 30 min followed by stewing for overnight. Then, the solution was centrifuged and dried to obtain a white powder product. After being fully ground and then calcined at 550 °C for 2 h with a heating rate of 10 ℃/min in a muffle furnace, the powder of g-C 3 N 4 /TiO 2 heterojunctions was obtained. The powder samples added by 2, 4, and 8 g urea are marked as UR-2, UR-4, and UR-8, respectively.
As a comparison, the pure g-C 3 N 4 powders were obtained by a facile thermal polymerization method Yan et al. 2009). Generally, 10 g urea was placed in an alumina crucible with a porcelain lid and calcined under exactly the same condition. Finally, the calcined bulk was cooled naturally to room temperature and ground into powders in an agate mortar.

Characterization
X-ray diffraction (XRD) analysis was investigated with a Shimadzu XRD-7000 X-ray diffractometer equipped with a Cu-Kα radiation source (λ = 1.5406 Å) over the 2θ range from 10° to 80°. Scanning electron microscopy (SEM) images were gained using a Zeiss Sigma 300 scanning electron microscope. High-Resolution Transmission Electron Microscope (HRTEM) and elemental analysis were estimated by a JEM-2100 transmission electron microscope. X-ray photoelectron spectroscopy (XPS) spectra were acquired by a Thermo Scientific ESCALAB 250Xi, and the binding energies were calibrated from C 1 s 284.8 eV. UV-vis and UV-vis diffuse reflectance (UV-Vis/DRS) spectra were carried out using a BeiFen Ruili UV-1601 spectrophotometer. Photoluminescence (PL) spectra were taken on a Hitachi F-4600 fluorescence spectrometer.

Photoelectrochemical measurement
The transient photocurrent (TPC) response and electrochemical impedance spectra (EIS) were characterized using an electrochemical analyzer (CHI660E, Shanghai Chenhua, China) in a standard three-electrode configuration. ITO glass substrate with a photocatalyst film, platinum plate and saturated calomel were, respectively, used as the working electrode, the counter electrode, and the reference electrode. 0.1 M Na 2 SO 4 aqueous solution was used as the electrolyte solution.
The electrode sample preparation process is as follows: 5 mg photocatalyst was dispersed in 900 μL ethanol and 100 μL Nafion solution (0.5 wt%) to form a uniform slurry after 30 min of ultrasonic processing. The slurry was dip-coated on a piece of ITO surface (1 cm × 1 cm), then dried at room temperature for 24 h.

Photocatalytic performance and active species measurements
The photocatalytic performance was assessed by degrading the pollutants TC-HCl (20 mg L −1 ) and MB (20 mg L −1 ) solution with 10 mg photocatalyst. Using a 300 W Xe lamp (PLS-SXE 300) to simulate sunlight during the reaction and using a cut-off filter of 400 nm to retain visible-light irradiation. The mixture solution was stirred vigorously for 30 min in the dark conditions to achieve adsorption/desorption equilibrium before irradiation. Every time after a given irradiation time, 2 mL solution was drawn and centrifuged to take away the photocatalyst. Finally, the concentration of solution concentration was recorded by UV-vis spectrophotometry (BeiFen Ruili UV-1601). The above methods were repeated to investigate the effects of pH (4, 6, 8, and 10), the initial amount of photocatalyst (0.2, 0.4, 0.6, and 0.8 g/L), and the initial concentration of TC-HCl (10, 20, 30, and 40 mg/L) on the photodegradation of TC-HCl. Stability tests were carried out as follows: UR-4 samples were filtered and centrifuged after the first photodegradation experiment, and then the samples were dried at 60 ℃ for 12 h. The dried UR-4 samples were used for the second photodegradation experiment. The photodegradation time was set at 30 min for five cycles of the experiment.
The total organic carbon (TOC) contents of the solution were determined by a TOC-V CPH (Thermo Fisher Scientific Vario TOC).
To identify the critical active species, several quenching experiments were also performed. IPA, p-BQ, Na-EDTA, and AgNO 3 were applied on the TC-HCl degradation as scavengers of ·OH (hydroxyl radical), ·O 2− (superoxide radical), h + (hole), and e − (electron), respectively.  (215) crystal faces of anatase TiO 2 (JCPDS 21-1272), respectively. Two peaks at 12.9°and 27.3° were found from the diffraction pattern of the annealed g-C 3 N 4 , which were consistent with the (100) and (002) crystal faces of a typical graphitic structure (Lu et al. 2010). The peaks at 27.3° can also be found in all samples, and they increased with the addition of urea content. In addition, the positions and shapes of TiO 2 characteristic peaks of UR samples were almost consistent with those of standard anatase TiO 2 , indicating that coupling with g-C 3 N 4 did not affect the lattice structure of TiO 2 . Therefore, the results of XRD can infer the g-C 3 N 4 /TiO 2 nanocomposite have been successfully prepared by a one-step concentrated adsorption method.

Structure and morphology characteristics
The XPS measurements were performed to analyze the chemical component and state of the samples. Combining the characteristic peaks of pure TiO 2 and g-C 3 N 4 , signals of elements C, N, Ti, and O were found in the full survey spectra of the UR-4 sample (Fig. 3a), indicating the formation of g-C 3 N 4 / TiO 2 composites. For the C 1 s spectrum in Fig. 3b, the peak at 284.8 eV could be derived from C-C or the adventitious carbon contamination (Dong et al. 2013;Liu et al. 2011). The peaks at 286.2 eV and 288.2 eV of g-C 3 N 4 were attributed to the C-N and (N) 2 -C = N bond species of graphene phase (Ni et al. 2022;Yu et al. 2013a). Three similar peaks with a slight shift could be observed in the spectrum of the UR-4 sample . The N 1 s spectrum of UR-4 in Fig. 3c can be deconvoluted into three different peaks at 398.0 eV, 399.4 eV, and 401.9 eV, presenting in characteristic peaks to sp 2 hybridized N-bonded C group (C = N-C) of the triazine ring within the g-C 3 N 4 structure, the tertiary nitrogen in the N-(C) 3 group, and C-N-H group, respectively (Jiang et al. 2018;Zhang et al. 2012). In comparison with bulk g-C 3 N 4 , the binding energy of N-(C) 3 and C = N-C in the UR-4 sample shifted to lower, and the Ti 2p (Fig. 3e) binding energy shaved a slight shift to higher energies, indicating the existence of the strong electronic interaction between g-C 3 N 4 and TiO 2 in the UR-4 composite and that electrons can be transferred from TiO 2 to g-C 3 N 4 . The O 1 s spectrum provided in Fig. 3d can be fitted to two peaks. The main peak located at 529.7 eV can be ascribed to Ti-O bonding of TiO 2 , and the weak peak located at 531.1 eV might be originated from the oxygen species on the surface, including the adsorbed H 2 O or hydroxyl group (Zhou et al. 2014). The peaks of Ti 2p showed in Fig. 3e centered at about 458.5 eV and 464.2 eV belonged to Ti 2p 3/2 and Ti 2p 1/2 , respectively, which confirmed the Ti 4+ species in the form of TiO 2 nanoparticles (Chai et al. 2012). Furthermore, there were no apparent characteristic peaks for Ti-N(C) could be observed, representing that the nitrogen and carbon elements did not enter the lattice of the TiO 2 .
To further substantiate the successful construction of the g-C 3 N 4 /TiO 2 heterojunctions, TEM and HRTEM characterization of the UR-4 samples was performed. As shown in Fig. 4a, the nearly spherical irregular polyhedral nanoparticles with sizes ranging from 20 to 80 nm were encapsulated by a layered substance and aggregated. And its HRTEM image exhibited distinct lattice fringes with a lattice distance of 0.356 and 0.326 nm, matching well with the (101) crystal plane of anatase TiO 2 and the (002) crystal plane of g-C 3 N 4 , respectively (marked by arrows in Fig. 4b). Moreover, the EDS mapping images (Fig. 4c) indicated the four main elements (C, N, O and Ti) were dispersed uniformly in the samples, suggesting that the g-C 3 N 4 and TiO 2 were tightly bound, with parts of the surface of the TiO 2 being covered by g-C 3 N 4 . From the local magnified image (Fig. 4d), it is evident that the lattice fringes of the g-C 3 N 4 layer were intimately coupled with the vague fringes of TiO 2 nanoparticles, also suggesting that the g-C 3 N 4 layer and TiO 2 were connected by the heterostructures rather than an ordinary physical contact . Thus, in line with the above XRD, XPS, TEM, and EDS spectra, the g-C 3 N 4 /TiO 2 heterostructure has been successfully achieved by a one-step in situ preparation, which demonstrated the good conformity of the concentrated adsorption process.

Optical and photoelectrochemical properties
As shown in the optical photo of the powder sample prepared by the concentrated adsorption process in Fig. 5a, the color deepened with the addition of urea, which corresponded to the increase of the range of photoabsorption region shown in UV-Vis diffuse reflectance spectra (Fig. 5b). The strong peaks of bare g-C 3 N 4 ranged across the ultraviolet and visible regions; nevertheless, the absorption spectra of bare TiO 2 were limited in the ultraviolet region. Upon loading TiO 2 with g-C 3 N 4 , the composites presented an evident hybrid adsorption trait, with a significant increase in light capture in the visible region as the urea content increases, especially with the highest redshift occurring at about 450 nm was found on UR-8. For the direct band-gap semiconductor, we converted the absorption spectra to reflectance spectra using the Kubelka-Munk and roughly estimated the optical absorption bandgap energy (Eg), as shown in Fig. S1. According to the intercept in the Tauc plots of (αhυ) 1/2 versus hυ, the Eg of pure g-C 3 N 4 and TiO 2 were calculated to be 2.7 and 3.2 eV, respectively, which were in line with what has been reported so far ). The Eg values of UR-2, UR-4, and UR-8 were smaller than that of TiO 2 . It can be concluded that the additional absorption in the ca. 400-450 nm region can be contributed to the interaction of g-C 3 N 4 and TiO 2 in the heterojunction. The g-C 3 N 4 as a photosensitizer of TiO 2 particles for promoting visible-light harvest and potentially improving photocatalytic activity. Additionally, the as-synthesized samples appeared the similar absorption features with P25 and the sharp turn in the curves of (αhν) 1/2 vs photon energy, indicating that the g-C 3 N 4 in the composite prepared by the concentrated adsorption process was only coupled with the TiO 2 particles instead of being incorporated into the lattice of TiO 2 , which was in good conformity with the XRD, XPS, TEM, and EDX mapping findings. Such heterogeneous structure between g-C 3 N 4 and TiO 2 would facilitate the separation and transfer of photogenerated e − -h + pairs in both semiconductors, thus improving photocatalytic efficiency.
The PL spectra were adopted to evaluate the separation behaviors and lifetime of light-excited electron-hole pairs in the photocatalyst. The lower PL emission intensity generally means the lower recombination possibilities of carriers and therefore higher photocatalytic activity (Yan et al. 2016;Zhou et al. 2012). Fig. S2(a) depicts the PL spectra of P25, bare g-C 3 N 4 , and as-synthesized samples. The strong peak of bare g-C 3 N 4 around 460 nm originated from the bandband PL phenomenon, which corresponded to its bandgap energy. In comparison with that of g-C 3 N 4 (partial enlargement of Fig. S2(b)), the emission intensities of UR-2, UR-4, and UR-8 were completely quenched, indicating there was a much lower charge recombination probability. It revealed that the energy-wasteful carrier recombination process of g-C 3 N 4 can be greatly inhibited by coupling with the TiO 2 credited to the redistribution of electrons and holes in the heterostructures (Tong et al. 2015). Besides, as shown in the partial enlargement of Fig. S2(c), the emission intensities of as-synthesized samples at 400 nm were also lower than P25, suggesting that the separation and transfer capability of carriers were further increased compared with P25. The results of PL characterization illustrated that the intrinsic band structure of g-C 3 N 4 and TiO 2 has been successfully tailored via the construction of heterojunction by a concentrated absorption process, which would naturally facilitate the separation and migration of photo-induced carriers, and a better photocatalytic performance can be expected.
To further investigate the electronic interactions between g-C 3 N 4 and TiO 2 , the transient photocurrent responses and electrochemical impedance spectroscopy of the P25, bare g-C 3 N 4 , UR-4, and mechanically mixed P25 and g-C 3 N 4 nanoparticles were recorded. As shown in Fig. 6a, the photocurrent response was fast and reversible for all samples under simulated sunlight irradiation. There was a minimal photocurrent response of the bare g-C 3 N 4 due to the rapid charge recombination. It was clear that the photocurrent value of UR-4 was significantly enhanced, being about 1.6 times as high as that of P25 and 2.0 times as high as that of P25 and g-C 3 N 4 mixed nanoparticles. This suggested that under simulated-sunlight irradiation of the P25, g-C 3 N 4 , UR-4, and mechanically mixed P25 and g-C 3 N 4 nanoparticles the as-prepared UR-4 exhibited higher charge separation and transportation efficiency through the interface between g-C 3 N 4 and TiO 2 . In the EIS Nyquist diagram (Fig. 6b), a smaller arc size usually means a lower charge transfer resistance and a more efficient electron-hole separation. Corroborating with the photocurrent results, the smallest radius of the UR-4 samples implied the highest photogenerated charge transfer efficiency. The results revealed that the tight interface created between g-C 3 N 4 and TiO 2 could provide a fast channel for charge migration. Also, the heterogeneous structures could indeed hinder the rapid charge recombination, thus contributing to a noticeable increase in photocatalytic activity.

Photocatalytic performances
To explore the photocatalytic properties of g-C 3 N 4 /TiO 2 heterojunctions prepared via the concentrated absorption method, degradation performance was conducted with TC-HCl and MB as the target water contaminant, under simulated sunlight, and visible light, respectively. Before exposing to the irradiation, the system was stirred continuously under the dark condition for 30 min to ensure an adsorption-desorption equilibrium over the photocatalysts achieved.
The pristine UV-vis spectra used to monitor the concentration of TC-HCl in the solution were shown in Fig. S3(a)-(f), two characteristic peaks of TC-HCl declined with time can be observed, demonstrating that as-prepared samples can completely degrade TC-HCl. Figure 7a presents the dynamic curves of TC-HCl degradation under simulated sunlight. TC-HCl was barely decomposed after light illumination for 30 min in the absence of a catalyst. Both bare g-C 3 N 4 and commercial TiO 2 (P25) exhibited a low degradation capacity, which has only a degradation rate of 62.4% and 45.3% within 30 min, respectively. By contrast, the degradation ratio over g-C 3 N 4 /TiO 2 heterojunctions with different contents of urea reached up to 72.0%, 84.5%, and 80.4% under the same conditions. Fig. S4 recorded the total organic carbon (TOC) results of UR-4 during TC-HCl photodegradation. It showed that about 60% of TOC removal was obtained in 30 min under simulated sunlight, indicating that the reduction in TC-HCl concentration was ascribed to the degradation of organic molecules rather than decolorization or decomposition. A similar trend was observed under visible light irradiation (Fig. 7c). After 60 min of degradation, pure g-C 3 N 4 only degraded TC-HCl by 63.1% and P25 did not possess significant degradation capacities. However, the removed efficiency over UR-2, UR-4, and UR-8 can reach 82.6%, 90.1%, and 84.0%, which indicated that combined with g-C 3 N 4 , the light absorption range of TiO 2 can be expanded from ultraviolet to visible light.
The photodegradation of TC-HCl over original g-C 3 N 4 , P25, and g-C 3 N 4 /TiO 2 heterojunctions was further described with pseudo-first-order kinetics reaction using a simplified Langmuir-Hinshelwood model (Fig. 7b, d), and the reaction rate constants (κ) were calculated. Among them, the UR-4 displayed the best photodegradation properties with the κ of Fig. 7 Photocatalytic degradation TC-HCl curves under simulated sunlight irradiation (a) and visible light irradiation (c), reaction rate constant for degradation TC-HCl under simulated sunlight irradiation (b), and visible light irradiation (d) over P25, g-C 3 N 4 , UR-2, UR-4, and UR-8 nanoparticles 58.9 × 10 −3 min −1 and 35.9 × 10 −3 min −1 under the full solar spectrum and visible light, respectively, which is 3.9 and 2.5 times higher than g-C 3 N 4 (23.2 × 10 −3 min −1 ) and TiO 2 (15.1 × 10 −3 min −1 ) under the full solar spectrum, 2.2 times higher than g-C 3 N 4 (16.4 × 10 −3 min −1 ) under visible light. This result was in good accordance with the characterization discussed above, which demonstrated that the construction of g-C 3 N 4 /TiO 2 heterojunctions by concentrated absorption process has extended wavelength range of light harvest and reduced recombination rate of carriers, leading to an impressive improvement in photocatalytic performance (Bian et al. 2021;Yu et al. 2013b). It should be noted that the content of urea should be prudently controlled. The downward trend of UR-8 performance might be owed to the complete coverage of TiO 2 with excess g-C 3 N 4 , which blocked out the light irradiation and obstructed the transfer of photogenerated carriers to the surface.
As listed in Table 1, a comparison of the photocatalytic performance for TC-HCl degradation was carried out between the UR-4 sample and other similar heterojunction photocatalysts synthesized by various methods. It could be evident that the g-C 3 N 4 /TiO 2 synthesized through the concentrated absorption method exhibited better or comparable photocatalytic performance in TC-HCl degradation.
For actual polluted water, the situation is often more complex. Therefore, the effects of various factors (catalyst dose, pollutant concentration, and pH value) on the on the photocatalytic activity of UR-4 to degrade TC-HCl were explored (Fig. 8a-c). As shown in Fig. 8a, increasing the UR-4 dose from 0.2 g/L to 0.6 g/L caused little change in the removal efficiency and rate constant of TC-HCl. However, when the UR-4 dose was added to 0.8 g/L, the efficiency dropped dramatically. This was because too much catalyst blocked light absorption, thus preventing the full performance of the individual photocatalysts. From an economic and efficiency point of view, 10 mg is the optimum amount of catalyst. Figure 8b shows the effect of initial TC-HCl concentrations on the photodegradation process. As the concentration increased, the rate constant first increased and then decreased. The fluctuation was due to the fact that the number of free radicals produced under the identical concentration of catalyst was constant and limited. Also, the intermediates increased with increasing TC-HCl concentrations, which might absorb some light. Therefore, 20 mg/L was chosen as the optimum TC-HCl concentration. The pH value was inclined to influence the charge distribution on the catalyst surface and the production of active molecules; therefore, it was crucial to study the degradation rate of TC-HCl at different pH conditions. As shown in Fig. 8c, the rate constant of the TC-HCl removal by UR-4 decreased significantly when the pH value increased. The drop might be due to the inhibition of the migration of photo-reduced electrons to the catalyst surface at high pH, resulting in increased recombination of e − and h + . It was revealed that the UR-4 has excellent photocatalytic activity in the acidic state, while its To better demonstrate the stability and recyclability of UR-4, five cyclic degradation experiments of TC-HCl were conducted. Figure 8d depicts that the photocatalytic efficiency of UR-4 did not decrease markedly and still maintained at 91.37% after 5 cycles. And there was little change in the XRD pattern after the cycles as well (Fig. S5). Therefore, the UR-4 was considered to be a promising photocatalyst that can be used repeatedly in practical applications.
Moreover, the potential application of the designed photocatalysts in dye-contaminated water treatment was also evaluated by MB degradation (Fig. S6). It can be stated that the samples prepared by the concentrated absorption method exhibited superior performance to photodegrade MB. Almost all MB was degraded within 20 min by the UR-2 and UR-4 under the full solar spectrum. Especially, UR-2 has the dramatic κ value of 244.1 × 10 −3 min −1 , being 8 and 3.2 times that of g-C 3 N 4 (30.5 × 10 −3 min −1 ) and TiO 2 (77.2 × 10 −3 min −1 ).

Reaction mechanisms
For the purpose of further ascertaining the plausible mechanism over as-prepared photocatalysts in the photodegradation reaction, the quenching experiments were conducted in the system of TC-HCl degradation. To ascertain the active species that play a dominant role in the photodegradation process, AgNO 3 , Na-EDTA, IPA, and p-BQ were used as radical scavengers for holes (h + ), electron (e − ), hydroxide (·OH), and superoxide radicals (·O 2 − ) species, respectively (Qu et al. 2020;Shao et al. 2021). As shown in Fig. 9, the degradation performances were markedly inhibited by the addition of IPA, followed by BQ. However, the effect caused by AgNO 3 and Na-EDTA can be almost ignored. It can be inferred that ·OH and ·O 2 − play a part in TC-HCl photodegradation, and the ·OH was the predominant reactive free-radical, which was clearly in conflict with the type-II heterojunction mechanism. By the principle of the type-II heterojunction, photoinduced h + will accumulate in the valence band (VB) of g-C 3 N 4 , while simultaneously e − will be abundant in the conduction band (CB) of TiO 2 according to the redox potential distribution (Tong et al. 2015). In this case, ·O 2 − radicals cannot be generated by the reaction of e − in the CB of TiO 2 with dissolved oxygen molecules near the catalyst surface, while ·OH radicals cannot be generated by the reaction of h + in VB of g-C 3 N 4 with the water molecules (or surface hydroxyls), because the CB edges of TiO 2 are not negative enough to the standard redox potential of O 2 /·O 2 − (− 0.33 eV vs. NHE) and the VB edges of g-C 3 N 4 are not positive enough to the standard redox potential of H 2 O/·OH (2.27 eV vs. NHE). The traditional g-C 3 N 4 /TiO 2 heterojunction cannot effectively generate ·OH and ·O 2 − , which will result in a lower oxidation ability and photocatalytic activity. Hence, we can certainly conclude that g-C 3 N 4 and TiO 2 do not form the conventional type-II heterojunction (Wu et al. 2022a).
Taken together with the results of MB photodegradation and optical characteristic, the direct Z-scheme heterojunction mechanism of the photocatalysts prepared by a concentrated absorption process was initially proposed and is illustrated schematically in Fig. 10. Under the illumination of simulated solar light, both g-C 3 N 4 and TiO 2 in heterojunction got excited and e − was generated on each CB, leaving h + on each VB. The photoinduced e − in the CB of TiO 2 can be transferred directly to the VB of g-C 3 N 4 across the contact interface and recombine with the h + therein, ensuring the safe separation of e − on CB of TiO 2 and h + on VB of g-C 3 N 4 . This special e − -h + pairs migration path not only efficiently separated photoinduced carriers but also maintained the high redox capability of the heterojunction photocatalyst system . Then, the e − can respond with the reduced O 2 to yield ·O 2 − radicals, and the h + can respond with H 2 O to compose ·OH radicals. Afterward, these active oxygen species (·O 2 − and ·OH radicals) can decompose pollutants into CO 2 , H 2 O, and other small inorganic molecules (Shi et al. 2021).

Conclusions
In summary, the g-C 3 N 4 /TiO 2 direct Z-scheme heterojunction photocatalyst can be one-step fabricated through an absorption-calcination process. The prepared g-C 3 N 4 /TiO 2 composites displayed considerable photocatalytic performance in the degradation of TC-HCl. The enhanced photocatalytic performance was due to the synergistic effect of multiple factors. On the one hand, the one-step preparation route can effectively avoid interface defects and establish close-knit interface contact between g-C 3 N 4 and TiO 2 , thus inhibiting the recombination of photogenerated carriers. On the other hand, by combining of UV activity of TiO 2 and the visible light response of g-C 3 N 4 , the light absorption range can be expanded to utilize more solar radiation for photocatalytic reactions. In addition, benefiting from the well-matched band structure, the Z-scheme heterojunction was constructed between g-C 3 N 4 and TiO 2 , which can facilitate the carrier spatial separation and retain the strong redox capacity. This work promises to be a simple, economical, and efficient way to design photocatalysts for environmental purification.
Acknowledgements Thanks are due to Dr. Yangfan Lu (School of Materials Science and Engineering) for her help on XPS measurement and analysis.
Author contribution All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Renke Bi, Zhe Liu, Jialong Liu, Yijie Shen, and Chutong Zhou. The first draft of the manuscript was written by Renke Bi and Zhiyu Wang. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding The generous support of the National Natural Science Foundation of China (51872258) and the Innovation Fund of the Zhejiang Kechuang New Materials Research Institute (ZKN-20-Z03) is gratefully acknowledged.

Data availability
The authors declare that the data supporting the findings of this study are available within the article.

Declarations
Ethics approval Not applicable.